CCaappaabbiilliittyy BBuuiillddiinngg ffoorr tthhee MMaannuuffaaccttuurree ooff PPhhoottoovvoollttaaiicc SSyysstteemm CCoommppoonneennttss iinn DDeevveellooppiinngg CCoouunnttrriieess
Anna Bruce
School of Photovoltaics and Renewable Energy Engineering The University of New South Wales Sydney, Australia
A thesis submitted to the University of New South Wales in fulfilment of the requirements of the degree of
Doctor of Philosophy
December 2007 Originality Statement
‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’
Signed ……………………………………………......
Date ……………………………………………......
i Abstract
The manufacture of photovoltaic (PV) system components has a role to play in the industrialisation and poverty reduction strategies of developing countries. It has also been suggested that small scale local manufacture of balance of systems components has the potential to improve the maintenance, installation and use of the technology. However, PV is a complex technology and most developing countries have not been able to build the capabilities required to manufacture PV system components of an appropriate quality and price, either in the modern or small scale sectors. The factors that determine the success of PV manufacturers in developing countries are therefore of interest. Previous studies on learning in the PV industry have been focused on industry-wide concerns and have not explicitly addressed enterprise-level capability building or challenges specific to developing countries. In particular, there has been very little published about small scale PV manufacture. This thesis therefore aims to improve understanding of the factors that influence capability building, with a view to assisting decision making in relation to PV manufacture in developing countries. The aims of the study have been fulfilled by the development and assessment of a software simulation training tool for PV cell production line engineers, the development of an analysis framework, and application of it to several case study PV enterprises. Through the application of the framework to the case studies, it has been possible to assess the role of software simulations, the suitability of countries with different types of infrastructure for hosting PV manufacturing and the institutional arrangements or interventions that could be used to promote capability building for PV manufacturers in developing countries. While further case studies are required to make more than tentative conclusions, the framework developed and tested in this thesis may now be used as a tool to systematically and rapidly analyse the appropriateness of different types of PV manufacture in particular countries, to identify the weaknesses in their PV technological systems and therefore to suggest where resources should be invested and where appropriate institutional changes could be made. The simulation software has been demonstrated to be an effective capability building tool, thus providing one of the key elements required for successful manufacturing.
ii Acknowledgements
I first wish to express my thanks to Giles for his love and support throughout the PhD years and I look forward to a wonderful post-PhD future together.
My supervisor Muriel deserves a medal for her patience, our many fruitful discussions and multiple careful edits, which have helped to bring some order into this work. Perhaps a bottle of wine will do!
I wish to thank Susan Kinne and the members of Grupo Fénix, Suni Solar and the Barefoot College for generously giving their time and assistance for this research. I also gratefully acknowledge the rare opportunity Suntech has provided to carry out a case study in a commercial environment.
I would also like to acknowledge the wonderful people at the PV Centre and the Centre for Energy and Environmental Markets at UNSW: Rob Passey for his interesting opinions and for all the lunches and pep talks, Alistair Sproul for many important discussions about energy efficiency and related topics, Hugh Outhred for his input into my theoretical framework, all the CEEM lunchtime meeting crew, the friendly and efficient PV office staff and especially Stuart Wenham for inspiring me to return to the university, giving me the opportunity to work on the Virtual Production Line and sitting through many interviews for the Suntech case study.
I also want to thank my gorgeous friends, who are all having babies now, for their belief in me, their support and especially for the good times. May they continue forever, babies and all!
Finally, I wish to acknowledge my parents, Maurice and Jane, who have, as always, shown great courage during a difficult year, and my Grandparents, Dicky and Grandpa, who have been wonderful role models.
iii Glossary of Terms
Development: Positive change. Industrialisation: A process of increasing capitalisation and more efficient use of labour and other resources, particularly through the mechanisation of production. Although more industrialised countries have progressed through mass production, automation and increasingly computerisation, ultimately relying less on manufacturing to generate income, the use of technology remains an important factor. Developing Countries: Under the UN system, countries may choose their status as ‘developed’, ‘developing’, etc. In the context of this thesis, ‘developing countries’ is taken to mean all countries apart from Japan, Canada, the United States, Australia, New Zealand, and Western Europe. Industrialising Countries: Countries that are technologically self-reliant in some standard modern technologies but remain dependent on advanced countries for technology, especially innovative technology. Enterprise: An enterprise (or "company") is a collection of organizations and people formed to create and deliver products to customers. An enterprise may be a business, service, or membership organization; consist of one or several establishments; and operate at one or several locations. Latecomer: A firm or country that enters an already established technological system (technological latecomer). May also refer to a firm or country that is not yet industrialised (industrial latecomer). Catching Up: The process of approaching the technological frontier. May refer to the acquisition of firm-level technological capabilities or the development of the national technological system. Knowledge: Justified true belief, distinct from information in that it has been interpreted and internalised. In this document, knowledge refers more broadly to human embodied skills, knowledge and expertise. Tacit Knowledge: Knowledge which is not easily codified (documented). Often used to refer to knowledge which is not codified. Explicit Knowledge: Knowledge which is codified (documented). Learning: In common usage refers to the acquisition of knowledge or skill. In economics, traditionally refers to improved performance, particularly labour productivity or cost reductions per unit output. In this thesis, refers to technological learning, which is the acquisition of technological capabilities. Technology: The skills, knowledge and procedures for doing useful things. May be embodied in equipment, documents and organisations, as well as people. Innovation: As defined by Schumpeter, bringing a new or improved product to market. In this document, taken more broadly to mean new creations of economic significance, which may be embodied in people, equipment, documents or institutions. Technological Learning: The acquisition of technological capabilities. In this thesis, the terms ‘technological learning’ and ‘technological capability building’ are used interchangeably. Learning By Doing: Learning that arises automatically through experience in carrying out an activity. Learning By Searching: Learning that arises through intentional searches for new knowledge or solutions to problems, whether internal to a firm, or external. Learning By Interacting: Learning through interactions with other actors in innovation systems. Learning By Using: Learning from the feedback of users of technology.
iv Technological Paradigm: A technological outlook which is a branch among the evolution of technological development. The paradigm includes perceptions of problems that need to be solved and the range of technologies that can provide solutions. It therefore constrains and enables technological development. Technological Trajectory: The activity of technological progress along the economic and technological trade-offs defined by a paradigm. Technological Capabilities: The ability to acquire, use and build on technology productively, resulting from the possession of the appropriate knowledge, procedures, documents and physical capital that embody the technology. National Technological The ability of nations to acquire, use and build upon technological Capabilities: knowledge, not equivalent to the sum of the capabilities of firms, but also dependent on the effectiveness of innovation systems within which firms operate. Technological Capability Acquisition of technological capabilities by a firm or technological Building: system. In this thesis, the terms ‘technological learning’ and ‘technological capability building’ are used interchangeably. Innovation System: A system of actors, institutions, and their interactions, within which innovations are created. Technological System: An innovation system for a particular technology or group of technologies. Institutions In Laws, rules and norms which govern the relations between Technological Systems: individuals or groups. Actors (Organisations) In Organisations that influence or participate in technological learning. Technological Systems: Networks In Innovation The sum of the linkages between actors in innovation systems. Systems: Clusters: Networks between firms, customers and suppliers in close proximity (usually physical proximity, but may refer to cultural or technological proximity). Public Good: Goods which are not depleted by use, so after consumption by one individual, are available for another’s consumption. In the case of pure public goods, it is not possible to exclude people from using the good. Externality: A cost or benefit not accounted for by the market.
v List of Acronyms
PV Modules, Systems and Components: PV Photovoltaic(s) BOS Balance Of Systems (components in a PV system) Wp Watts peak (rated capacity of PV measured under standardised test conditions) SHS Solar Home Systems LED Light Emitting Diode CFL Compact Fluorescent Lamp SLI Starting, Lighting, Ignition (battery) IV Curve Current-Voltage Curve Voc Open Circuit Voltage Isc Short Circuit Current
Circuits: DC Direct Current AC Alternating Current PWM Pulse Width Modulation FET Field Effect Transistor PCB Printed Circuit Board MOSFET Metal Oxide Semiconductor Field Effect Transistor IC Integrated Circuit
Organisations: IEA International Energy Agency UN United Nations NGO Non-Governmental Organisation GEF Global Environment Facility WB World Bank ESCO Energy Service Company UNSW The University of New South Wales PV-GAP International PV Standards Organisation
vi Solar Cells, Materials, Manufacturing Techniques and Products: c-Si Crystalline Silicon (solar cells) mc-Si Multicrystalline Silicon (solar cells) sc-Si Single Crystal Silicon (solar cells) a-Si Amorphous Silicon (solar cells) Cz-Si Czochralski Silicon (single crystal Si produced using the Czochralski method) CIS/CIGS Copper-Indium-Selenide / Copper-Indium-Gallium-Selenide (solar cells) CdTe Cadmium Telluride (solar cells) GaAs Gallium Arsenide (solar cells) LGBC Laser-Grooved, Buried Contact (solar cells) EFG Edge-defined Film-fed Growth (silicon ribbon solar cells) HIT Hetrojunction with Intrinsic Thin Layer (solar cells) p-n junction Junction between p-type and n-type silicon material p-type Positively doped silicon material n-type Negatively doped silicon material AR coating Antireflection Coating SiNx Silicon Nitride (antireflection coating) APCVD Atmospheric Pressure Chemical Vapour Deposition PECVD Plasma-Enhanced Chemical Vapour Deposition EVA Ethyl Vinyl Acetate RTV Room Temperature Vulcanising (silicone) BIPV Building-Integrated Photovoltaics
vii Business, Technology, Investment and Manufacturing Terms: CEO Chief Executive Officer CTO Chief Technical Officer R&D Research and Development FDI Foreign Direct Investment PR Progress Ratio LR Learning Rate IP Intellectual Property IT Industry Information Technology Industry QC Quality Control JIT Just-In-Time (logistics) OEM Own Equipment Manufacture ODM Own Design and Manufacture OBM Own Brand Manufacture TNC Transnational Corporation BDS Business Development Services TT Technology Transfer VAT Value Added Tax IPO Initial Public Offering (of shares)
Barefoot College SWRC Social Work Research Centre (Barefoot College) REW Rural Electronics Workshop BSE Barefoot Solar Engineer VEEC Village Energy and Environment Committee
Other: UV Ultraviolet VPL Virtual Production Line
viii Table of Contents
CHAPTER 1. INTRODUCTION TO THE STUDY ...... 1 1.1. Problem Statement ...... 2 1.2. Aim and Scope...... 4 1.3. Thesis Overview ...... 6 REFERENCES...... 7
CHAPTER 2. THE PV INDUSTRY AND MODERN-SECTOR MANUFACTURING IN DEVELOPING COUNTRIES...... 9 2.1. The Role of Industrialisation in Development...... 10 2.1.1. Appropriate Technology for Industrialisation ...... 10 2.1.2. The Challenge of Catching Up ...... 12 2.1.3. A Special Role for PV Manufacture...... 12 2.2. Markets and Firms in the International PV Industry ...... 14 2.2.1. The Evolution of Markets for PV ...... 14 2.2.2. The Location and Concentration of PV Firms...... 16 2.2.3. Types of Firms in the PV Industry...... 18 2.3. Competitive Strategies and Barriers in the International Photovoltaics Industry.... 19 2.3.1. Expansion of Market Share ...... 19 2.3.2. Cost Reduction...... 20 2.3.3. Quality, Product Differentiation and Brand Equity ...... 20 2.3.4. Technology Leadership ...... 21 2.3.5. Vertical Integration...... 23 2.4. PV Cell and Module Manufacture in Developing Countries ...... 25 2.4.1. The Status of PV Module Manufacture in Developing Countries ...... 25 2.4.2. The Status of PV Cell Manufacture in Developing Countries...... 26 2.4.3. Sources of Technology for Cell and Module Manufacture in Developing Countries...... 27 2.4.4. Advantages and Barriers for Firms in Developing Countries...... 30 2.5. Conclusion ...... 32 REFERENCES...... 33
CHAPTER 3. LEARNING IN THE INTERNATIONAL PV MANUFACTURING INDUSTRY ...... 37 3.1. A Review of the Literature on Technological Learning and Technological Change in Photovoltaic Cell Manufacture...... 38 3.1.1. Top-Down Studies: Trends in Cost Reductions...... 38 3.1.2. The Distribution of Costs in PV Production...... 42 3.1.3. Bottom-Up Studies: Sources of Cost Reductions ...... 43 3.1.4. The Rate and Direction of Technological Change in Photovoltaics ...... 51 3.1.5. A Summary of Technical Challenges for PV Cell Manufacturers in Developing Countries...... 54 3.2. National Experiences with PV Manufacture ...... 55 3.2.1. The Case of Germany...... 55 3.2.2. The Case of Japan...... 58 3.2.3. The Case of the US...... 61 3.2.4. The Case of Australia ...... 64 ix 3.2.5. Factors Influencing the Success of PV Industries in Different Countries...... 66 3.2.6. Virtuous Circles of Policy Induced Development...... 75 3.3. Conclusion...... 77 REFERENCES ...... 79
CHAPTER 4. THE PV INDUSTRY AND SMALL SCALE MANUFACTURE IN DEVELOPING COUNTRIES ...... 83 4.1. The Role of Electricity from Photovoltaics in Development...... 84 4.1.1. Energy and Poverty...... 84 4.1.2. The Impacts of Electricity on Poverty...... 86 4.1.3. Decentralised PV Systems and Development ...... 86 4.2. The Role of Small Scale Enterprises in Developing Countries...... 88 4.3. Markets and Actors in the PV Industry in Developing Countries ...... 92 4.3.1. PV Markets in Developing Countries ...... 92 4.3.2. Actors in the Small Scale PV Industry...... 92 4.3.3. Government Ministries...... 95 4.3.4. Financiers...... 96 4.4. Technology for Small Scale BOS and Module Manufacture in Developing Countries...... 97 4.4.1. Charge Controllers ...... 97 4.4.2. Failures of Charge Controllers ...... 99 4.4.3. Fluorescent Lamps ...... 99 4.4.4. Failures of Fluorescent Lamps ...... 100 4.4.5. Problems in the Manufacture and Use of Electronic BOS Components...... 100 4.4.6. Modules...... 101 4.4.7. Module Failure Modes ...... 102 4.5. Experiences with Small Scale BOS and Module Manufacture in Developing Countries...... 104 4.5.1. Bolivia...... 104 4.5.2. Brazil...... 105 4.5.3. China...... 106 4.5.4. Indonesia...... 106 4.5.5. Kenya ...... 107 4.5.6. Kiribati and Tuvalu ...... 108 4.5.7. Nepal...... 109 4.5.8. Sri Lanka...... 109 4.5.9. Zambia ...... 110 4.5.10. Zimbabwe ...... 110 4.6. Barriers and Advantages of Small Scale Manufacture ...... 112 4.6.1. High Upfront Costs ...... 112 4.6.2. Quality of the Hardware...... 113 4.6.3. Local Manufacture and Local Technical Capabilities...... 115 4.6.4. Local Manufacture and Project-Based Diffusion...... 116 4.6.5. Local Manufacture and Diffusion via Cash-Markets ...... 117 4.6.6. Non-Technical Capabilities of Small Scale Manufacturers ...... 117 4.6.7. Supporting Small Scale PV Manufacture...... 117 4.7. Conclusion...... 119 REFERENCES ...... 121
x CHAPTER 5. A FRAMEWORK FOR THE ANALYSIS OF CAPABILITY BUILDING IN DEVELOPING COUNTRIES ...... 127 5.1. Knowledge and Technology ...... 128 5.1.1. Knowledge...... 128 5.1.2. Tacit and Explicit Knowledge ...... 128 5.1.3. Knowledge Creation...... 129 5.1.4. Innovation...... 130 5.1.5. Technology...... 130 5.2. Technological Capabilities ...... 131 5.2.1. Types of Technological Capabilities ...... 131 5.2.2. A Typology of Capabilities ...... 133 5.3. Technological Learning...... 134 5.3.1. Improved Performance via Learning by Doing ...... 134 5.3.2. Traditional Views of Technology and Technological Change ...... 135 5.3.3. Technological Change via Evolutionary Processes ...... 136 5.4. Types of Learning ...... 138 5.4.1. Learning by Doing...... 138 5.4.2. Learning by Searching...... 139 5.4.3. Learning by Interacting ...... 141 5.4.4. A Learning Framework ...... 143 5.5. Capability Building in Modern Sector Latecomers ...... 145 5.5.1. Equity-based Technology Transfer...... 146 5.5.2. Subcontracting...... 147 5.5.3. Purchase of Technology ...... 148 5.5.4. Hiring and Training ...... 149 5.5.5. Imitation ...... 150 5.5.6. Research Collaborations and International R&D ...... 151 5.5.7. Early Entry and Technological Niches ...... 151 5.5.8. Supplier Interactions...... 152 5.5.9. Clusters and Special Industrial Districts...... 152 5.5.10. Barriers and Advantages to Capability Building Strategies...... 153 5.5.11. The Role of Learning and Technological Systems Literature in the Analysis of Capability Building...... 153 5.6. Capability Building in Small Scale Enterprises ...... 155 5.6.1. Learning by Searching...... 156 5.6.2. Subcontracting...... 157 5.6.3. Purchase of Equipment...... 157 5.6.4. Small Scale Technology Development...... 157 5.6.5. Support for Small Enterprises...... 158 5.6.6. South-South Technology Transfer...... 160 5.6.7. Barriers to Small Scale Capability Building...... 161 5.6.8. Lessons Learnt in Small Enterprise Technology Development and Support...... 162 5.6.9. Concluding Remarks ...... 166 5.7. Technological Systems...... 167 5.7.1. Evolutionary Learning and Technological Trajectories...... 167 5.7.2. Innovation Systems...... 168 5.7.3. Technological Systems ...... 169 5.7.4. Components of Technological Systems and their Roles...... 169 5.7.5. Networks and Institutions in Developing Countries...... 174 5.7.6. Small Scale Enterprises and Technological Systems...... 175 5.7.7. The Role of Policies in Influencing Learning...... 177 5.8. A Capability Building Framework...... 180 5.9. Conclusion ...... 182 REFERENCES...... 184
xi CHAPTER 6. A VIRTUAL PRODUCTION LINE FOR THE MANUFACTURE OF SCREEN-PRINTED SOLAR CELLS...... 189 6.1. The Value and Development of the Virtual Production Line...... 190 6.2. Processing Virtual Solar Cells with the Virtual Production Line ...... 193 6.2.1. Saw Damage Removal Etch...... 195 6.2.2. Texturing...... 196 6.2.3. Rinse and Acid Clean...... 197 6.2.4. Diffusion ...... 198 6.2.5. Diffusion Oxide Removal ...... 199 6.2.6. Plasma Edge Isolation...... 199 6.2.7. Antireflection Coatings...... 200 6.2.8. Screen Printing and Firing of the Metal Contacts ...... 202 6.3. Simulation of Quality Control Tests ...... 205 6.3.1. Simple Tests which Return Stored Parameters ...... 205 6.3.2. Tests Which Call PC1D ...... 206 6.4. Educational Features of the VPL...... 210 6.4.1. Saved Settings...... 210 6.4.2. Help Files...... 210 6.4.3. Assignments...... 210 6.5. The Effectiveness of the Virtual Production Line...... 212 6.5.1. Software Simulations as Learning Tools...... 212 6.5.2. Measuring the Effectiveness of the Virtual Production Line at UNSW...... 212 6.5.3. The Effectiveness of the VPL at Suntech Power...... 214 6.6. Conclusion...... 216 REFERENCES ...... 217
CHAPTER 7. CAPABILITY BUILDING AT SUNTECH POWER ...... 219 7.1. Background to the Case Study ...... 220 7.1.1. PV Manufacturing in China ...... 220 7.1.2. Inputs to the PV Industry in China...... 225 7.1.3. Recent Industry Developments ...... 227 7.1.4. The Chinese Market for PV ...... 229 7.1.5. Support for the PV Industry in China...... 231 7.2. Suntech Power ...... 234 7.2.1. Origin and Start Up of Suntech Power...... 234 7.2.2. Production Capabilities at Suntech ...... 235 7.2.3. Innovative Capabilities...... 239 7.3. Capability Building Strategies at Suntech...... 243 7.4. Analysis of the Case Study Using the Framework...... 252 7.4.1. The Development of Chinese PV Technological System ...... 252 7.4.2. Learning at Suntech ...... 257 7.5. Conclusion...... 263 REFERENCES ...... 264
xii CHAPTER 8. CAPABILITY BUILDING AT GRUPO FÉNIX...... 267 8.1. Background on Nicaragua ...... 268 8.1.1. Rural Electrification in Nicaragua ...... 268 8.1.2. Markets for PV in Nicaragua...... 269 8.1.3. Small PV Enterprises in Nicaragua ...... 272 8.2. Background to Grupo Fénix ...... 274 8.2.1. History and Philosophy of Grupo Fénix...... 274 8.2.2. Grupo Fénix Activities ...... 276 8.2.3. Group Fénix Staff ...... 277 8.3. Capabilities at Grupo Fénix...... 279 8.3.1. Production Capabilities and Activities at Grupo Fénix ...... 279 8.3.2. Innovative Capabilities and Activities at Grupo Fénix...... 286 8.3.3. Investment and Linkage Capabilities at Grupo Fénix...... 288 8.4. Analysis of the Case Study Using the Framework ...... 294 8.4.1. The Nicaraguan Technological System for Small Scale PV Manufacture ...... 294 8.4.2. Learning at Grupo Fénix...... 297 8.5. Conclusion ...... 304 REFERENCES...... 305
CHAPTER 9. CAPABILITY BUILDING AT THE BAREFOOT COLLEGE ...... 307 9.1. Background on India ...... 309 9.1.1. Rural Electrification in India ...... 309 9.1.2. Support for PV Rural Electrification in India...... 310 9.1.3. A Critique of the Indian PV Programmes...... 312 9.1.4. Rural Infrastructure and the Small Scale PV Industry...... 313 9.1.5. NGOs and the Panchayat System ...... 314 9.2. Background on the Barefoot College...... 317 9.2.1. History & Philosophy of the Barefoot College...... 317 9.2.2. Organisation of the Barefoot College...... 319 9.2.3. Solar Energy Projects ...... 322 9.2.4. Impacts, Sustainability and Replicability ...... 334 9.3. Capabilities at the Barefoot College...... 336 9.3.1. Production Capabilities...... 336 9.3.2. Innovative Capabilities ...... 338 9.3.3. After-Sales Service Capabilities...... 340 9.4. Analysis of the Case Study using the Framework...... 342 9.4.1. The Indian Technological System for Small Scale PV Manufacture ...... 342 9.4.2. Learning at the Barefoot College...... 345 9.5. Conclusion ...... 352 REFERENCES...... 353
xiii CHAPTER 10. DISCUSSION AND CONCLUSIONS ...... 357 10.1. Technological Systems for PV Manufacture in Developing Countries...... 358 10.1.1. Resources for Production and Innovation ...... 359 10.1.2. Opportunities and Incentives for Investment in Production...... 364 10.1.3. Knowledge Creation and Exchange ...... 367 10.1.4. Incentives for Innovation ...... 368 10.1.5. Interactions Between Networks, Policies and Technological Trajectories...... 372 10.2. Capability Building Strategies in PV Manufacturing Enterprises in Developing Countries...... 375 10.2.1. Learning By Doing, Production Capabilities and Routines ...... 376 10.2.2. Learning by Searching and Innovative Capabilities...... 381 10.2.3. Learning by Interacting and Linkage Capabilities ...... 383 10.3. Conclusion...... 387 10.3.1. Tentative Conclusions about Capability Building for PV Manufacturers in Developing Countries...... 387 10.3.2. The Value of the New Framework, Limitations of the Study and Further Research ...... 389 REFERENCES ...... 391
Appendix 1: Virtual Production Line CD
Appendix 2: Experiences with the Diffusion of Photovoltaics in Developing Countries
Appendix 3: An Introduction to Solar Cells
Appendix 4: Processing Algorithms for the Virtual Production Line
Appendix 5: Suni Solar Module Encapsulation Procedure
Appendix 6: Cost of Materials for Grupo Fénix Module Manufacture
Appendix 7: SWRC/EU/UNDP Training Modules 2000
Appendix 8: Barefoot Solar Engineer Final Test Paper – Theoretical Component
Appendix 9: Estimate of Cost of Equipment for a New Rural Electronic Workshop (REW) in Ladakh in 2000
Appendix 10: Details of repairs carried out on PV Systems installed in Leh district (1993-94)
Appendix 11: Details, Parts and Costs for Systems Produced in the Barefoot College REWs
xiv
CChhaapptteerr 11.. IInnttrroodduuccttiioonn ttoo tthhee SSttuuddyy Chapter 1. Introduction to the Study
1.1. Problem Statement
Photovoltaic (PV) systems have been widely used to supply electricity to remote rural areas of developing countries1. They are often the least cost solution to rural electrification and generally require less maintenance than conventional technologies such as fossil fuel generator sets. However, the technological and institutional capability for the maintenance and operation of PV systems is not easy to build locally, while system components are generally imported from industrialised countries and are therefore expensive and require foreign currency. The local manufacture of PV system components has the potential to contribute to a more sustainable use of the technology in developing countries through the avoidance of imports, lower cost manufacture, and improved technical knowledge and skills, availability of spare parts and access to information locally. There is a generally decreasing level of complexity in the technology for the manufacture of photovoltaic system components, from the conversion of sand to silicon, crystallisation and cutting into wafers, device fabrication, module assembly and balance of systems components (BOS) manufacture. The manufacture of silicon and solar cells requires advanced, research-based technological capabilities, whereas the manufacture of modules can rely on technology that is standard, or acquired externally. The manufacture of BOS for small DC systems can be more simple and experience-based. The manufacture of technologically complex PV system components, such as cells, may have a role to play within the industrialisation strategies of more advanced developing countries, particularly in view of PV’s technological proximity to the electronics industry, its export potential, and its potential to provide energy to remote areas, contribute to energy security, and reduce the use of foreign currency for energy. The manufacture of components that involve more simple technology, such as BOS, may provide opportunities for poverty reduction via the establishment of small industry in a range of developing countries. The manufacture of BOS components in developing countries is of particular interest because it has the most potential to increase local knowledge, skills and access to spare parts through the interactions of manufacturers, technicians and users in close geographical and social proximity. While PV cell and module manufacture may be a beneficial industry for some developing countries, enterprises in most countries have had difficulty employing the technology, and the majority of the global manufacture of PV cells and modules still occurs in
1 While there is no established convention for the use of the term ‘developing country’ (for example, under the UN system, countries may choose their status), Japan, Canada, the United States, Australia, New Zealand, and Western Europe are consistently considered "developed". The remainder of countries are usually considered “developing”, and will be referred to as such in this thesis. It should, however, be noted that the East Asian Tigers (Hong Kong, Singapore, South Korea and Taiwan) are considered to be developed countries in some classifications, such as that of the IMF, while some Eastern European and former Soviet Union countries are not classified as either “developed” or “developing” in some instances. The terms “emerging” and “transitional” economies are also sometimes used. 2 Chapter 1. Introduction to the Study
industrialised countries. The factors influencing the successful use of PV manufacturing technology in developing countries are therefore of interest. While previous studies have been carried out on technological learning and innovation in the PV cell manufacturing industry, these studies have been focused on industry-wide concerns, such as predicting learning rates and the direction of technological change. They have not specifically addressed enterprise-level capability building strategies or learning mechanisms. Some other studies have analysed the success of different countries in promoting the development of a PV industry, but they have not linked national-level factors to enterprise capabilities and have not identified challenges specific to developing countries. There is therefore a lack of knowledge about both the enterprise-level and national-level determinants of capability building in the manufacture of PV cell and modules in developing countries. Despite speculation about the potential benefits of local manufacture of BOS, particularly at a small scale and local level, there has been very little published about the capabilities required for the manufacture of PV system components in small-scale enterprises or the conditions under which they may be built. There is also little solid evidence about the extent to which local manufacture can contribute to local development and improve the quality and sustainability of the use of the technology. This thesis aims to fill these gaps in the current knowledge and assist decision making in relation to industry development efforts.
3 Chapter 1. Introduction to the Study
1.2. Aim and Scope
This study aims to identify suitable capability building strategies, enabling environments, and interventions to support the manufacture of PV system components in developing countries. The aim of the study will be fulfilled through: The development of an analytical framework that facilitates assessment of the factors that influence capability building, both within enterprises and externally. The application of the framework to several case studies in order to identify the typical requirements and constraints of PV manufacturing enterprises in developing countries, successful capability building strategies, characteristics of enabling environments and therefore opportunities for intervention. The development and assessment of a software training tool for photovoltaic cell production line engineers, which can be used as a means of technology transfer as well as capability building in developing countries.
A review of the literature on PV manufacture has revealed a gap in the knowledge about capability building within enterprises. The focus of the research is therefore at the enterprise level, covering research questions such as: How can enterprises build capabilities? Why are some environments more enabling than others? How can enterprises be supported by interventions? The analysis framework developed for this research is therefore centred on technological learning processes within enterprises. It has drawn upon concepts and frameworks from the literature on technological capability building in developing countries, organisational learning, and innovation systems. Case study research and histories are deemed to be appropriate tools for explanatory studies where the questions of how or why something occurs are central to the research. Survey- based and archival studies, conversely, are most appropriate to questions of what, how many, how much (Yin, 1989). Case studies have therefore been selected as the primary method in this research. These case studies are designed to be what Stake (2000) calls instrumental, in that they facilitate understanding of a general case, and hence provide a step towards theory building. The manufacture of PV cells is more technologically complex, and hence requires a different set of capabilities to the assembly of modules or BOS components. Small-scale and large-scale producers operate with a different set of resources and capabilities, as do those in rural, compared to urban locations. In order to include variety across these attributes, it is necessary to study the capability building of different types of enterprises. Critical test cases
4 Chapter 1. Introduction to the Study
ranging from small-scale manufacture using simple technology to large-scale high-tech manufacture have been selected, as illustrated in Table 1-1.
Table 1-1: The Selection of Cases for the Study Technology Scale Location Simple Medium Complex Small Large Rural Urban Suntech Power, China Grupo Fénix, Nicaragua Barefoot College, India
For the purposes of this thesis, small-scale manufacturers are defined be small in size; to operate locally, utilise local resources and produce products for local markets; using relatively small amounts of capital and simple production methods. They may be either rural or urban. Modern sector manufacturers are located in urban areas, are large scale, service international markets and use complex technologies. Suntech Power, a solar cell and module manufacturer in China has been chosen as a revelatory modern-sector case study, since a high level of capabilities for cell manufacture has been rapidly and successfully built. Suntech Power has expanded rapidly to become the fourth largest manufacturer of solar cells in the world by 2006. Two small-scale cases have been selected: the Barefoot College in India, and Grupo Fénix in Nicaragua. The Barefoot College manufactures electronic balance of system components for PV systems within a network of rural workshops, closely linked to the community of users. Grupo Fénix supports a small business that manufactures balance of system components and assembles modules at a cottage industry scale in an urban workshop and sustains an unregistered rural enterprise that assembles modules. The two cases offer the opportunity to study differences in ownership structure, scale, technology linkages, location, resources and capabilities. By using a case study approach to build an evolutionary picture of the learning processes through which PV enterprises have built capabilities, the research identifies the types of capabilities that have been important and the factors that have influenced successful capability building. Through analysis of the case studies in the context of what is known from the literature, the findings are tentatively generalised, providing the opportunity to suggest effective policies, institutional arrangements and support programmes.
5 Chapter 1. Introduction to the Study
1.3. Thesis Overview
In chapter 2 of this thesis, the role of PV manufacture in the industrialisation strategies of developing countries is established. The status of markets, enterprises and technology in the PV industry are described, identifying barriers to the entry of manufacturers from developing countries. In chapter 3, existing studies on technological learning and development in the PV industry, and the experiences of different countries in fostering a successful PV industry are reviewed, establishing the need for further research on enterprise-level capability building in the PV industry. Chapter 4 reviews experiences with the manufacture, dissemination and use of PV technology for rural electrification in developing countries, concluding that there is a role for the small scale manufacture of PV system components in developing countries and identifying barriers to the successful commercialisation of the technology. In chapter 5, a framework is developed for the analysis of capability building in PV manufacturers in developing countries. The framework identifies types of learning that build different types of capabilities, and the factors within the external technological system that influence learning. Chapter 6 documents the development and use of a virtual solar cell production line for the training of production line engineers and assesses its effectiveness and application in the context of Suntech Power. In chapter 7, learning at Suntech Power, a PV cell manufacturer in China, is analysed through the structure of the framework. Through the study of how capability building occurred in this case, explanations are proposed for the success of Suntech and the Chinese technological system that may be applicable more broadly. In chapters 8 and 9, two cases of small scale manufacture of PV system components are examined through the framework. The first is the Barefoot College in India and the second is Grupo Fénix in Nicaragua. The ways that learning occurred within these enterprises, and the ways in which the technological system supported and constrained them are identified. Chapter 10 analyses the case studies with reference to the existing literature in order to identify the typical requirements and constraints of PV manufacturing enterprises in developing countries, successful capability building strategies and characteristics of enabling environments. Suitable policies and interventions to support manufacturers are suggested. The thesis concludes in chapter 10 by identifying the contribution that this work has made to the knowledge about PV manufacturing in developing countries and discussing the effectiveness of the framework and the software that have been developed and their usefulness for future industry development efforts.
6 Chapter 1. Introduction to the Study
References
Stake, R.E. (2000), Case Studies, in Denzin, N.K. & Lincoln, Y.S. (eds), "The handbook of qualitative research", Sage Publications, Thousand Oaks, Calif., pp xx, 1065 , [1057] p. Yin, R.K. (1989), Case study research : design and methods, Rev. ed, Sage Publications, Newbury Park, [Calif.].
7 8 CChhaapptteerr 22.. TThhee PPVV IInndduussttrryy aanndd MMooddeerrnn--SSeeccttoorr MMaannuuffaaccttuurriinngg iinn DDeevveellooppiinngg CCoouunnttrriieess
The purposes of this chapter are to identify the role for modern-sector photovoltaics manufacture in developing countries; to provide background on markets, firms and technology in the international PV industry and specifically in developing countries; and to demonstrate that technological learning1 is a critical factor for competitiveness in the industry. Section 2.1 describes the role of industrialisation in the development strategies of countries, and investigates the suitability of PV manufacture as an appropriate industry for developing countries. Section 2.2 describes the evolution of markets for PV and the types and locations of firms in the industry. Section 2.3 analyses competitive strategies in the PV industry. Section 2.4 then details the current status of PV cell and module manufacturing in developing countries, and on the basis of the preceding analysis, identifies advantages and barriers for ‘latecomer’ firms entering the PV industry.
1 Technological learning is the acquisition of the ability to acquire, use and build on technology productively. The concept of technological learning is explored in detail in chapter 5. 9 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
2.1. The Role of Industrialisation in Development
Prior to industrialisation, most societies lived a subsistence existence, and lived through continual cycles of famine. Industrialisation has enabled higher productivity via improved efficiency, and can be a source of increased food production, improved medical care, and communication technologies that can contribute to a healthier, more fulfilling life. Industrialisation is also seen as an avenue to solve the problems of low export earnings that result from low prices obtained in international markets for many primary products (Todaro, 1982). Industrialisation, however, often degrades the natural environment, and the process of modernisation may be socio-culturally destructive, particularly as rural communities migrate to urban areas. Industrial development is not a panacea for the elimination of poverty, as the very poor are usually isolated from processes of industrialisation, and economic growth is not often accompanied by increased equality (Ingham, 1993). There is therefore a duality in current approaches to development, with recognition that national economic growth will not automatically enable those in extreme poverty to improve their productivity; but also recognition that for those individuals, groups and nations with sufficient capital, industrialisation provides opportunities for further investment in infrastructure, health, education and safety nets. Despite the destructive nature of industry, and its limits in addressing extreme poverty, industrialisation is essential for developing countries to widen their development base and meet growing needs, and will doubtless be a part of the development strategy of all countries (Bruntland, 1987).
2.1.1. Appropriate Technology for Industrialisation Industrial development approaches have evolved since World War II. Development efforts immediately after World War II were largely directed at the reconstruction of Europe and emphasised growth in productivity via savings and capital investments (Wanmali, 1998). The process of industrialisation is closely related to the acquisition of technology, and there is therefore intense interest in the acquisition of technology by developing countries. Following the success of the reconstruction of Europe, it was believed that developing countries would be able to benefit from the experiences of industrialised countries and follow the same path to industrialisation, using the same technologies. It was believed that the task of catching-up would be easier and faster than it had been in the development of industrialised countries, resulting in convergence (Juma & Clark, 2002). Progress in achieving modern rates of economic growth, however, was slow for most developing countries, despite efforts to industrialise. There was a pattern of divergence, rather than convergence appearing.
10 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
In the 1960s, the route of industrialisation via technologies from advanced countries was seen to be inappropriate since developing countries were comparatively rich in labour and sometimes in raw materials, but poor in capital resources (Juma & Clark, 2002; Todaro, 1982). Mainstream economists hence viewed labour-intensive, rather than capital-intensive techniques for medium and large scales of production as appropriate for developing countries (Grieve, 2004; Stewart & James, 1982). Underpinning this idea was the assumption that labour-intensive, low-capital techniques were either available, or could be developed. Bhalla (1975) recommended “wherever technologies appropriate to the factor endowments of developing countries do not exist, they should be created through massive research and development” (Bhalla, A.S. 1975 cited in Grieve, 2004, p 174). Studies beginning in the 1970s, however, found that the development of technology does not depend primarily on factor availability and prices, but on a complex interaction of industry and location-specific variables. The technology appropriate to all available factors of production is therefore not necessarily available or likely to be developed. Dahlman & Westphal (1982, p 109) believe that “it is now well established that there is scope for choosing between techniques with differing levels of labour intensity and productivity, but that the scope is by no means uniform.” With technological improvements, industry is becoming more capital and less labour- intensive, and the importance of cheap labour as a source of competitive advantage is also declining. Factors of production such as unskilled labour and primary products do not provide sustained advantage. Industries that rely on such factors are called ‘footloose’, since firms often move to a more favourable location if it becomes more profitable to do so. The jobs, growth and other inter-industry spillovers associated with industrial development are therefore not sustained. Slow rates of progress toward industrialisation in developing countries have more recently been exacerbated by increasingly rapid technological change. The demand for increasing diversity and more widespread quality of products drives the innovation process. Facilitating this technological change is the increasing globalisation of markets and production chains, made possible by improved transport and communications technologies and the use of information technology to coordinate distributed and flexible manufacturing systems (Sharif, 1992). As technological innovations change more rapidly, they are also becoming more costly to develop and more pervasive. In the absence of efficient labour-intensive production techniques, the lack of sustained advantage gained from cheap labour and the need to continually change in order to satisfy consumer preferences, innovation is accepted as the most important source of competitive advantage and productivity growth. Firms in developing countries cannot rely on labour-
11 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
intensive production methods and must develop innovative high technology industries if they are to access the most profitable parts of global value chains.
2.1.2. The Challenge of Catching Up Since innovative industries are the source of sustained advantage, developing countries must move from simple technologies to complex ones and from imitation of existing technologies to innovation, in order to successfully industrialise. Both incremental and radical change may be part of competitiveness in a rapidly changing environment. Enterprises need to employ new technology, whether to improve incrementally, or to innovate radically. Continual learning is therefore the basis for sustaining the capabilities required by enterprises to keep up with change and to compete (Dutrénit, 2004). Hobday (1995) developed the idea of the latecomer firm which enters an already established technological system (technological latecomer), and described the process of technological catch-up required as ‘followers’ attempt to approach the ‘leaders’. Catching up has proved elusive for the majority of developing countries, while the rare success of the east-Asian tiger economies has been referred to as the east-Asian miracle. Latecomers must build the capabilities to find and use technology, to adapt it to local conditions and eventually to be able to build on it to create new technology. These capabilities are collectively referred to in this thesis a ‘technological capabilities’. Technology and technological capabilities cannot be easily transferred. The development of capabilities is a path dependent process, and depends on the technological capabilities already possessed, the learning processes undertaken and the external technological system (Lall, 1992). The development of technological capabilities also depends on an effective national technological system. Support for catch-up is generally required, as firms must overcome market and technological barriers to entry and acquire technological capabilities to become competitive. This support must be selective, due to limited government resources for financial and infrastructural support, foreign currency, human, physical and financial capital, and entrepreneurial capabilities. This thesis investigates how these critical but elusive technological capabilities may be built in PV manufacture. The concepts of technological capabilities, technological systems and the characteristics of technological learning are therefore further explored in chapter 5, forming the basis for an analytical framework that facilitates assessment of the factors that influence capability building in latecomer PV manufacturers.
2.1.3. A Special Role for PV Manufacture PV cell manufacture; requiring high levels of technological capabilities, including a degree of innovative capability, is likely to be too challenging for developing countries with a
12 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
low level of technological and industrial infrastructure. For newly industrialising countries, however, possessing a more sophisticated industrial and technological base, cell manufacture may be an appropriate strategic niche industry for developing countries to support. PV cell manufacture has historically occurred mainly in Japan, Germany and the USA. Asian newly industrialising countries, dominated by China, are now emerging as significant manufacturers. Manufacturing offers opportunities for firms in industrialising countries to capitalise on their advantage in low cost labour and their potential to produce the inputs to production at lower cost than advanced economies. Renewable energy technologies such as photovoltaics are generally of interest because of their potential to address global issues of climate change and energy security. In addition, photovoltaic cell manufacture may be a particularly appropriate industry for industrialising countries to select for a number of reasons. Firstly, photovoltaics can contribute to the fulfilment of goals of energy security and rural electrification in most developing countries, while conserving foreign currency. Local manufacturers can benefit by accessing local markets for PV rural electrification. Secondly, photovoltaics is technologically close to electronics, which, at its most advanced, is a high return, dynamic and growing manufacturing industry. Thirdly, the technology is at a relatively mature stage of the product cycle, where standard production technology and capital equipment can be purchased, giving new entrants an opportunity to learn the technology. Fourthly, although the global market is relatively immature, it is growing and has a large and increasingly certain potential. Export markets are an important driver for industrial development because they promote technological learning through interacting in demanding and rapidly changing global factor and product markets. Export industries also provide opportunities to earn valuable foreign currency. Having established the importance of high technology industries in industrial development, and a particular role for photovoltaics manufacture, the following sections describe the international PV industry and the nature of competition in the industry in order to identify the advantages and barriers for PV manufacturers in developing countries.
13 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
2.2. Markets and Firms in the International PV Industry
Markets are the main source of incentives for PV cell and module manufacturers to improve technologically and make investments, and will decide the success or failure of latecomers. The location and type of firms involved in the PV technological system will determine the knowledge, resources and linkages they bring with them and hence influence their ability to compete and their strategic behaviour. The following sections describe the emergence of niche markets for photovoltaics up to the period of recent subsidy-based urban market growth and the location, concentration and type of firms in the PV industry.
2.2.1. The Evolution of Markets for PV The cost of electricity from photovoltaics is presently too high to compete on an equal basis with conventional generating technologies in central grid-connected markets. As the cost of photovoltaics has reduced, however, it has become viable in new niche markets, including high-value satellite applications, remote rural solar home systems, remote industrial applications, particularly telecommunications, and increasingly in diesel or other mini grids. Several authors have focused on the importance of niche markets as stepping stones towards competitiveness with conventional electricity generation technologies. Balaguer & Marinova (2006) call these markets ‘nursing markets’, in which customers are generally willing to pay more, buying the technology time to improve its cost and performance, as well as providing incentives for firms and suppliers to enter the market, bringing new knowledge and resources. The market for solar cells is described by Jacobsson et al. (2002) as evolving through five periods: ‘the space age’, when photovoltaics first found a commercial niche in the high- value satellite application; ‘oil crisis and solar vision’, a period of interest in alternative energy as a result of the oil crisis in the 1970s; ‘commercial off-grid markets’, a period dominated by markets for off-grid applications, where photovoltaics was competitive on cost; followed by ‘roof-top programmes’, a policy-driven growth of the grid-connected market that has occurred during the last decade. These are illustrated in Figure 2-1 (note the log scale).
14 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
Figure 2-1: Five Phases of Solar Cell Diffusion
Figure has been removed due to copyright restrictions.
Source: (Jacobsson et al., 2002, p 10)
In 2005, photovoltaic module production reached 1.759 GWp, a 9 billion euro/year business (Jäger-Waldau, 2006). Business analysts suggest that the market will be between
6GWp (Mints, 2006) and 10GWp (Rogol, 2006) by 2010, a 40 billion euro/year business. The grid-connected market grew to over 85% of the total market in IEA countries by 2005 (Figure 2-2), and is predicted to continue to be the most important market segment (Hoffmann, 2004).
Figure 2-2: Cumulative installed grid-connected and off-grid PV power in the IEA reporting countries
Figure has been removed due to copyright restrictions.
Source: (IEA, 2006)
Central to the recent period of market expansion in the PV industry has been market formation programmes, particularly in Japan and Germany. There is a strong incentive for the
15 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
industry to continue to reduce costs in order to retain public support for the technology as a viable option for cost-effective electricity production in the future. The CEO of Q-Cells, the second largest manufacturer of PV cells, believes that the industry has a 4-5 year window to significantly reduce costs and become competitive in some markets in order to maintain momentum and support (Milner, 2006). The industry is currently experiencing a shortage of silicon feedstock. Continuing growth rates will therefore also depend on increasing production of and reduced usage of silicon material. It is expected, however, that the shortage will be resolved within a few years as new production capacity comes on line. With fossil fuel prices expected to rise, many countries increasingly dependent on imports of energy, more and more countries committing to renewable energy targets, and to measures including feed-in tariffs for PV, the long-term demand for PV is expected to remain high (Jäger-Waldau, 2006). However, although markets for PV are growing, and there are niche markets achieved or within sight, the market is not yet self-sustaining and requires ongoing policy interventions.
2.2.2. The Location and Concentration of PV Firms Photon Magazine’s annual worldwide production survey data (Hirshman et al., 2007; Schmela, 2005c, 2006) has been used to reveal the location of PV cell and module manufacturing worldwide, as displayed in Figure 2-3 and Figure 2-4. The majority of cell production in 2006 came from Japan (36%), Germany (20%) and China (15%). Developing countries, including those from Africa, the Middle East, Eastern Europe and Russia, Central and South America, China, India and the rest of Asia contributed 27.5% in 2006 and continue to produce an increasing fraction of worldwide cell production.
Figure 2-3: Global 2004, 2005, 2006 and Planned 2007 Cell Manufacture by Region
1200
1000
800
600 MW
400
200
0 Africa Middle Germany Rest of Russia Eastern North Central & China India Japan Rest of Australia East Europe Europe America South Asia America
2004 Cell Production 2005 Cell Production 2006 Cell Production Planned 2007 Cell Production
Source: data from (Hirshman et al., 2007; Schmela, 2005c, 2006)
16 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
The majority of module assembly worldwide is also carried out in industrialised nations, primarily in Japan (2006, 30%) and western and central Europe (2006, 30%). Developing countries produce a larger share of modules than cells, assembling 32% of modules in 2006 and expected to produce 42% in 2007. Most of this increase will come from China, which has rapidly increased module production to 22% of world supply in 2006, projected to increase to 32% in 2007.
Figure 2-4: Global 2004-2006 Module Manufacture and Planned 2007 Module Manufacture by Region
1400
1200
1000
800 MW 600
400
200
0 Africa Middle Germany Rest of Russia Eastern North Central & China India Japan Rest of Australia East Europe Europe America South Asia America
Planned 2007 Module Production 2006 Module Production 2005 Module Production 2004 Module Production
Source: data from (Hirshman et al., 2007; Schmela, 2005c, 2006)
The PV industry is relatively concentrated, with the top 10 manufacturers producing 74% of global output of cells in 2005 (Figure 2-5). Mergers and acquisitions have and are continuing to consolidate the PV industry structure (Jäger-Waldau, 2006). Many of the largest firms in the PV industry have been around for decades, and have accumulated significant experience in production and R&D. There is also a steady stream of new entrants to the industry, some of which have been able to rapidly expand, becoming significant players. An increase in industry concentration via mergers and acquisitions may be expected to lead to a decrease in competition, an increase in market power, and higher profit margins for remaining firms, but the threat of new entrants is likely to keep the competitive pressure on.
17 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
Figure 2-5: Share of 2006 Global Cell Manufacturing by Manufacturer
Sharp 18%
Others 35%
Q-Cells 11%
Kyocera 7% Deutsche Cell 3% Suntech Power RWE Schott Solar 6% 3% Motech Sanyo BP Solar 4% 6% 3% Mitsubishi Electric 4%
Source: data from (Hirshman et al., 2007)
2.2.3. Types of Firms in the PV Industry There are a variety of multinational firms that have extended to photovoltaics, particularly in Japan, Europe and the US, where photovoltaics manufacture has been dominated by firms that had electronic equipment, semiconductors, chemical materials or energy as their core business (Balaguer & Marinova, 2006; Linden et al., 1977). Most of the top 10 manufacturers are divisions of large multinationals, including Sharp, BP Solar, Kyocera, Shell, RWE Schott, Sanyo and Mitsubishi, and have benefited from strategic mergers and acquisitions. Some of these companies, in particular, semiconductor and microelectronics firms, have entered the industry in order to take advantage of related competencies in some part of the PV value chain, such as know-how in manufacturing, microelectronics and materials. Others, such as oil and energy companies, are likely to have diversified into PV in order to preserve long term market share, since new energy technologies such as photovoltaics are potential substitutes for fossil fuels. Multinational firms have high-profile brand names established through their core business. Market access may be influenced by such brand advantage, creating a barrier to the entry of new firms. The only pure PV firms in the top 10 are Suntech (China), Motech (Taiwan) and Q-cells (Germany). Older PV startups Isofoton and Photowatt were in the top 10 firms in 2001. There are also a large number of smaller photovoltaics start-ups, such as Evergreen Solar, First Solar, Sunpower, Global Solar Energy, Daystar, Nanosolar in the US; and Antec Solar Energy, CSG Solar, ErSol Solar Energy, Photovoltech, Solar World, Sunways, among many others in Europe. Most of these firms were formed in order to commercialise research from universities or research institutes, while a few vertically expanded from PV systems companies.
18 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
2.3. Competitive Strategies and Barriers in the International Photovoltaics Industry
PV manufacturers, until recently, have indicated that they have been absorbing losses (Schaeffer et al., 2004; Schmela, 2006). PV manufacturing operations (especially those that are small parts of multinational companies) have been seen as a strategic future investment, and producers have forfeited current profits for market share and position in the future. According to Schaeffer et al. (2004), these manufacturers had aimed to start making profits by 2004-2005. The recent silicon shortage and market expansion has created demand conditions that have enabled manufacturers to sell all of their production without significant price reductions, and therefore helped them to make the transition to profits on target. This section describes the role of a number of competitive strategies in the industry and the impact on latecomers of these strategies: expansion of market share, cost reduction, product differentiation and brand equity, and technology leadership. Barriers to entry and advantages for latecomer firms in the PV industry are then summarised.
2.3.1. Expansion of Market Share Expansion of operations can increase the market share of manufacturers and give them access to cost reductions via scale economies. Since cell manufacture is highly capital-intensive, most producers have large centralized plants in one or a few locations, and export to access international markets. Existing plants are commonly between 10-50MWp/year capacity, and new facilities being built are often greater than 50MWp/year (Eberhardt, 2005), with 11 firms planning to expand plants to over 200MWp in 2007 (Jäger-Waldau, 2006). Some predict factories of 1GWp capacity by 2010 (Lüdemann, 2005; Podewils, 2007). A capital investment of
US$0.65-1.3M per MWp capacity of crystalline silicon (c-Si) production is required (Lüdemann, 2005; Solarbuzz, 2007) and US$2M per MW or more for thin film capacity (Solarbuzz, 2007). As firms expand production rapidly, PV manufacturing plants are more than doubling in size each year (Nemet, 2006), allowing firms to realise economies of scale, but also increasing the upfront investment required to enter the industry on a competitive footing. Production lines for wafer-based crystalline technologies can easily be purchased and be up and running in a short period of time (Jäger-Waldau, 2006). Solarbuzz (2007) suggests that they can be operational within 18 months to two years of project approval, and running at
19 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
full capacity after a further year. Cell plants often have parallel production lines, allowing for sequential expansion. However, modular expansion may exclude existing producers from some of the benefits of scale economies that can be obtained when a new large-scale plant is installed. In response to recent market growth, many firms have invested heavily in expansion, leaving them vulnerable to changes in the level of political support for market formation programs. Since future market growth is largely reliant on continuing government support, expansion of capacity may be a high risk investment proposition for manufacturers. Suppliers to photovoltaics firms may also be cautious to invest. Indeed, silicon feedstock suppliers have been reticent to ramp up capacity without long term purchase contracts.
2.3.2. Cost Reduction Since there is little physical differentiation between PV modules from different sources, the primary basis for competition is cost (Haase, 2005). The cost of photovoltaics is usually expressed as cost per watt peak (Wp). The two main routes to cost reductions in photovoltaics 2 are via increasing the generation efficiency (Wp/m ) of modules (while maintaining production costs), or reducing the cost of production (while maintaining efficiency). With accumulated experience and technological progress, prices for PV have steadily decreased and performance has increased. Cost leadership in photovoltaics may be obtained either through the development of new types of solar cells that are more efficient or cheaper to make; or through incremental improvements in efficiency or production cost using conventional technology. Follower firms may lower the cost of the product or production by learning from the experience of technological leaders and thereby avoiding R&D costs (Porter, 1985). The sustainability of the competitive advantage of cost leaders depends on competitors being unable to duplicate the technology or leader firms innovating faster than competitors can catch up. The PV industry is not yet a mature industry, and therefore experiences high levels of technological change. Even for firms using mature technology, building capabilities and sustaining the improvements required to keep up with progress in the industry will be challenging.
2.3.3. Quality, Product Differentiation and Brand Equity Apart from cost, different brands of PV modules are in most markets comparatively undifferentiated. Areas of differentiation include efficiency, physical durability and aesthetic qualities of the product.
For a given size (Wp) PV installation, efficiency determines the surface area occupied by the modules, which is a strong determinant of the cost of mounting structures and BOS for PV systems (Jaus et al., 2005). The costs of module encapsulation are also reduced if cell
20 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
efficiencies improve (Jäger-Waldau, 2006). Efficiency often becomes more important in the built environment, especially in countries such as Japan, where the amount of space available to accommodate PV is limited and can be a boundary to total generating capacity. While the cost-effectiveness of a PV module in terms of electricity production is primarily reliant on its $/Wp cost, leadership in technology and high quality can also create a good brand reputation. A good brand image can carry benefits such as access to better employees, market access, access to facilities (sites), inputs or other scare resources, special status with the government, and early profits and expansion. Quality implies a longer life for the product, often associated with longer warranty periods. The lifecycle cost of electricity production is reduced if the PV module continues to produce electricity for longer. Photovoltaics markets are largest in Japan and Germany, where political support for the industry has been strong. The German and Japanese markets are sophisticated and demand high quality, good performance and certification of products. Producers in other countries are likely to export a good proportion of their product to these markets and other fast growing markets in Europe, where the stiff competition and demanding consumers will require continuous improvement. For example, demanding export markets have encouraged Chinese manufacturers to improve their quality and certify their products and production lines, increasing the confidence of foreign buyers (Hug & Schachinger, 2006). Building-integrated PV modules are an emerging market segment in these countries, where product differentiation is more important, since aesthetic considerations and good integration with building materials and construction methods are significant in this high value market. 2 Quality and efficiency (Wp/m ) are likely to be secondary considerations after cost
($/Wp) in the choice of PV module for rural applications in developing countries (Li, 2004). Consumers are usually unable to discern the difference in quality between one module and another, and must rely on certifications or brand as an assurance of quality. However, consumers in these markets are unlikely to understand the meaning of certifications (Real, 2006), and unmonitored markets are vulnerable to fraudulent use of the certifications (Wassen, 2006). In rural electrification markets in developing countries, where PV systems are not affordable for most of the people without access to grid electricity, the cheapest modules are therefore likely to find favour with consumers. Demand for low prices may make it difficult for less technologically capable firms to secure market share from the cost leaders, but some may be able to produce lower quality modules cheaply.
2.3.4. Technology Leadership The widely anticipated emergence of thin-film PV technology is likely to result in what Andersson & Jacobsson (2000) call a minor discontinuity in technological progress. Such technological discontinuities often change the balance of competitive advantage and make old
21 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
knowledge, resources and linkages obsolete. Firms must therefore monitor technology changes and attempt to stay at the forefront of these changes. Thin film technologies use radically different production techniques; so much of the production knowledge for wafer-based cells will not be useful for thin-film manufacture. The materials and processes for thin-film devices are also very different. If thin-film technology becomes dominant, current cell-handling and encapsulation technology will become obsolete, since thin-film technology is not based on the production of individual cells. For the time being, specialised equipment for thin film deposition and processing must often be custom built, appropriate materials may not be available off-the-shelf, and manufacturing issues with the new technology must be resolved. Thin film plants, even those employing well-developed technology, currently take slightly longer than wafer-based technologies to become operational and several years to achieve full production capacity (Solarbuzz, 2007). Production may be constrained in the short term by lower demand for the new technology, because of market uncertainties about quality and economies of scale are therefore unlikely to be achieved. Breakthroughs in device designs could have dramatic impacts on the PV technology trajectory2 (Margolis, 2003) and learning curve, but the potential for breakthroughs in the various PV technologies is difficult to quantify, and new technologies require investment in new materials and equipment and suffer from scale and supplier disadvantages. Uncertainty and the investments required are disincentives to commercialise the alternative technologies. Nevertheless, there are a number of firms investing in a variety of alternative technologies, including some of the large firms already involved in production with the dominant technology, such as Shell Solar, the first company to commercialise CIS/CIGS solar cells (Baumann et al., 2004); Sharp, who announced large scale production of thin films beginning in September 2005 (Jäger-Waldau, 2006); and BP Solar (although they have recently abandoned their CdTe manufacture). A significant and increasing number of start-ups have also been formed to commercialise new technology, usually developed in research institutes. There are less potential benefits in the improvement of mature technology, but the return is more certain (Porter, 1985). According to Schott Solar‘s CEO Winifred Hoffmann, many of the current improvements in standard solar cells arise from patents registered during the 1970s and 1980s, which were often technologies developed for space applications, for which the patents have expired (Siemer, 2005). There may therefore be scope for firms to continue to use mature technology for a number of years to come, but an important technological breakthrough could come at any time. The European Commission roadmap for PV R&D predicts that crystalline silicon will continue to dominate the market for a decade beyond 2004, thin films will start to penetrate significantly after 2010 and novel technologies will not significantly penetrate markets until 2020 (de Moor et al., 2004).
2 A technological trajectory is the direction of technological progress. This concept is further explored in chapter 3. 22 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
2.3.5. Vertical Integration Vertical integration into key parts of the value chain that embody technology or give clues to technology may be an important source of competitive advantage (Porter, 1985). Shum & Watanabe (2006) suggest that major vertically integrated firms can facilitate learning more effectively and cost-efficiently than those within a non-integrated value chain. Vertical integration may also enable access to scarce materials or reduce the cost of procuring materials and equipment. Balaguer & Marinova (2006) point out that upstream vertical integration into silicon feedstock and wafer manufacture can insulate firms from those parts of the value chain that currently have high value added and have the potential to impact the cost structure of the product. Silicon supply shortages have therefore increased the tendency towards vertical integration (IEA, 2006). A number of the largest PV cell manufacturers are vertically integrated into upstream parts of the value chain. Kyocera is engaged in casting silicon ingots, slicing wafers, producing cells and assembling modules (Kyocera, 2004), while German manufacturers SolarWorld and RWE Schott Solar also manufacture wafers, cells and modules. SolarWorld has created a joint venture to develop a solar grade silicon production process. Most large cell manufacturers are vertically integrated downstream (assemble cells into modules). Of the top 10 cell manufacturers, only Q-cells, Motech and Deutsche Cell did not make modules in 2005. The remainder converted most of their cell production to modules. Of the 10 largest cell and module manufacturers, Suntech in China and Motech in Taiwan are the only ones based primarily in a newly industrialising country.
Table 2-1: Large Cell Manufacturers’ Cell and Module Production 2006 Worldwide Countries with Plants Worldwide Cell Module Production Production 2006 by 2006 by Manufacturer Manufacturer Company (MW) (MW) 1 Sharp 435 362 UK, Japan 2 Q-Cells 253 0 Germany 3 Kyocera 180 102 Czech Rep, China, Mexico, Japan 4 Suntech Power 160 120 China 5 Sanyo 155 57 Hungary, Mexico, Japan 6 Mitsubishi Electric 111 111 Japan 7 Motech 102 0 Taiwan 8 BP Solar 86 92 Spain, USA, China, India, Australia 9 RWE Schott Solar 96 55 Germany, Czech Republic, USA 10 Deutsche Cell 70 0 Germany Others 889 1434 2536 2333 Source: (Schmela, 2006)
Many Japanese manufacturers, such as Sanyo and Kyocera are highly vertically integrated downstream, including cell, module and BOS manufacture, and sometimes
23 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
installation and maintenance (Foster, 2005). Some have purchased housing or construction companies or collaborated with existing firms to access the grid-connected solar home system market. Shum & Watanabe (2006) believe that this type of vertical integration can facilitate learning in areas such as module durability and quality.
24 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
2.4. PV Cell and Module Manufacture in Developing Countries
Having described the basis for competition in the international PV manufacturing industry, this section summarises the status of manufacture in developing countries and important competitive advantages and barriers for latecomers.
2.4.1. The Status of PV Module Manufacture in Developing Countries 16% of the manufacture of modules in 2004 listed in the Photon Magazine market surveys (Hirshman et al., 2007; Schmela, 2005c, 2006) took place in developing countries, with the market share increasing to 26% in 2005, and 40% in 2006. China has more than doubled its module production between 2004 and 2006, and plans indicate that they will again in 2007. The largest of more than 25 Chinese module producers are Suntech Power, Shenzhen Jiawei, Ningbo Solar Cell Factory and Shanghai Solar. In India, there are at least 13 module producers, but most of the production comes from TATA BP Solar. There is also significant expansion in production in Eastern European countries, mainly through Kyocera and Schott Solar in the Czech Republic and Sanyo in Hungary, but there are also smaller operations such as Solar Cells in Croatia and Energy Solutions in Bulgaria. Sanyo is also operating a new plant in Mexico. There are also a number of smaller producers each in Thailand, South Korea and Taiwan. In Africa, most of the production comes from Total Energie / Tenesol in South Africa, but there are small manufacturers in some less developed African countries, such as Namibia and Uganda. In Namibia, a $6.4m solar module factory, Nopasika Electronic, was opened in March 2002 (Hirshman, 2002). Racell Uganda Ltd., a Danish-Ugandan joint venture based in Kampala, has started manufacturing modules, while Solar Energy Uganda is constructing a 1 MW plant, scheduled for completion in 2007, with an investment of US$3.4 million. The factory will produce small multicrystalline modules for the regional market with power sizes of 10, 20, 32, and 64 W (Brand, 2005). Racell Uganda, and Liselo also operate in South Africa.
25 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
Figure 2-6: PV Module Production in Developing Countries 2004 - 2007
1400
1200
1000
800 MW 600
400
200
0 Africa Middle Russia Eastern Central & China India Rest of East Europe South Asia America
Planned 2007 Module Production 2006 Module Production Planned 2006 Module Production 2005 Module Production 2004 Module Production
Sources: (Hirshman et al., 2007; Schmela, 2005c, 2006)
2.4.2. The Status of PV Cell Manufacture in Developing Countries Almost 16% of the worldwide PV cell production in 2005 was in developing countries, increasing to 27.5% in 2006. China and other Asian nations make up the bulk of this production. The share of global manufacturing in China was projected to increase from 15% in 2006 to 23% in 2007, and the rest of Asia’s share to rise from 10% to 13.5% over the same period. Eastern Europe and the Middle East are also beginning to break into cell production. Many of the countries with cell manufacturing facilities already have capacity in related high-tech industries, such as electronics manufacture. Of the largest vertically integrated cell and module manufacturers, Suntech in China is the only one based primarily in a newly industrialising country. Of the firms in industrialising countries, only Sunpower (Phillipines FDI, high efficiency), Orion (Israel, Dye cells) and Ulicia (China, recycled solar cells) are producing cells using a radical design.
26 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
Figure 2-7: PV Cell Production in Developing and Industrialising Countries 2004 - 2007
1200
1000
800
600 MW
400
200
0 Africa Middle Russia Eastern Central & China India Rest of East Europe South Asia America
Planned 2007 Cell Production 2006 Cell Production 2005 Cell Production 2004 Cell Production
Sources: (Hirshman et al., 2007; Schmela, 2005c, 2006)
2.4.3. Sources of Technology for Cell and Module Manufacture in Developing Countries PV manufacturers in developing countries have acquired technology from a range of sources, ranging from equity-based technology transfer, such as foreign investment and joint ventures, to independent technology acquisitions such as technology purchases, and for more advanced manufacturers, research collaborations.
Equity-based Technology Transfer There are a few examples of foreign-owned cell manufacturing start-ups in industrialising countries, including Sunpower in the Philippines and Sinonar in Taiwan. There are also a number of joint ventures in cell manufacture, the oldest of which is BP Solar’s joint venture with Tata Power and Maharishi in India. Jing Ao Solar and Nanjing PV Tech Co. Ltd (CEEG) in China are recent foreign-domestic joint ventures. Harbin-Chronar and Yu Kang Solar were Chinese-US joint ventures that failed in 2003 and 1997 respectively. In many cases the foreign partner provides technology, plant, commissioning, training, while the local joint venture partner provides the facility, installation, labour, utilities, services, local marketing and sales for the factory. Almost all the foreign direct investments (FDI) made by multinational PV companies in developing countries are in module manufacturing facilities, which give the large cell manufacturers access to regional markets with reduced transportation costs (Jordan, 1995), as 27 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
well as cheaper labour in many cases. RWE Schott Solar has a plant in the Czech Republic and Sanyo assembles modules in Mexico and Hungary. BP Solar has module lines in a number of Asian countries. Kyocera has consciously pursued a strategy of globalisation, establishing local module production companies in Mexico, to supply the US market; the Czech Republic to supply the European market; and in China to access the growing Chinese market; while the Japanese plant will continue to supply cells to all the module assembly plants, and make modules for the Japanese market (Kyocera, 2004). The technology for module assembly is much simpler than that for cell manufacture, and in the absence of investments in transferring the cell technology, there is little chance for module manufacturers to make the transition to the more sophisticated cell manufacturing technology. Since firms require innovative capabilities in order to keep up with the incremental improvements in the PV industry, the developing country manufacturer in a joint venture will need to build these capabilities or rely on continuous inputs of technology from the foreign partner. Despite low cost manufacturing advantages, the small number of successful PV firms set up under FDI or joint ventures in developing countries reflects the difficulty of building this level of capabilities and accessing the resources required for manufacturing PV in developing countries.
Subcontracting Subcontracted module assembly, usually with the brand of the cell manufacturer, has become common, particularly in China. For instance, Sharp and SolarWorld have modules produced in China (Schmela, 2005b) and REW Schott Solar and Sunpower have modules produced in India (Hirshman, 2006c) through this type of arrangement. The benefits of equity- based and subcontracting arrangements include access to markets, technology and scale economies, and may provide an opportunity for firms to ‘learn by doing’. Equity based arrangements, however, generally involve less capability building and more dependence on the foreign firm than does subcontracting.
Purchase of Technology Mature process technology for wafer-based silicon solar cells is freely available without licenses, since any patents for the common processes have expired (Siemer, 2005). Startups in developing countries are likely to have competitive advantages in manufacturing costs, rather than access to innovative technology, and may therefore begin by using mature technology or purchasing technology. The technology is often acquired through a turnkey provider who generally provides technology, equipment, installation, commissioning, training and warranty to produce modules; or through a contractor who provides advice on procurement, commissioning and operation of the plant. The focus of these firms in the early stages of production will be to fine-tune the production line, reduce costs and eventually improve processes.
28 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
More capable firms may license technology through a patent agreement, whereby they are legally permitted to employ innovative patented technology for a limited period of time or in a limited region (Lewis & Wiser, 2005). Licensing agreements often do not include transfer of the associated tacit knowledge and complementary technologies required to use the patented technology, and the firm must therefore obtain these capabilities in some other way. Innovative device designs in the PV industry normally come from universities, research organisations or technology institutes. Laboratory solar cell efficiencies, for instance, have largely been achieved through R&D in universities and research institutes, rather than in firms, as illustrated in Figure 2-8. Much of the dramatic cost reduction in PV cell manufacture between the 1950s and the 1990s was a result of public R&D investment. With low levels of production, learning by doing probably played a fairly minor role during this period.
Figure 2-8: Record Solar Cell Efficiencies and Organisations for Laboratory Devices
40.0
Spectrolab 35.0 Spectrolab NREL/Spectrolab
NREL 30.0 NREL Japan Energy
25.0 UNSW UNSW Spire UNSW UNSW UNSW Spire Stanford Georgia UNSW 20.0 Tech
ARCO Varian Georgia Sharp NREL NREL NREL Tech Efficiency (%) Westinghouse N. Carolina Uni NREL Uni S Solarex NREL 15.0 Florida Solarex Astropower Boeing Euro CIS ARCO United Solar Kodak Boeing AMTEK Uni Lausanne Photon 10.0 Masushita Kodak Astropower Energy United Solar Monosolar Boeing
Boeing Solarex Uni Lausanne RCA Siemens 5.0 Univ. Maine RCA Princeton RCA Groenigen Uni Linz RCA Uni Linz RCA Kodak UCSB RCA RCA Cambridge Berkeley 0.0 1970 1975 1980 1985 1990 1995 2000 2005 2010
Cu(In,Ga)Se2 Single Crystal Multicrystalline CdTe Amorphous Si (stablised) Dye Cells Organic Cells Multijunction Concentrators Thin Si
Source: Data from (Baumann et al., 2004; Surek, 2003)
Many large PV companies have licensed innovative device designs or process technology developed at universities or research institutions. For example, BP Solar licensed the laser-grooved, buried contact (LGBC) cells used to make their Saturn modules from the University of New South Wales. The access that firms have to the science and technology available in public institutions is likely to be important, particularly since firms are unlikely to licence technology they have developed to competitors. New technologies are often developed using processes that are impractical in a commercial environment and need to be adapted. Firms who implement new technology,
29 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
whether it is acquired internally or externally, will therefore usually have to make some of their own investments in R&D and pilot lines in order to get the technology into commercial production. Production experience with related technologies will help them to make the most of new technology.
Research Collaborations and International R&D R&D collaborations with universities and public research institutes have been an important source of learning for PV manufacturers in Europe, Japan and the US (Balaguer & Marinova, 2006; Jäger-Waldau, 2003). Technical collaborations allow firms to put new technology into production with reduced R&D investments and risks, and reduced dependence on technology licensing. Research collaborations are generally undertaken by more established manufacturers. These collaborations can help firms build innovative capabilities and make the transition to independent device R&D. Some of the more advanced PV cell manufacturers in developing countries involve collaborations, including: Shanghai Topsola Green Energy, set up to commercialise technology from Shanghai Jiao Tong university; Suntech Power, the subject of the case study in chapter 5; Central Electronics Ltd, who has collaborated with IMEC, Belgium; Photon Semiconductor, which was founded to commercialise technology from a Korean university; and DelSolar and E-Ton Dynamics in Taiwan, who have collaborations with Taiwanese research institutes and firms.
2.4.4. Advantages and Barriers for Firms in Developing Countries The extent of manufacture of PV cells and modules in developing countries is increasing, and much of the production, particularly from China and some other Asian countries is being now being exported (Wang, 2006). Firms in developing countries may benefit from competitive advantages such as low labour costs and the ability to cheaply manufacture inputs as well as the ability to install the latest technology at a large scale. Foreign firms may have difficulty working in developing countries and accessing the same advantages, since administrative procedures in these countries are often complex and obstructive, and foreign firms may not have access to local markets and industry actors. Notwithstanding the low cost of labour and in some cases, materials and equipment, manufacturers in developing countries have had difficulty remaining cost competitive until the past few years. In China, despite efforts to develop a PV manufacturing industry since the
30 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
1970s, low efficiencies (Dai et al., 1999), poor capacity utilisation (Zhao, 2001) and failure to invest in the latest capital equipment (Yang et al., 2003) prevented manufacturers from achieving cost reductions on par with international industry leaders (Yang et al., 2003; Zhao et al., 2006). Similarly, despite indigenous technology development and a long history of PV manufacture, manufacturers in India have failed to keep up with international improvements in cell efficiency, manufacturing line yield, capacity utilisation and therefore price (Kathuria, 2002), and have not scaled up at the same rate as international manufacturers (Hirshman, 2006a, b). Clearly, the ability of manufacturers in developing countries to compete on the basis of price has been hampered by their inability to access, use, and build upon the latest technology. The quality of modules produced in developing countries has also been insufficient for demanding European markets (Hug & Schachinger, 2006; Schmela, 2005a). Manufacturers in developing countries have therefore been unable to access large export markets and therefore opportunities for economies of scale, although some manufacturers have recently been able to improve the quality of their products in order to access these markets. The presence of multinationals with recognisable brand names associated with quality may be a barrier for manufacturers in developing countries, even in domestic markets. The high rate of technological change within the industry and the expectation of new technologies that may be cost effective within decades imply that manufacturers in developing countries must not only catch up with the industry leaders in terms of price and quality, but also be prepared to improve continually to stay abreast of these changes. The ability of developing country manufacturers to use PV technology successfully is also impacted by their ability to operate in factor markets. New firms may have difficulty attracting the investment capital required for the large scale plants that can achieve cost- effective manufacture, particularly in the context of uncertain subsidised markets. They are also likely to have difficulty attracting the highly skilled technical staff they require, particularly in an environment of scarcity of trained personnel. The silicon shortage and the high cost of materials increases the importance of both reductions in silicon use and the ability to negotiate over supplies. New firms may be unable to operate at full capacity, or have to resort to the spot market, since most silicon suppliers have committed their product to long term contracts. The tendency of large firms to vertically integrate into silicon production may increase difficulty for other firms to negotiate supply agreements.
31 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
2.5. Conclusion
This chapter has identified roles for PV manufacture in the industrialisation strategies of more advanced developing countries, described the PV manufacturing industry, both internationally and in developing countries, and identified barriers to successful manufacture in developing countries. The largest PV markets internationally are for grid-connected systems in industrialised countries, which are strongly dependent on government market support. The majority of PV cell and module manufacturers are located in industrialised countries, but developing countries, led by China, are increasing their share of production. Competition between manufacturers is primarily based on price. Expansion of market share gives manufacturers opportunities to reduce price via economies of scale, while good quality is a requirement for export markets, and brand equity and technology leadership may also provide better access to markets and to skilled personnel. Vertical integration has enabled some manufacturers to reduce their vulnerability to changes in supply markets. The manufacture of photovoltaic cells offers opportunities for firms in developing countries to capitalise on their advantage in low cost labour and their potential to produce the inputs to production at lower cost than advanced economies. New enterprises may also benefit by installing the latest technology at a large scale. The factors influencing the successful use of PV manufacturing technology in developing countries are therefore of interest. To date, however, PV cell manufacturers in developing countries have either had a high dependence on foreign owners of joint venture partners, or have had little success in the large export markets. The most significant barrier to manufacturers in developing countries successfully using PV technology is the inability to achieve sufficiently low costs and high quality and therefore access export markets and scale up production. It is likely that accessing technology, including that embodied in skilled personnel, and accessing investment capital have been problematic. Latecomers are also at a competitive disadvantage due to low levels of brand recognition, while small scale and lack of leverage are likely to restrict their ability to interact effectively with supply markets. This chapter has established that while there is a role for PV cell and module manufacturers in developing countries, they have had difficulty developing sufficient capabilities to acquire, use and build upon PV technology in order to overcome market and technological barriers to entry. The following chapter will review the literature on technological learning and technological systems in the PV industry, in order to discover what is known about the factors influencing the successful use of PV manufacturing technology in developing countries and identify opportunities for further research.
32 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
References
Andersson, B.A. and Jacobsson, S. (2000), Monitoring and assessing technology choice: the case of solar cells, Energy Policy, 28 (14), p 1037. Balaguer, A. and Marinova, D. (2006), Sectoral Transformation in the Photovoltaics Industry in Australia, Germany and Japan: Contrasting the Co-evolution of Actors, Knowledge, Institutions and Markets, Prometheus, 24 (3). Baumann, A., Bharghava, Y., Liu, Z.X., Nemet, G.F. and Wilcox, J. (2004), Photovoltaics Technology Review, Presentation in Course MSE / ER C226: Photovoltaic Materials: Modern Technologies in the Context of a Growing Renewable Energy Market, UCLA, Berkeley, California, USA. Brand, B. (2005), Solar Energy Uganda announces plans for module factory in Entebbe, PHOTON International, September 2005. Dahlman, C. and Westphal, L. (1982), Technological Effort in Industrial Development: An interpretive survey of recent research, in Stewart, F. & James, J. (eds), "The Economics of new technology in developing countries", F. Pinter; Westview Press, London; Boulder, Colorado., pp vi, 282 p. Dai, Y., Shi, Z. and Xi, X. (1999), Technological Innovation of the Chinese Photovoltaic Industry, Prepared for the Center for the Integrated Study of the Human Dimensions of Global Change, Carnegie Mellon University, Beijing, China. de Moor, H., Hagemann, I., Herrero, J., Malbranche, P., Siebentritt, S., Schedel-Niedrig, T., Riedle, T., Poortmans, J., Nijs, J., Szlufcik, J., Jäger-Waldau, A., Karg, F., Teksolar, E.Ö., Pietruszko, S., Haywood, S., Topic, M., Schropp, R., Rath, J., Dimmler, B., Powalla, M. and Springer, J. (2004), PVNET European Roadmap for PV R&D, R&D for PV Products Generating Clean Electricity Contract Reference: ENK6-CT2001- 20400, Jäger-Waldau, A. (ed), European Communities, Luxembourg. Dutrénit, G. (2004), Building Technological Capabilities in Latecomer Firms: A Review Essay, Science, Technology & Society, 9 (2). Eberhardt, K. (2005), PV Production Facilities adapted to Technology Requirements - a Prerequisite for cost effective Mass Production, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. Foster, R. (2005), Japan Photovoltaics Market Overview, Sandia National Laboratories, U.S. Department of Energy. Grieve, R.H. (2004), Appropriate Technology in a Globalizing World, International Journal of Technology Management and Sustainable Development, 3 (3), pp 173-187. Haase, J. (2005), High Volume Solar Cell Production Lines - Concepts and Cost Aspects, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. Hirshman, W.P. (2002), Namibian company starts module production, PHOTON International, May 2002. Hirshman, W.P. (2006a), Bharat Heavy Electricals Ltd., Photon International, May 2006, pp 94-96. Hirshman, W.P. (2006b), Tata BP Solar, Photon International, May 2006, pp 86-88. Hirshman, W.P. (2006c), Titan Energy Systems, Photon International, May 2006, pp 98-100. Hirshman, W.P., Hering, G. and Schmela, M. (2007), Gigawatts - the measure of things to come: Market survey on global solar cell and module production in 2006, Photon International (March 2007), pp 136-166. Hobday, M. (1995), East Asian latecomer firms: Learning the technology of electronics, World Development, 23 (7), p 1171. Hoffmann, W. (2004), PV Solar Electricity Industry: Market growth and perspectives, 14th International PVSEC, Bangkok, Thailand, 26-30 January, 2004. Hug, R. and Schachinger, M. (2006), Chinese Solar Modules Penetrating the German Market, Solar Reports, Solarserver Forum for Solar Energy.
33 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
IEA (2006), Trends in photovoltaic applications. Survey report of selected IEA countries between 1992 and 2005, Report IEA-PVPS T1-15:2006, International Energy Agency PVPS Task 1. Ingham, B. (1993), The meaning of development: Interactions between "new" and "old" ideas, World Development, 21 (11), p 1803. Jacobsson, S., Andersson, B.A. and Bångens, L. (2002), Transforming the energy system - the evolution of the German technological system for solar cells, SPRU Electronic Working Paper Series Paper No 84, Science and Technology Research, University of Sussex, Brighton, U.K. Jäger-Waldau, A. (2003), Research, Solar Cell Production and Market Implementation in Japan, USA and the European Union, PV Status Report, European Commission. Jäger-Waldau, A. (2006), Research, Solar Cell Production and Market Implementation of Photovoltaics, PV Status Report, European Commission. Jaus, J., Rentsch, J. and Preu, R. (2005), Determining Trade-Offs Between Costs of Production and Cell/Module-Efficiency with the Concept of Indifference Curves, 20th European Photovoltaic Solar Energy Conference and Exhibition, Barcelona, 6.- 10 June 2005. Jordan, D. (1995), Conversation during BP Solar plant visit, Sydney. Juma, C. and Clark, N. (2002), Technological Catch-Up: Opportunities and Challenges for Developing Countries, SUPRA Occasional Paper, Research Centre for the Social Sciences, University of Edinburgh, Edinburgh. Kathuria, V. (2002), Technology transfer for GHG reduction: A framework with application to India, Technological Forecasting and Social Change, 69 (4), p 405. Kyocera (2004), Kyocera Establishes a Quadriparite Global Production Framework for Solar Modules, Press Release 24-09-2004, Kyocera Corporation. Lall, S. (1992), Technological capabilities and industrialization, World Development, 20 (2), p 165. Lewis, J. and Wiser, R. (2005), Fostering a Renewable Energy Technology Industry: An International Comparison of Wind Industry Policy Support Mechanisms, Environmental Energy Technologies Division, Ernest Orlando Lawrence Berkeley National Laboratory. Li, Z. (2004), Made In China, Renewable Energy World, 7 (1), pp 70-80. Linden, L.H., Bottaro, D., Moskowitz, J. and Ocasio, W. (1977), The Solar Photovoltaics Industry: The status and evolution of the technology and the institutions, MIT Energy Laboratory Report - MIT-EL-77-021, US Department of Energy. Lüdemann, R. (2005), Experience and Expectation of Silicon Solar Cell Mass Production - Requirements for Next Generation Equipment, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. Margolis, R.M. (2003), Photovoltaic Technology Experience Curves and Markets, NCPV and Solar Program Review Meeting, Denver, Colorado, March 24, 2003. Milner, A. (2006), Industry Competitiveness and PV R&D: An Industry point of view, 3rd EPIA PV Industry Forum, in the Frame of the 21st EUPVSEC, Dresden, Germany. Mints, P. (2006), PV - The story so far, Refocus, November/December 2006, pp 32-36. Podewils, C. (2007), Dreaming of Dinosaurs, Photon International (March 2007), pp 114-123. Porter, M.E. (1985), Technology and Competitive Advantage, Journal of Business Strategy, 5 (3), p 60. Real, M. (2006), Presentation to Workshop: PV for Development - Ensuring highest quality for sustainability and scale-up, 21st European Photovoltaic Solar Energy Conference, Dresden, Germany, 4-8 September, 2006. Rogol, M. (2006), Interest in Solar, Photon International, June 2006, pp 102-105. Schaeffer, G.J., Alsema, E., Seebregts, A., Beurskens, L., de Moor, H., van Sark, W., Durstewitz, M., Perrin, M., Boulanger, P., Laukamp, H. and Zuccaro, C. (2004), Learning from the Sun Analysis of the use of experience curves for energy policy purposes: The case of photovoltaic power, Final report of the Photex project: Report ECN-C-04-035, ECN Renewable Energy in the Built Environment.
34 Chapter 2. The PV Industry and Modern-Sector Manufacturing in Developing Countries
Schmela, M. (2005a), Green light for solar power: China‘s PV industry awakens, Photon International, September 2005, pp 56-57. Schmela, M. (2005b), Shanghai Solar: OEM production for Sharp and others, Photon International, September 2005, pp 62-63. Schmela, M. (2005c), Super Sonic Solar Market - Worldwide Market Survey - cell & module production, Photon International (March 2005), pp 66-82. Schmela, M. (2006), Silicon Shortage - so what! Market survey on cell & module production 2005, Photon International (March 2006), pp 100-124. Sharif, N. (1992), Technological dimensions of international cooperation and sustainable development, Technological Forecasting and Social Change, 42 (4), p 367. Shum, K.L. and Watanabe, C. (2006), Photovoltaic deployment strategy in Japan and the USA-- an institutional appraisal, Energy Policy, In Press, Corrected Proof. Siemer, J. (2005), Saving intellectual property: Patents for photovoltaic applications, Photon International, November 2005, pp 36-40. Solarbuzz (2007), Solar Cell Manufacturing Plants, Accessed from: http://www.solarbuzz.com/Plants.htm, on: Jan 2007. Stewart, F. and James, J. (1982), The Economics of new technology in developing countries, F. Pinter; Westview Press, London; Boulder, Colorado. Surek, T. (2003), Progress in U.S. photovoltaics: looking back 30 years and looking ahead 20, pp 2507-2512 Vol.2503. Todaro, M.P. (1982), Economics for a developing world : an introduction to principles, problems and policies for development, 2nd ed, Longman, Harlow, Essex. Wang, S. (2006), Current Status and Future Expectation of PV in China, China-EU Energy Collaboration Forum, Shanghai, China, Feb. 20-21, 2006. Wanmali, S. (1998), Participatory Assessment and Planning for Sustainable Livelihoods, UNDP. Wassen, W. (2006), Presentation to Workshop: PV for Development - Ensuring highest quality for sustainability and scale-up, 21st European Photovoltaic Solar Energy Conference, Dresden, Germany, 4-8 September, 2006. Yang, H., Wang, H., Yu, H., Xi, J., Cui, R. and Chen, G. (2003), Status of photovoltaic industry in China, Energy Policy, 31 (8), pp 703-707. Zhao, Y. (2001), The present status and future of photovoltaic in China, Solar Energy Materials and Solar Cells, 67 (1-4), pp 663-671. Zhao, Y., Wang, S., Li, X., Wang, W., Liu, Z. and Song, S. (2006), Report on the Development of the Photovoltaic Industry in China (2004ዉ2005), China Renewable Energy Development Project Office.
35 36 CChhaapptteerr 33.. LLeeaarrnniinngg iinn tthhee IInntteerrnnaattiioonnaall PPVV MMaannuuffaaccttuurriinngg IInndduussttrryy
The aims of this chapter are to identify what is known about technological learning in the PV industry and national-level factors which affect learning within PV enterprises; and to therefore identify opportunities for new research into technological learning in PV manufacture which may have relevance for developing countries wishing to enter the PV market. Technological learning is defined in this thesis as the acquisition of the abilities to acquire, use and build upon technology effectively. Section 3.1 reviews existing studies on technological learning in the PV industry.1 Through this review, important characteristics of PV technology and likely sources of cost reduction in the industry are identified, enabling the identification of technical challenges that are likely to constitute barriers for manufacturers in developing countries. Section 3.2 reviews the literature on technological systems for PV in Germany, Japan, the US and Australia, which enables the identification of national-level factors influencing their relative success. While the literature identifies the technological changes and barriers that are likely to be important to PV manufacturers, it does not explain how learning occurs within enterprises, so cannot assist the identification of suitable capability building strategies. Additionally, the national-level factors cannot therefore be linked to specific learning processes, so their impact on learning cannot be fully understood.
1 The screen printed, wafer-based crystalline silicon solar cell technology that is most widely used in commercial solar cell production today is a mature technology that was first used in the 1970s Green, M.A. (2000), Photovoltaics: technology overview, Energy Policy, 28 (14), pp 989-998.. The technology still provides 90% of the total world production (Jäger-Waldau, 2006). Although it is believed that the industry will move towards thin-film technologies in the medium term, crystalline wafer based technologies are expected to dominate until at least 2010 (de Moor et al., 2004; Hoffmann, 2004b; Surek, 2003; Swanson, 2004). This chapter is therefore mainly concerned with the technology and industry requirements for the manufacture of wafer-based crystalline cells, since this technology is by far the most common and has been the primary technology implemented by Suntech (the subject of the case study in chapter 7). The operation of and typical manufacturing process for these cells is described in the Virtual Production Line (VPL) Software CD contained in Appendix 1 and summarised in the Chapter describing the VPL. 37 Chapter 3. Learning in the International PV Manufacturing Industry
3.1. A Review of the Literature on Technological Learning and Technological Change in Photovoltaic Cell Manufacture
In this section, the literature on technological learning in PV cell manufacture is reviewed. Through this review, the learning rate, sources of cost reduction and the direction of technological change are established, enabling identification of the technological challenges for latecomers to PV cell manufacture. A number of studies have been undertaken on technological learning in the photovoltaics industry. Some studies, such as those reviewed in section 3.1.1, have predicted cost reductions for photovoltaics manufacture on the basis of the historical learning rate. Others, reviewed in sections 3.1.2-3.1.3, have taken a bottom up approach; distinguishing the types of changes that have been, or will be responsible for cost reductions. Roadmaps and industry experts have also attempted to predict the direction and rate of radical technological change, as described in section 3.1.4.
3.1.1. Top-Down Studies: Trends in Cost Reductions Learning curves model empirically the reductions in the cost of production gained with experience. The first conception of the learning curve was formalised by Arrow (1962), and related to the productivity increases achieved with cumulative experience of the labour force in a plant. Later a more general expression, sometimes called the experience curve, including all the costs of manufacturing was applied, expressing the current unit cost of production as a function of the number of units produced:
−b C q S = D x T x CC 0 D T (Nemet, 2006a), E q0 U where:
Cx = the cost of producing the xth unit;
C0 = the cost of the first unit produced;
qx = the cumulative amount of units produced; and b = the experience index.
The progress ratio (PR) is the ratio of final cost to initial cost (Cx/Cx-1) with each doubling of output (PR = 2-b). The learning rate (LR) can be expressed as the percentage reduction in unit cost: LR=(1-PR) x 100%.
38 Chapter 3. Learning in the International PV Manufacturing Industry
Figure 3-1: Learning Curve showing reduced hours for aircraft assembly with cumulative production
Figure has been removed due to copyright restrictions.
Source: (Argote & Epple, 1990) Numerous empirical studies demonstrate that the more products have been produced in a firm, the more learning (cost reduction) has occurred. Learning, or experience curve studies are often used to predict the future cost of photovoltaics after a given amount of production has occurred, or to forecast the amount of time or industry support that is required to achieve learning sufficient to place photovoltaics on a price-competitive footing in a particular market. ‘Learning investments’, illustrated in Figure 3- 2, are the financial investments needed to reach a particular production volume and/or price.
Figure 3-2: An Experience Curve and Conceptual Illustration of the Learning Investment and the Role of Niche Markets
Figure has been removed due to copyright restrictions.
Source: (Shum & Watanabe, 2006)
39 Chapter 3. Learning in the International PV Manufacturing Industry
The learning rate for PV technology is faster than for many other energy technologies, implying that with sufficient learning investments, PV has the potential to become competitive in the longer term. Masini & Frankl (2003) compared the learning rates for different technologies (Figure 3-3), on the basis of cost per kWh of electricity generated by a system.
Figure 3-3: Learning curves and progress ratios (PR) for electricity generation technologies in the EU between 1980 and 1995
Figure has been removed due to copyright restrictions.
Source: (Masini & Frankl, 2003, p 41)
Although cost data for photovoltaics have produced fairly consistent learning rates, with an R2 value of >0.95 (Nemet, 2006a), there is some variation in the learning rate calculated for PV by different studies. Performance ratios of 74% (Maycock, 1997) and 83% (Strategies Unlimited, 2003, cited in Nemet, 2006a) have been published for module manufacture. A small variation in data upon which the figure is based may lead to a large variation in the amount of investment and cumulative production required, as illustrated in Figure 3-4. The more recent data is more reliable, but recent price stabilization, and in some markets, price increases due to silicon shortages, may provide at least an anomaly in learning curves, if not resulting in a new learning rate after the supply constraints have abated.
40 Chapter 3. Learning in the International PV Manufacturing Industry
Figure 3-4: The Sensitivity of the Learning Rate to Data
Figure has been removed due to copyright restrictions.
Source: (Nemet, 2005)
Government support for PV technology is dependent on confidence in the technology as a future cost competitive option for electricity generation. Market growth therefore depends on convincing and reliable forecasts of learning rates and technological progress generally. Assumptions based on learning rates are problematic because of sensitivity to variations in data and also because the learning rate has not been consistent historically (Figure 3-5), and therefore cannot be expected to remain constant and be predictable in the future. Although we cannot blindly rely on past learning rates for PV to continue, Andersson & Jacobsson (2000) note that a progress ratio of 0.8 is not unusual in the electronics industry. Since PV is a physical, rather than information technology, the cost reductions that have been achieved as a result of dematerialisation and miniaturisation in the electronics industry do not apply. Andersson & Jacobsson (2000) believe that the potential for increases in the scale of production and better cell designs could, however, feasibly maintain a PR better than 0.8.
Figure 3-5: Learning curve, with different sections, for PV modules
Figure has been removed due to copyright restrictions.
Source: (van der Zwaan & Rabl, 2004, p 1550)
41 Chapter 3. Learning in the International PV Manufacturing Industry
Learning curves tell us how fast to expect technological change to occur and facilitate comparison between progress in different technologies and calculations of the investments required in a particular technology. Learning curves are therefore useful tools for technology choice and policy formation. Learning curves, however, do not tell us anything about how these cost reductions are achieved or how the capability building strategies of firms may impact learning. Bottom up studies, which identify the sources of cost reductions indicate the nature of technological change in the PV industry. These studies are now examined, and the implications of the technological trends for latecomer firms are identified.
3.1.2. The Distribution of Costs in PV Production Cost reductions are the main source of competitive advantage in the industry. The most cost intensive aspects of PV production offer the greatest opportunities for learning and are often the focus for search-based learning. It is instructive to identify the most costly aspects of production, since the ability to compete on cost will be a strong determinant of the ability of latecomer firms to survive. The costs of PV production are the sum of the equipment and materials costs for each process; the cost of any associated IP; the utilities and services (such as electricity, deionised water, compressed air, exhaust systems, waste treatment etc.) required by the plant, the costs associated with the establishment and use of the premises; and the skilled and unskilled labour costs. A study carried out by Arthur D Little in 2000 (Frantzis et al., 2000) found that the cost of materials is the largest in the production of photovoltaic modules, particularly in the case of the dominant wafer-based technologies. Authors have attributed 54% (Ghannam et al., 1997), 52.5% (Rohatgi, 2003), 50% (Jester, 2002) and 40% (Frantzis et al., 2000) of the total direct costs of manufacturing wafer-based multicrystalline modules to crystal growth and wafer slicing. When the cost of multicrystalline wafer production was examined more closely in the Arthur D Little study of US manufacturers in 2000 (Frantzis et al., 2000) (Figure 3-6), it was found that materials (primarily silicon feedstock) accounted for 70% of the wafer cost and 30% of the total cost of the module. Cell production has been found to account for 17% (Rohatgi, 2003), 20.7% (Frantzis et al., 2000) and 30% (1996 music-FM study, (Bruton, 2005)) of module manufacture, around US$0.34-0.44/Wp. In US cell manufacture, labour (38%) and depreciation of plant and equipment (32%) were found by Frantzis et al. (2000) to be the largest costs, while materials (20%) and yield losses (10%) were also significant. Developing countries with low labour costs therefore have a potential advantage, as do countries that can manufacture low-cost equipment and materials. Reduced material use and yield improvements are likely to emerge from
42 Chapter 3. Learning in the International PV Manufacturing Industry
improvements in manufacturing technology, such as improved wafer handling and quality control, as well as new device designs.
Figure 3-6: Manufacturing Costs by Process for a 10MW mc-Si Plant in 2000
Figure has been removed due to copyright restrictions.
Source: (Frantzis et al., 2000)
3.1.3. Bottom-Up Studies: Sources of Cost Reductions The cost of production has declined throughout the development of PV technology as a result of changes in production such as: Process innovations and automation improvements; Economies of scale; Changes in cell design; and Changes in prices for materials, equipment and labour. Learning curves aggregate all of these factors and the cost reductions are often interpreted as automatic improvements resulting from learning by doing. Work has been carried out by Nemet (2005; 2006a) on disaggregating the sources of cost reduction seen in the empirical data of learning curves. Nemet used data from the literature documenting the historical cost reduction of different factors, and constructed a numerical model of the learning rate for PV from these figures (Table 3-1).
Table 3-1: Cost Reductions and Sources of Cost Reduction in PV Manufacture 1980-2001 Factor Change Effect on Module Cost ($) Plant size 125kW/yr 14 MW/yr -9.22 (43%) Module Efficiency 8.0% 13.5% -6.50 (30%) Si Cost 131 $/kg 25 $/kg -2.69 (12%) Wafer Size 45 cm2 180 cm2 -0.67 (3%) Si Consumption 28 g/W 18 g/W -0.59 (3%) Yield 82% 92% -0.43 (2%) Poly-crystal 0% 50% -0.38 (2%) Sum of Factors -20.48 (95%) Actual Change -21.62 (100%) Residual -1.14 (5%)
Source:(Nemet, 2005, pp 13, 14)
43 Chapter 3. Learning in the International PV Manufacturing Industry
Scaling up and increases in module efficiency together account for 73% of the cost reductions between 1980 and 2001 in Nemet’s model. Reductions in the cost and consumption of silicon material, increased wafer size, improved yield and the transition to multicrystalline material were found to account for almost all of the remaining cost reductions. Frantzis et al. (2000) predicted that direct manufacturing costs for multicrystalline modules (in 2000 US dollars, excluding overhead costs and profits) could fall from US$2.10/Wp to $1.15/Wp by 2020. Scale is expected by Frantzis et al. to remain the biggest factor in cost reduction, contributing 25%. Increases in cell efficiency are expected to continue, reaching 18- 20% within the next few years, and reducing manufacturing costs by 17%. Reduced silicon costs and utilisation are projected to continue, reducing costs by 8% and yields to continue to improve, reducing costs by 4%. Rohatgi (2003) also performed cost analysis on the impact of improvements in cost-sensitive factors in PV manufacture. His cost reduction projections are compared with those of Frantzis et al. (2000) in Figure 3-7. Although the two authors emphasise the importance of different sources of future cost reductions, the sources are consistent with each other and with past cost reductions, and can be accepted as the most likely sources of future cost reductions. Factors influencing the achievement of these cost reductions will be discussed in the following sections.
Figure 3-7: Cost reduction prospects for silicon modules
2.50
2.00 0.36 0.31 0.07
0.27 1.50 0.42 Present Cost - Efficiency Improvements - Yield Increses - Scale Increases 0.10 0.33 - Materials Cost Reductions
Cost ($US/Wp) Cost 1.00 New Cost
1.15 0.50 1.07
0.00 Frantzis et. al., 2000 Rohatgi, 2003
Sources: (Frantzis et al., 2000; Rohatgi, 2003)
44 Chapter 3. Learning in the International PV Manufacturing Industry
3.1.3.1. Scale Increases Nemet (2005) found that the largest single factor influencing cost reductions in photovoltaics between 1980 and 2001 was increased plant size. The commissioning of increasingly high throughput production plants, resulting in continuing scale-related cost reductions is expected (Bruton, 2002; Rohatgi, 2003; Swanson, 2004). According to Frantzis et al., (2000) future scale-related cost reductions will be achieved through high volume purchasing of materials and equipment, better potential for balancing of production lines at larger scales and the use of larger equipment that represents a smaller depreciation cost per unit produced. Although cost reductions have been achieved through scale increases, the production technology, as yet, has not undergone radical change to accommodate larger scale (Lüdemann, 2005; Menanteau, 2000). Technologies such as EFG and silicon ribbon technologies have the potential for continuous, rather than batch processing and would constitute a radical transformation, accessing opportunities for larger cost reductions. Although most successful PV firms are expanding, in many cases they expand by adding incremental capacity, rather than installing a new, large-scale plant, thereby missing out on potential scale economies. New firms have the advantage of starting from scratch with the latest technology and often at a large scale. The capacity of single lines is currently 15-20 MWp, but is expected to increase to 100 MWp, within facilities containing 10 of these production lines
(Lüdemann, 2005). These 1GWp plants will be capable of producing a module per second (Podewils, 2007). Firms with little PV specific experience have been able to scale up very rapidly. Q-cells began manufacturing in 2001 and is already the second largest producer worldwide. Mitsubishi Electric had experience in PV production and R&D, but in 1997 had very little industrial-scale production experience in terrestrial PV, and was able to expand to 230MW by 2006 (Nemet, 2006b). For developing countries, this implies that new large-scale production facilities could potentially be established and, combined with lower labour costs, could produce cost- competitive product.
3.1.3.2. Efficiency Improvements Efficiency improvements may result from improved fine-tuning of processes, or the development of new processes. The large gap between the efficiency of the best laboratory cells (25% in the case of sc-Si) and mass-produced cells (around 15%) illustrated in Figure 3-8 suggests that the commercialisation of novel, advanced processing techniques will bring down
$/Wp costs substantially.
45 Chapter 3. Learning in the International PV Manufacturing Industry
Figure 3-8: Crystalline PV Efficiency: Highest laboratory Cells vs. Average Commercial Modules.
Figure has been removed due to copyright restrictions.
Source: (Nemet, 2005, p 15)
Experts agree that efficiencies of crystalline silicon cells will improve, but predictions on the size of the improvement vary. Frantzis and Rohatgi predicted that commercial cell efficiency would increase to at least 17% for multicrystalline silicon and 18% for single crystal silicon by 2010 (Frantzis et al., 2000; Rohatgi, 2003). Swanson (2004) predicted efficiencies of 21% by 2012 and the European Commission Roadmap for PV R&D identifies a target of 26% by 2010 (de Moor et al., 2004). It is certain that PV firms will need to improve substantially and continually to remain competitive. Some of the processes used for laboratory cells are not cost effective in the manufacturing environment, but companies such as BP Solar, Sanyo and Sunpower have already introduced advanced processing techniques for their laser-grooved buried contact (LGBC), Hetrojunction with Intrinsic Thin Layer (HIT) and A-300 high-efficiency, all back contacted technologies respectively. Each of these technologies have resulted in appreciable efficiency increases. Frantzis et al. (2000) expect that better process control and improvements in cell design, such as reductions in front contact shadowing (for Cz wafers), hydrogen passivation (for mc-Si wafers) and back surface modification (for mc-Si wafers) will be sources of efficiency improvements in the wafer-based technologies. Swanson (2004) also predicts reduced shading losses through finer screen printed contacts and improved back surface fields to reduce recombination at the back surface. He additionally anticipates efficiency improvements through advanced cell designs such as a selective emitter design that will decrease sheet resistance and recombination, better texturing and high lifetime silicon through improvements in crystal growth technology.
46 Chapter 3. Learning in the International PV Manufacturing Industry
In view of the strong material dependence of PV costs, it is likely that latecomer firms will need to improve efficiencies in order to stay cost competitive. The required improvements in cell design are likely to be obtained only through deliberate and sustained R&D efforts, particularly challenging for latecomer firms. Latecomers, however, do have the advantage that implementing new technology in new plant is likely to be cheaper than retrofitting existing capacity.
3.1.3.3. Material Cost Reductions Since the quality of the silicon material required for solar cells is lower than that required in the microelectronics industry, off-specification wafers discarded from the electronics industry can be used to make solar cells. Until recently, this waste has been sufficient to provide all the feedstock for the PV industry (Schaeffer et al., 2004). With the expansion of the PV industry, however, the consumption of feedstock has surpassed the waste silicon available from the microelectronics industry. This, combined with the resurgence of the IT industry, has created a severe shortage. Silicon manufacturers are in the process of developing new capacity for solar-grade (lower purity) silicon, which will be produced specifically for the PV industry to cover the shortfall. Building production capacity for silicon feedstock is extremely capital intensive and has a long timeframe (Mints, 2006a). Nevertheless, Jäger-Waldau (2006) predicts that the shortage should be overcome within the next 2-3 years, and Schaeffer et al. (2004) speculated that prices of $10-15/kg, reduced from $25/kg (Nemet, 2006a), could be realised with the entrance of solar grade silicon, which does not need to comply with such stringent purity requirements as the silicon used for the microelectronics industry. As a result of the short-term shortages, however, the price of silicon has risen and cell manufacturers have been forced to sign long term contracts for silicon material (Jäger-Waldau, 2006). In 2004, silicon feedstock prices ranged from $20-65/kg, increasing to $90/kg on the spot market at the end of 2005, with reports of $200-400/kg on the spot market in 2006 (Mints, 2006a; Schaeffer et al., 2004). The shortage of silicon feedstock in the industry is a barrier to the entry of new firms, as many suppliers have already committed their future production to existing contracts. Countries that are latecomers to the PV industry may benefit by setting up Si manufacturing capacity, which may also have cost advantages. There are currently a small number of suppliers of silicon feedstock. New demand and technology for solar grade silicon is increasing the number of new entrants into the silicon market. Rohatgi (2003) believes that economies of scale in materials production and increasing competition between a larger number of suppliers will bring the prices for materials down. Reductions in the amount of silicon material used and increases in efficiencies will also help to reduce expenditure on silicon. The use of thin-film technologies is one route to reduction
47 Chapter 3. Learning in the International PV Manufacturing Industry
in silicon use. Thin film technology experienced an increase in production of over 50% from 2004 to 2005 (Jäger-Waldau, 2006), with new players entering production and established conventional cell manufacturers diversifying into thin films. Less mature thin-film technologies are likely to be more challenging for developing countries, as discussed in the following section (3.1.3.4). In the wafer based technologies, manufacturers are reducing materials costs by using progressively thinner wafers. Silicon use can also be reduced through innovative processes such as the recycling of slurry during wafer slicing (Rohatgi, 2003) vapour etching of wafers and reduced kerf loss from 200 m to 130 m through improved slicing of wafers (Frantzis et al., 2000; Swanson, 2004). Nemet (2006a) reports that the amount of silicon used in PV production has decreased from 30g/Wp to 18g/Wp between 1975 and 2001. Schaeffer et al. (2004) speculate that the amount of silicon material used may be reduced from 17g/Wp to 10g/Wp by 2010. The use of thinner wafers in response to silicon shortages is constrained by yield losses as the thinner wafers are more fragile and difficult to handle (Wenham & Bruce, 2002). Reducing material use will require firms to improve cell handling. Sourcing production equipment and materials other than silicon is also problematic in the infant PV industry. A US industry workshop (NCPV, 1999) stated that the lack of dedicated materials suppliers, lack of standardisation in the materials and equipment requirements of manufacturers to the industry and lack of a well-established distribution infrastructure for materials were impediments to the manufacturing industry. Milner (2006) also sees standardisation as a potential source of cost reduction.
3.1.3.4. Manufacturing Cost Reductions Wafers must be moved from one process (piece of equipment) to another in a solar cell production line. In some cases the transfer is done manually, particularly when the cells are transferred from a continuous process, such as a belt furnace, to a cartridge for a batch process, such as a chemical etch or a plasma etch. Between two continuous processes, belts can be used to transfer the cells. Automated cell handling is usually preferred because yield losses and labour costs are generally reduced. Automation also facilitates inline production monitoring and control. Swanson (2004) cites increased automation ‘from none to some’ as one of the most important eight trends facilitating cost reduction in manufacture between 1979 and 2002. Manufacturers are actively looking to further develop continuous-type processes to replace batch processes and the European crystalline R&D roadmapping project ‘Crystalclear’ focuses on inline processing as one of the two most important sources of future cost reduction (Sinke et al., 2006).
48 Chapter 3. Learning in the International PV Manufacturing Industry
Higher Yield through Manufacturing Improvements The use of expensive silicon material in photovoltaics manufacture makes high manufacturing yields a particularly important factor in cost reduction. Yield losses in solar cell manufacture constituted 10% of total costs for US manufacturers in 2000 (Frantzis et al., 2000), where the average yield for finished modules was found to be 85%. Nemet (2005) quoted 92% yield for modules, an improvement of 7% over five years. Hegedus & Luque (2003) gave a figure of 95% yield for the production of sc-Si cells, similar to Lüdemann’s (2005) estimate of less than 90-95% for cells. Lüdemann (2005) expects yield for cell production to ultimately reach 99%. Higher yields may be achieved through manufacturing improvements such as: Reduced breakage through improved cell handling. Improved consistency in processing through improved monitoring and adjustment of processes, such that more wafers are processed according to specifications. Better maintenance and operation of equipment. (Frantzis et al., 2000) believes that the elimination of gross failures of wire saws for slicing wafers, for example will be a significant factor in improving yields.
The difference between poor yield and good yield may be 80-95% (15% of output). Consequently, manufacturers who cannot achieve high yields will not be able to produce cells cost competitively.
More Efficient Plant Operation Good industrial engineering and coordination can also lead to higher equipment utilisation, labour savings and space savings. Computer production control systems are not commonly used in PV production, but are expected to become standard. These systems will enable in-line process control and testing; and simulation of processes, materials and personnel flow and equipment layout; improving process monitoring and control, revealing bottlenecks and enabling coordination of supplier contracts (Eberhardt, 2005). In addition to better efficiency and yield, good production control can lead to cost savings through better line balancing and utilisation of the capacity of equipment. The uptime of equipment is currently 80- 85%, but is expected to increase to 98% (Lüdemann, 2005), through production control and the implementation of improved and preventative maintenance.
The Use of Thinner, Larger Wafers
The use of larger wafers, which may be processed at lower cost per Wp, or thinner wafers, which consume less material, also depends on improved cell handling. Manufacturing costs have been found to be highly dependent on the number of cells processed, and to have a low dependence on the area of the cell (Ghannam et al., 1997), since the same equipment set
49 Chapter 3. Learning in the International PV Manufacturing Industry
can process the same number of cells, regardless of size (Haase, 2005). Larger cell sizes can therefore reduce costs in the operation of equipment, and save resources in transferring cells about the plant. Larger cells, however, suffer from increased series resistance and both thin wafers and large wafers are more difficult to handle, making yields harder to maintain. Nevertheless, cell sizes are projected by US experts to continue to increase from 125mm to 200mm square (Swanson, 2004), while wafers are expected to become thinner as discussed in the previous section.
Reduced Labour Costs through Automation Continuous processes are often more cost effective than batch processes; since personnel are not required to transfer the wafers, reducing labour costs (see Figure 3-9), as well as increasing throughput, as previously discussed.
Figure 3-9: Labour Productivity using Manual compared to Automated Processing at Siemens Plants
Figure has been removed due to copyright restrictions.
Source: (Jester, 2002)
Lüdemann (2005) expects that employees per MWp will be reduced from 8 employees to 0.5 employees as PV production is fully automated. Reduced handling is also expected to increase yields and reduce the risk of contamination of the wafers. For firms in developing countries, the cost of labour is not a key issue, so automation may not provide cost reduction benefits to the same extent as it does in industrialised countries and hence reduce one of their competitive advantages. Nevertheless, manufacturers relying on manual cell handling would be expected to be at a disadvantage as they would not obtain improved yields through reduced breakage, reduced contamination, more consistent processing and improved maintenance; particularly important since many advanced firms are likely to be using larger, thinner wafers.
50 Chapter 3. Learning in the International PV Manufacturing Industry
3.1.4. The Rate and Direction of Technological Change in Photovoltaics The rate and direction of technical change will determine the extent to which firms need to adopt or develop innovative products, and in which direction they need to develop. Technical change in photovoltaics is primarily concerned with reducing the $/Wp cost, aiming to achieve cost competitiveness with conventional electricity sources, either by improving the dominant silicon wafer based technology, or by developing new device designs. Many firms in the industry, including many of the largest established producers as well as new entrants, continue to manufacture and install new capacity in the older technology, relying on incremental improvements to justify the long term investment. The development of innovative processes for standard solar cells has allowed these manufacturers to continue to improve efficiencies and yields and reduce material costs for the wafer-based technology. The efficiency of commercial cells has improved from 8.0% to 13.5% between 1980 and 2001 (Nemet, 2006a), and improved efficiencies are predicted to comprise between 15% (Rohatgi et al., 2003) and 17% (Frantzis et al., 2000) of future cost reductions by 2010. The difference in the efficiency of cells produced by a good manufacturer and a poor one is around 2-4%, which, given efficiencies of 15%, translates to around 20% of the Wp output of the product. Firms who are not able to continually improve their efficiency will not be competitive in PV markets.
Expert Predictions Hoffman (2004b) classifies the available device technologies as crystalline silicon, thin film, III-IV compounds and new concepts. Within the crystalline silicon area, the majority of production uses silicon wafer-based technology, but some firms are producing cells via a continuous ribbon process. The thin film area includes amorphous silicon and II-VI compounds, used to create thin films of semiconductor materials. More recently, the deposition of polycrystalline silicon material on substrates has been commercialised. Thin film PV modules made from a number of II-IV materials such as CdTe and GaAs have also been commercialised. New concepts include dye cells, organic cells, and high efficiency (third generation) technologies. Experts predict that although wafer-based technologies will continue to dominate for the next decade (de Moor et al., 2004; Surek, 2003; Swanson, 2004) some of the thin film technologies are likely to become more feasible in the medium term (de Moor et al., 2004; Green, 2003; Hoffmann, 2004a, b; Jäger-Waldau, 2004, 2005, 2006), as they have the potential to achieve a better $/Wp outcome, use less silicon material and are better suited to mass production techniques (Harmon, 2000). Silicon shortages provide opportunities for accelerated development of thin film technologies. The deposition of semiconductor material on large areas of glass or other substrates and later cutting to size involves processes that are better suited to
51 Chapter 3. Learning in the International PV Manufacturing Industry
large scale production (fewer processing steps, continuous processing and larger production units) than wafer slicing and processing of many small, fragile units (Koch et al., 2005; Menanteau, 2000). Greater scale benefits may accrue to firms who can implement these technologies. Experts also expect that II-IV compounds and other new cell types such as dye-cells and organic cells will take over in the long term (Hoffmann, 2004b). (Hartley, 2006) believes that there are a range of technologies with the potential to reach similar cost competitiveness. The second largest PV cell manufacturer, Q-cells, has adopted a policy of a broad ‘technological footprint’, investing in technologies including silicon ribbon, thin film and a-Si technologies.
Production Data Production data (Figure 3-10) illustrates the continuing dominance of the wafer based technologies, while silicon ribbon technologies, a-Si and CIS technologies have failed to deliver the promised $/Wp decreases and have fallen out of favour in the commercial environment whereas Cd-Te technology may be finding a foothold. There are a variety of other technologies being manufactured at smaller scales (mostly less than 10MW), but the novel technologies have been unable to scale-up effectively to date (Rohatgi, 2003).
Figure 3-10: Percentage of Production by Technology
Figure has been removed due to copyright restrictions.
Source: (Photon, 2006)
52 Chapter 3. Learning in the International PV Manufacturing Industry
Patent Data Patents are indicators of innovative activity, and may be used to get an idea of the rate and direction of technical change (Andersson & Jacobsson, 2000).2 Patent data indicates that there is a significant amount of innovative activity in the PV industry. The worldwide total number of patents pending is much higher than the average for all industries during the period 1990-2004 (Visentin et al., 2005). The number of applications is increasing, from 5-600 per year in the early 1990s, to more than 1900 applications per year between 2002-2004 (Siemer, 2005). This would indicate that firms are unlikely to survive in the long term unless they are able to commercialise innovative products. Patent application figures from 1990-2004 (Figure 3-11) reveal a large increase in innovative activity; particularly in the area of a-Si technology, but also in crystalline and polycrystalline silicon devices, organic cells and PV building integration.3 The patent data appears to confirm the expert predictions that thin films will supersede bulk technologies in time.
Figure 3-11: Global development of patent applications in the area of PV between 1990 and 2004
Figure has been removed due to copyright restrictions.
Source: (Siemer, 2005, p 40)
2 It must, however, be recognised that patents are an imperfect measure of innovative capability, since some innovations (such as innovative production processes) may not be visible from the product, but are easy to imitate once they reach the public sphere via a patent application, thus discouraging patenting. 3 The majority of patents have been registered by Japan. Although Japan is indeed an important centre for PV R&D, Japanese researchers tend to patent every detail of an innovation, whereas in Europe and the US, researchers try to include as much as they can under one patent. Since Japan is particularly interested in a-Si technology, and less interested in III-V compounds and organic cells, the data is skewed towards a-Si (Visentin et al., 2005). 53 Chapter 3. Learning in the International PV Manufacturing Industry
3.1.5. A Summary of Technical Challenges for PV Cell Manufacturers in Developing Countries The probable sources of future cost reductions in PV manufacture and the routes by which these are likely to be achieved have been identified. Table 3-2 summarises these routes to achieving cost reductions and the specific barriers or challenges for latecomer firms. The latter will be examined in more detail in later Chapters, as the environment for capacity building, technology innovation and manufacture in developing countries are explored.
Table 3-2: Technical Challenges and Sources of Improvement in PV Manufacture Challenge for Cost Reduction Route to Better Performance latecomers High volume purchasing of materials & equipment Scale Cost More balanced production line High capital Reductions Larger equipment requirements New technology better suited to large scale Fine tuning of processes and production line Advanced QC tests leading to better process Clean environment & stability high purity materials Efficiency Increases in Good quality or improved equipment (must be Complex processes Existing Process imported) and interactions Technology Better efficiency (consistency) through better Manual handling may process control result in Reduced contamination through automation contamination (reduced handling) Less shadowing via thinner contacts (improved screen printing, photolithography) R&D and innovative High lifetime through gettering, front & back capabilities difficult to surface passivation (Hydrogen passivation via Efficiency Increases develop SiNx or other, Back surface field via Al alloying or via New Processes Customised Boron doping, Rapid firing of contacts) equipment required Selective emitter for new processes Texturing High lifetime through crystal growth improvements Reduced wafer thickness Reduced kerf and slurry recycling (in wafer slicing) Obtaining silicon Materials Cost Vapour etching of wafers supplies Reductions Solar grade silicon Procuring materials More competition in silicon supply locally New technology using less silicon (thin film etc) Production monitoring (routinized QC leading to higher yields) & automated production control Quality control not (better utilisation, reduced bottlenecks via line well understood balancing, JIT logistics) Sophisticated Better maintenance & preventative maintenance equipment Manufacturing Cost (reduced downtime and better process stability) Manual handling is Reductions Better (automated) cell handling (Higher yield and likely to result in lower cost through better handling of large thin lower yields. wafers, less labour requirement) Expensive material Standardisation of equipment and materials requires high yields New processes better suited to mass production R&D and innovative New Device Design Commercialise thin film or high efficiency devices capabilities difficult to develop
54 Chapter 3. Learning in the International PV Manufacturing Industry
3.2. National Experiences with PV Manufacture
Because of the high price of photovoltaics relative to conventional sources of electricity in most markets and the positive environmental externalities that may be obtained through the increased generation of electricity from renewable sources, the photovoltaics industry has been strongly affected by political support for market development, R&D, and also by environmental regulations. The PV industry has been able to steadily expand in the past decade, as costs have been reduced and new markets have become accessible; often with the support of public programs. The cases of Germany, Japan, the US and Australia, which have all been important in the development of PV, are described in this section.
3.2.1. The Case of Germany The federal government in Germany has provided significant R&D funding for PV since 1978. Between 1977 and 1990, 18 universities, 39 firms and 12 research institutes were involved in the German R&D programs (Jacobsson et al., 2002). The first German demonstration project was a 300kWp system installed in 1983 (Jacobsson & Lauber, 2006). Strong public opposition to nuclear power materialised after the Chernobyl nuclear accident in 1986, boosting support for PV as an alternative. The German Solar Energy Industries Association, formed in 1978, was joined by the German Society for Solar Energy in 1986. The new society included members from industry and the public, and disseminated information and provided advice to most political parties. Eurosolar was formed in 1988, and operated within the political structure, with members from all political parties except the Liberals. In the context of the anti-nuclear sentiment, these groups were able to effect a number of institutional changes (Jacobsson & Lauber, 2006). The first German renewable energy feed in law was introduced into some regions in 1991, and along with the 1000 roofs programme (1991-1995) (Stryi-Hipp, 2006), encouraged the entry of many new installers and led to the development of new inverter designs and standards for installation and grid-interconnection (Jacobsson et al., 2002). The initial feed-in law, under which 90% of the average consumer tariff was paid to renewable energy generators, was supported by all political parties, but did not have much direct impact on investment in photovoltaics, since PV electricity was still too expensive (Jacobsson & Lauber, 2006) and the market depended on imports. It was, however, an indication of the political strength of the renewable energy lobby and the growing acceptance of the technologies that had occurred through effective lobbying over these programmes (Balaguer & Marinova, 2006). Government R&D funding continued at a steady level (IEA, 2006a), although until the 1990s there remained
55 Chapter 3. Learning in the International PV Manufacturing Industry
an unwillingness to invest in bringing prototypes into production (Jacobsson & Lauber, 2006). Despite the 1000 roofs program and the feed-in tariff, the domestic market remained small, and manufacturers were still operating with losses. There was little investment in PV production facilities (Jacobsson et al., 2002). Firms preferred to expand production in the US, where the industry and market had been established much longer. Significant expansion of cell and module production did not occur in Germany until the late 1990s, when a critical mass of political pressure from the infant industry that had emerged through the initial market formation programs, and from anti-nuclear and renewable energy lobbyists, encouraged large market formation programs, including the large scale 100 000 roofs program and the revision of the first feed in law in 2000 (Jäger-Waldau, 2006) (Figure 3-12). The new feed in tariffs started at about 50 euro cents per kWh and were guaranteed for 20 years (EPIA, 2004). The tariffs have since been reduced by 5% per year in order to encourage cost reductions. The 100 000 roofs program also included low-cost loans that were attractive to investors.
Figure 3-12: The German Market for Photovoltaics 1990-2003
700 Second Feed-in Law 600
500
400
MWp First 300 Feed-in Law
200 100 000 Rooftop Programme
100
0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Sources: (IEA PVPS, 2007b; Stryi-Hipp, 2006)
In addition to support for photovoltaics, the strength of wind industry actors and networks enabled advocates of renewable energy to maintain the feed in laws despite opposition from the existing utilities (Haas, 2003). Significant market growth resulted in the entry of many firms, including solar cell manufacturers. Jacobsson et al. (2002) report that ASE and Shell invested in new production facilities in Germany in 1998, but only after assurances of sustained
56 Chapter 3. Learning in the International PV Manufacturing Industry
market support from government. Between 1996 and 2000, the number of manufacturers increased from two to six (Jacobsson et al., 2002), with the entry of Schott Solar, Ersol, Solarworld and Q-cells. During this period, the government directly funded 40% of plant establishment costs, excluding building costs, in three of these cases (Balaguer & Marinova, 2006). Other firms have also benefited from public as well as private investment. Between 1990-1999, an additional 15 universities, 41 firms and 17 research institutes received federal R&D funding (Jacobsson et al., 2002). This funding was designed to encourage research that could be applied in industry and required industry-academic collaborations, and industry representatives have been represented on academic boards (Jacobsson et al., 2002). The strong knowledge base and incentives for cooperation in German research institutes and universities has been attractive to firms seeking research collaborations (Jacobsson et al., 2002). For example, Sunways has carried out cooperative research at the University of Konstanz; Antec Solar was started by members of the Batelle Institute in order to commercialize technology developed there; and Wurth Solar was established when researchers at the University of Stuttgart were seeking finance for pilot production of CIS technology (Jacobsson et al., 2002). Researchers from the university moved to Wurth at its inception, and Wurth has relied on (largely federally funded) R&D at the university since, preferring not to establish their own R&D facilities. Germany also continues to provide incentives for manufacturers to locate there (Mints, 2006b). The strong industrial base in Germany has enabled beneficial interactions with suppliers. For example, joint ventures between PV firms and chemical firms such as Wacker Chemie and Degussa have been formed to develop solar grade silicon (Balaguer & Marinova, 2006). German vertically integrated manufacturer Schott Solar is part of a multinational whose base product is special glass. SolarWorld grew out of a renewable energy distribution business, and acquired Bayer Solar, a subsidiary of the multinational chemical and health corporation. Significant cost reductions in system integration and the development of innovative BIPV products have also been achieved as knowledge has been created through the interactions of designers, builders and systems integrators (Balaguer & Marinova, 2006). The feed in tariff has created in Germany the largest market for PV in the world. In 2005, 47% of all modules were sold in Germany (Mints, 2006b). Certain and sustained markets created through regulatory measures have given manufacturers the confidence to invest in new production capacity, and has created space for learning by doing throughout the value chain. The growing size and number of firms in the industry and the increasing strength of political advocates for the technology have in turn increased the legitimacy of the technology, and given the PV industry strength in the battle over the institutional framework.
57 Chapter 3. Learning in the International PV Manufacturing Industry
3.2.2. The Case of Japan Japan has had a long history of involvement in the development of PV. Sharp began to mass produce PV cells in 1963 (Jäger-Waldau, 2006), produced cells for a Japanese satellite in 1974, and incorporated cells in a calculator in 1980 (Foster, 2005). Sanyo began R&D in 1970, entered a-Si production in 1980 and began to mass produce their advanced hybrid HIT technology in 1997 (Foster, 2005). Japan recognises the importance of PV in the context of its energy policy which is based on the principles of energy security, economic efficiency and harmony with the environment (Jäger-Waldau, 2006). The PV industry is also valued because of its interdisciplinary nature and the potential for spillovers with other technologies. Shrinking markets for heavy machinery, a traditional Japanese industry, and the growth of PV led to PV being classified as a “key industry”. The Japanese MITI (Ministry of International Trade and Industry) recognises PV as a generic strategic technology, and has supported PV R&D and market development through the Sunshine Project, initiated in 1974 and the New Sunshine Program (1994). The Japanese Sunshine programs have intentionally engaged actors from a range of industry sectors, and have aimed to stimulate cross-sectoral and inter-disciplinary technology spillovers (Watanabe et al., 2000). Under the first program, an R&D association (PVTEC) was established in 1980 (Nagamatsu et al., 2006). The association included firms from chemical, petroleum, textiles, coal, ceramics, metallurgical and electrical machinery industries as well as public institutes, as listed in Table 3-3. There are many firms that have been involved in PV R&D, but not in solar cell production. Most of the Japanese PV manufacturers have electronics as their core business (e.g. Sharp, Sanyo, Mitsubishi Electric, Canon). Traditional heavy industry companies such as Mitsubishi Heavy Industries and car makers such as Honda have also entered the market (Jäger-Waldau, 2006). The technological proximity of PV to the semiconductor industry has both given Japan an advantage, and encouraged the government to see its strategic value in terms of inter-industry spillovers (Nagamatsu et al., 2006).
58 Chapter 3. Learning in the International PV Manufacturing Industry
Table 3-3: Firms participating in the Photovoltaic Power Generation Technology Research Association (PVTEC) Textiles (1) Teijin Ltd Chemicals (5) Kanegafuchi Chemical Industry Co., Ltd, Shinetsu Chemical Co., Ltd, Diado-hoxan Co., Matsushita Battery Industrial Co., Ltd, Mitsui Toatsu Chemicals Inc. Petroleum and coal Showa Shell Sekiyu K.K., Tonen Co., Japan Energy Co. products (3) Ceramics (3) Asahi Glass Co., Ltd, Kyocera Co., Nippon Sheet Glass Co., Ltd Iron and steel (1) Kawasaki Steel Co. Non-ferrous metals Osaka Titanium Co., Hitachi Cable, Ltd, Mitsubishi Materials Co. and products (3) Electrical machinery Oki Electric Industry Co., Ltd, Sanyo Electric Co., Ltd, Sharp Co., Sumitomo (8) Electric Industries Ltd, Hitachi, Ltd, Fuji Electric Corporate Research and Development Ltd, Matsushita Electric Industrial Co., Ltd, Mitsubishi Electric Co. Public institutes (2) Japan Measurement and Inspection Institute, Central Research Institute of Electric Power Industry Source: (Watanabe et al., 2000)
The PV R&D expenditure of all the firms involved in the Japanese PVTEC R&D program increased with a time lag of approximately one year after the MITI funding (Figure 3- 13), demonstrating the success of the Japanese program in stimulating private investment in R&D (Watanabe et al., 2000). The government also provided direct subsidies for the establishment of manufacturing, bringing production costs and system prices down (Mints, 2006b).
Figure 3-13: The relationship between public and private R&D investments
Figure has been removed due to copyright restrictions.
Source: (Watanabe et al., 2000, p 301)
In 1994, Japan initiated the 70 000 roofs program, under which PV system subsidies were around 50% of the cost of a system for 3-4 kW residential systems. The market exploded, and Japanese prices for PV have been reduced by more than half during the decade following (Figure 3-14). The subsidies have gradually been phased out. They were at 20% in 2004 (BCSE,
59 Chapter 3. Learning in the International PV Manufacturing Industry
2004) and have now been reduced to zero, with a renewable energy quota currently implemented. The growth of the Japanese PV manufacturers, however, has not been dampened. Immense confidence due to technological leadership, the development of a strong network of manufacturers and sales people and the expectation that MITI will intervene if the growth in the PV market decreases significantly have buoyed the industry (Jäger-Waldau, 2006).
Figure 3-14: Japanese Rooftop Program Experience, 1994-2000
18 120 16 100 14
12 80
10 Figure has been removed due to copyright restrictions. 60 8
6 40
PV System Price (1998$/Wp) 4
20 PV Rooftop Systems Installed (MW) 2
0 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 Year
Source: (Margolis, 2003) Much of the recent expansion in Japanese production has been exported to growing markets in other countries. In 1995, Japanese companies exported 528MW of production, with 386.8 MW being exported to Europe (Jäger-Waldau, 2006). The Japanese PV roadmap 2030 aims for “securement and maintenance of global competitiveness based on technological advantages” (NEDO, 2004). The goal of the roadmap is to achieve mass introduction of PV systems, specifically through mass production and reduced costs in manufacturing (Goto et al., 2004), expansion of markets, enhancement of industrial infrastructure and securing of raw material supply (NEDO, 2004). NEDO’s R&D sponsorship is selective, only supporting firms that have shown the best results (Ristau, 2001). NEDO has sponsored 100% of R&D of these firms and institutions, including the development of new technologies to pilot production (Balaguer & Marinova, 2006). Japan has been successful in building the largest PV manufacturing industry worldwide, and now has a sustainable competitive lead in the industry. Watanabe (Watanabe et al., 2000) has created an technological system model for the case of Japan, illustrating the virtuous circle of investments in R&D, cost reductions and increases in market size that have occurred in this case (Figure 3-15).
60 Chapter 3. Learning in the International PV Manufacturing Industry
Figure 3-15: A 'virtuous cycle' for PV development in Japan 1976-1995
Figure has been removed due to copyright restrictions.
Source: (Watanabe et al., 2003)
The R&D initiated by the Sunshine projects has encouraged an increase in private R&D investment, building up the technological knowledge stock. The increase in manufacturing knowledge and decreased cost of PV systems have combined with market formation programmes to promote market growth, further increasing solar cell production and encouraging further investment in R&D. The participation of large numbers of manufacturers and research institutes participating in photovoltaics R&D in Japan have created a particularly diverse knowledge base and a competitive environment (Balaguer & Marinova, 2006). There are now seven Japanese cell manufacturers, including Sharp, Kyocera, Mitsubishi Electric and Sanyo in the top 10 producers by volume. There are also four silicon producers (Tokuyama, Mitsubishi Materials, Sumitomo Titanium and JFE Steel), a few module-only manufacturers, including MSK, and inverter and glass producers. Other companies are researching and running pilot lines (Jäger-Waldau, 2006). Suppliers of materials for making solar cells: silicon feedstock, ingots, wafers, metal pastes; modules: encapsulation materials, glass, metal frames; as well as equipment manufacturers have also emerged (Ikki et al., 2004). Japanese renewable energy policy is notable for its market orientation and its long term timeframes, setting 15 year goals in 1994 (Jäger-Waldau, 2005). The Japanese government has driven the legitimisation process for PV, recognising its strategic importance, particularly in relation to energy security.
3.2.3. The Case of the US The US has a long and distinguished record of involvement in the PV industry. The application of PV in space programs began in the 1950s (Norberg-Bohm, 2000); and for
61 Chapter 3. Learning in the International PV Manufacturing Industry
terrestrial applications, particularly remote power supply, in the 1970s. PV R&D was very strongly supported by the national government Department of Energy (DOE) in response to the oil crisis in the late 1970s and early 1980s (IEA, 2006a). The Solar Energy Research Institute, which in 1991 became NREL, a national laboratory, was established in 1977. The first R&D program was the Flat-Plate Solar Array project from 1975-1985. By 1983, ARCO Solar and Solarex had MW-scale manufacturing plants and much of the production was sold to government funded demonstration projects (Surek, 2003). By 1983, however, prices for oil had dropped and US government funding for PV R&D also collapsed from over 300 million in 1981 to less than 100 million in 1983 (IEA, 2006a). After 1983, R&D funding was more focused on high-risk, potentially breakthrough technology, as there was less urgency and less optimism about using PV for utility-scale generation (Surek, 2003). The US has invested the most heavily of any country in PV R&D and has produced many of the most important technological breakthroughs. Thin film and multi-junction concentrator cell technologies emerged from US PV research programs (Surek, 2003). NREL holds the laboratory efficiency records for Cadmium Telluride and Copper Indium Gallium Selenide cells, while Spectrolab holds the record for multi-junction concentrators (35.2%). Incremental progress and improvements in module reliability and manufacturing cost reductions have also emerged from US R&D programs (Witt et al., 2001). Research institutions are traditionally focused on the technological frontier, and are less likely to promote incremental improvements that lead to cost reductions in mature technologies. In the US, however, the PVMaT program, which began in 1990 has explicitly targeted manufacturing cost reductions, and involved government funding for industry through competitive processes. The program aimed to (BCSE, 2004): improve PV manufacturing processes and equipment; accelerate manufacturing cost reductions for PV modules, balance of systems components, and integrated systems; increase commercial product performance and reliability; and enhance the investment opportunities for substantially scaling up US manufacturing capacity and increasing US market share. The PVMaT project has included up to 50% public funding for some projects, and publicly funded research is transferred to industry early in the development cycle, and with minimal royalties (Watt, 2003). Mobil Solar (now RWE Schott Solar), AstroPower (now GE Solar) and MIT (now Evergreen Solar) benefited from this funding for EFG, Silicon Film and String Ribbon technologies (Surek, 2003). The Thin-Film Partnership and the PV Manufacturing Program (an extension of the PVMaT program) have continued to facilitate inter-firm cooperation and collaborations between firms and research organisations (Norberg- Bohm, 2000).
62 Chapter 3. Learning in the International PV Manufacturing Industry
The US was the largest manufacturer of PV cells worldwide until 1995, when Japan overtook it (Mints, 2006b). Notwithstanding large and sustained investments in R&D, US production market share has decreased, due to slow domestic market growth (Shum & Watanabe, 2006). There has been little support for market development in the US, and what support there has been has been inconsistent and often state-based, not coordinated at the federal level. There has also been failure to invest in national incentives for manufacturers to locate in the US (Mints, 2006b), and state-based incentives have not been sufficient to encourage a high level of investment in manufacturing in the absence of strong, consistent markets. The US market is still the third largest worldwide, according to 2005 figures, but is made up of regional markets where states and utilities subsidise purchases, and niche markets (Jäger-Waldau, 2006). There are also renewable energy quota programs in various states (Hoffmann, 2006). The most ambitious of the state programs is the California “million roofs initiative”, announced in August 2006, which includes a 10 year rebate program. Hawaii,
Illinois, New York and North Carolina also have schemes that make PV cost effective at $7/Wp. A further 10 states and Washington DC have initiatives that make PV cost effective between
$4.50 and $7/Wp (Jäger-Waldau, 2006). At the federal level, in 2005, the US increased the 10% business energy tax credit for solar to 30% for a period of two years and extended it to residential systems. There are some promising signs regarding the growth of the PV technological system in the US. Despite previous failures of firms in commercialising thin-film technologies, due to manufacturing, reliability and marketing problems (Surek, 2003), firms promoting thin-film solar are now attracting venture capital. In 2005, thin-film startups, companies such as Miasolé, Nanosolar, and HelioVolt successfully raised venture funding to enter thin-film commercialisation (EIA, 2006). There were 12 manufacturers of PV cells in the US in 2005 (EIA, 2006). Japanese companies Kyocera, Sanyo and Sharp now also have subsidiaries or module production plants in the US. Production at Shell and GE decreased in 2005, but Evergreen, First Solar and United Solar increased production significantly (Jäger-Waldau, 2006). Norberg-Bohm (2000) identifies the strengths of the US solar R&D program: Strengths: Parallel path strategy (many research strands), Collaborations between industries, universities & national labs, including public- private partnerships with cost sharing, Attention to full range of R&D (basic scientific – manufacturing). Weaknesses: Lack of consistency of funding, Manufacturing R&D not started soon enough.
63 Chapter 3. Learning in the International PV Manufacturing Industry
3.2.4. The Case of Australia In the 1970s and 1980s, Australia was a leader in the use of photovoltaics for remote industrial and solar home system applications. Telecom Australia and Australian National Railways installed large PV repeater and signalling systems, and in the 1980s, the market for off-grid hybrid systems in Australia expanded (Watt, 2003). The market for remote PV in Australia has remained steady, but while other countries have rapidly expanded their markets for grid connected PV over the past decade, Australia’s grid connected market has remained a fraction of the size. Although there has been no overall strategy for the support of PV technology development, the government has funded basic R&D at a reasonably high level. There are two major research centres for PV in Australia. The University of New South Wales (UNSW) began research on photovoltaics in 1975, has led the world in efficiencies for laboratory crystalline silicon solar cells for over a decade and invented the successfully commercialised high efficiency Laser Grooved Buried Grid (LGBG) cell (UNSW, 2006). The Australian National University (ANU) began PV research in 1991 (Watt, 2003), and is commercialising their Sliver Cell technology. There are a four other universities with small groups of 3-4 people conducting PV research. UNSW has been continuously funded at a high level via its status as an Australian Research Council (ARC) research centre. ANU has also received support from the ARC in the form of discovery and linkage grants, but must regularly prepare grant applications in order to secure continued funding (Watt, 2003). Watt (2003) recommends a more co-ordinated approach to the establishment of research facilities, so that these facilities do not compete directly with each other for scarce funds, allowing for more diversity in research. BP Solar is currently the only commercial scale manufacturer of PV cells in Australia. In 1979, Tideland Energy built a PV manufacturing plant in Sydney (BCSE, 2004)and licensed the LGBG technology from UNSW in 1985, but was bought by BP Solar later that year (Balaguer & Marinova, 2006). Solarex also established a manufacturing plant in Sydney in the 1980s, but merged with BP Solar in 1998. BP Solar expanded manufacturing capacity to 40MWp per year in 2001 (BCSE, 2004). Pacific Solar was founded in 1995 to commercialise crystalline silicon on glass technology developed at UNSW, with investment from Pacific Power, a state utility and later Eurosolare, a European PV firm. Pilot line production was established, but changes to government funding in 2003 forced the company to look for additional injections of capital (BCSE, 2004). The company was unable to secure finance for commercialisation within Australia, and announced plans for commercial production in Germany in 2004 after raising capital through an equity arrangement with Q-cells and with German government support. There are two additional PV firms in pilot line production. In 2004 Origin Energy started pilot line production of an innovative silicon ‘Sliver’ technology, licensed from ANU (Watt et al., 64 Chapter 3. Learning in the International PV Manufacturing Industry
2006). Dyesol, part of an international group, carries out research and development into dye solar cells and has a pilot line. Watt (2003) identified a gap between funding for basic research funded by the ARC, and industry funded commercialisation. Some support for the commercialisation of PV technology is available through the Renewable Energy Equity Fund (REEF), which provides AU$26.5 million in venture capital for renewable energy technologies. Of this, AU$18 million is provided by the Australian Government and $8.5 million is provided by private sources (BCSE, 2004). The amount that may be obtained for one project, however, is limited to AU$3 million (Watt, 2003), much less than the investment required to build a PV plant. There are also competitive investment grant programs in NSW and Victoria and further support for remote PV systems. In an environment of small and uncertain domestic markets and scarce public funding available to commercialise fledgling technologies developed within Australian research groups, it has been difficult to obtain private investment to commercialise technology and update facilities (BCSE, 2004). Local research institutions are exporting their intellectual property to countries where investment capital is available and future markets are more promising. New government funding for commercialisation activities was announced in the Energy White Paper but a gap remains in pre-commercialisation technology development funding. There are some IP issues arising from the dependence of Australian PV R&D on government funding, since competitive grant processes must be transparent and therefore may expose otherwise confidential research. Negotiations over IP from universities, where nearly all of the research takes place, is also becoming costly and time consuming, since Australian universities are progressively more self-funding (Watt, 2003). Renewable energy market development programs in Australia include the Green Power scheme and Mandatory Renewable Energy Targets (MRET), introduced in 2000 (AGO, 2007). Under the Green Power scheme, utilities offer green products via higher tariffs. These products include a selected percentage of energy from renewable sources and retailers must produce or purchase the equivalent amount of renewable energy based electricity. A few large grid- connected demonstration systems have been supported by Green Power, including a 400 kW grid connected system at Singleton, a 199 kW BIPV system on a retail market in Melbourne, 50kW systems in Dubbo and Queanbeyan, 640 x 1 kW grid connected home systems as part of the Sydney Olympic village as well as larger grid-connected systems on the Olympic facilities (Watt et al., 2006). MRET is a GWh quota scheme. There is no differentiation between renewable technologies, so PV, being more expensive is commonly overlooked. The target is not ambitious enough to achieve significant production from renewables and has not been increased despite calls to do so from the renewable industry (Balaguer & Marinova, 2006). There are two national market development schemes specifically aimed at photovoltaics. Under the Photovoltaic Rebate Program (PVRP), operating from 2000, purchases
65 Chapter 3. Learning in the International PV Manufacturing Industry
of PV systems up to 1 kW are subsidised. The subsidies were reduced from $5.50 per Wp to 4 per Wp after 6 months due to oversubscription but have recently been increased to $8/Wp. Under the Renewable Remote Power Generation Program (RRPGP), grants of up to 50% of the purchase price of renewables are available for diesel replacement in remote areas (BCSE, 2004). In general, the grid-connected market development incentives to date have been too short term, too uncertain and too small to establish a self-sustaining market and supply industry. It remains to be seen whether the new extended programs will result in increased industry investment for the long term. Australia was a world leader in PV technology development, manufacturing and market development in the 1970s and 1980s. In Australia, cell manufacturing has not expanded in recent years, despite hosting leading research and educational facilities. The industry has suffered from intermittent and insufficient political support for market formation, leading to small and uncertain markets; inconsistent, uncertain and often insufficient support for R&D that has been concentrated in two institutions; and almost no support for commercialisation. These policies, Australia’s support for the fossil fuel industries and failure to sign the Kyoto protocol have left PV as a fringe industry. A mass of manufacturers and other industry actors are required to entrench favourable institutions and legitimise the technology.
3.2.5. Factors Influencing the Success of PV Industries in Different Countries Having described the experiences of Germany, Japan, the US and Australia in fostering a PV industry, this section identifies factors that have influenced their relative success.
3.2.5.1. Comparing Learning
Learning in PV manufacture may be measured by reductions in the $/Wp cost for solar cells, which have been primarily achieved through manufacturing cost reductions, material cost reductions, efficiency improvements or manufacturing scale increases. Module price does not necessarily follow cell manufacturing costs, since module encapsulation costs and marketing and distribution costs also contribute to module prices. There are other factors affecting price, such as profit margins, subsidies and tax concessions for manufacturers in different locations and direct funding for manufacturers, which have been available for firms locating in Japan and Germany. There are also likely to be spillovers between national technological systems, so price reductions cannot automatically be attributed to learning within the national technological system. However, in the absence of manufacturer’s cost data, module prices may be the best proxy available to compare the overall performance of technological systems in achieving cost reductions.
66 Chapter 3. Learning in the International PV Manufacturing Industry
Module prices decreased between 1992 and 2004 by 50% in Germany4, 55% in Japan, 15% in the US and have not changed appreciably in Australia during that time (Figure 3-16). There is a fairly clear indication that learning has been more effective in Germany and Japan than in the US and Australia. It should be noted that in the last few years module price in Germany have increased, due to huge market growth and module shortages, which have enabled producers to take profits, and is not related to learning.
Figure 3-16: Indicative Module Prices in National Currencies per Watt in Australia (AUS), Germany (DEU), Japan (JPN) and the US (USA) Country 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 EUR USD AUS - - 7 - 8 - - 8 8 8 7 7 8 8 4.9 6.1 4,0 2,5 - 3,0 - - 4,0 - 4,9 - DEU 5.98 5.93 5.42 4.91 4.5 4.14 3.73 3.63 3.58 3.53 3.04 9,7 9,6 6,0 6,0 7,4 JPN 966 950 927 764 646 656 670 600 548 484 463 446 439 431 3.2 3.9 USA 4.25 4.25 4 3.75 4 4.15 4 3.5 3.75 3.5 3.25 3 3.5 3.6 2.9 3.6 Source: (IEA PVPS, 2006)
8
7
6
5 AUS DEU 4 JPN USA 3
Price normalisedPrice to 2005 Euros 2
1
0 1990 1992 1994 1996 1998 2000 2002 2004 2006
The following sections seek to explain why the German and Japanese technological systems have been more successful in building technological capabilities within PV manufacturing firms.
3.2.5.2. Comparing Domestic Markets and Production According to IEA PVPS statistics (IEA PVPS, 2006), in 1992, the cumulative PV installations in Australia amounted to 7.3 MWp, in Japan 19 MWp, Germany 5.6 MWp and the
4 The 2002 price is used for Germany, since a range is quoted from 2003 onwards, the highest figure in the range being for high-cost, custom modules, such as those for BIPV applications. 67 Chapter 3. Learning in the International PV Manufacturing Industry
USA 43.5 MWp. Between 1992 and 2004, Australia and the USA have respectively increased cumulative PV installed by 6.2 times (to 52.3 MWp) and 7.4 times (to 365.2 MWp). Germany and Japan have respectively increased cumulative PV installed 141 times (to 794.0 MWp) and 59 times (to 1132.0 MWp). Figure 3-17 compares the annual cell production with the annual PV sales in the four countries. It is clear that market size is strongly correlated with and provides a strong incentive for investment in production. In countries such as the USA and Australia which have not been able to grow local markets strongly and consistently, the development of the cell manufacturing industry has been stifled, even while there has been a large stock of technological knowledge. In the US, markets have slowly but consistently grown, whereas investment in new manufacturing has slowed. In Australia, incremental manufacturing increases have continued despite very slow market growth.
Figure 3-17: Cell Production and Market Size (MW/year) 1995-2005 in Australia (AUS), Germany (DEU), Japan (JPN) and the US (USA)5
800
700
600
AUS Market 500 DEU Market JPN Market USA Market 400 MW AUS Cell Prod DEU Cell Prod 300 JPN Cell Prod USA Cell Prod
200
100
0 1994 1996 1998 2000 2002 2004 2006
Sources: (IEA PVPS, 2006, 2007a, b; Maycock, 2006; Mints, 2006a; Schmela, 2005, 2006; Stryi-Hipp, 2006)
It is clear that a strong and consistent domestic market, as well as government commitment to the industry signalled by financial incentives for manufacturers provides a long- term planning horizon for firms, encouraging investment in manufacturing and R&D.
5 Since different sources have been used to compile the data, there may be variations in the data collection methods and therefore the data may not be entirely equivalent, but the general trend can still be relied upon. 68 Chapter 3. Learning in the International PV Manufacturing Industry
3.2.5.3. The Role of Government Market Support The role of market growth in promoting manufacturing cost reductions in cell manufacture is questioned. Menanteau (2000) considers that deployment policies tend to favour the incumbent technology, and do not contribute significantly to innovation. Schaeffer et al. (2004) believe that market support programs favour downstream learning, such as the development of knowledge related to BOS and system integration, rather than cost reductions in module manufacture, which are more dependent on R&D investments. Jacobsson et al. (2002), however, believe that market formation programs, when successful, build confidence in the future of the technology and encourage the entry of actors and investment in R&D, including in non-dominant technologies. Some markets will be more effective at inducing firms to invest in innovation. The German feed in tariff and the Japanese subsidy models do not limit the size of the market as do quota and competitive tender approaches, and may therefore provide more incentive for manufacturers to make the large and long-term investments required to increase production, providing opportunities for scale economies and learning by doing. The prospect of long term markets is also more likely to encourage investment in R&D. Technology specific subsidies have been criticised for failing to encourage cost reductions. Policy mechanisms such as renewable energy quotas and competitive tenders that force suppliers to be competitive with other technologies on cost are expected to promote innovation more effectively than other measures (Anderson et al., 2006). Since PV is at an early stage of its development, however, it is likely to lose out to other technologies and fail to attract investment without some differentiation in support for technologies that are currently more expensive. Some argue that the promise of higher profits available in subsidy or tariff systems is an equally strong inducement to innovate as is cost pressure. In Germany, a combination of strong market growth and limited supplies of silicon feedstock have meant that demand has outstripped supply, and manufacturers have had little incentive to reduce prices. However, given the expectation that the silicon shortage will come to an end within a few years, and the growing production capacity worldwide, it is unlikely that German manufacturers have ceased investing in cost reduction. Because the market for PV modules is an international market (for example, manufacturers from all the producing countries sell modules in the German market); a protected domestic market situation is unlikely to reduce spending on R&D.
Support for Local Manufacture Hoffmann (2004b) believes that although a local market is a necessary condition for a production industry to become established, the existence of a local market does not guarantee growth of a local industry. This has been found to be true in the case of wind turbine manufacturing (Lewis & Wiser, 2005), and in the German PV industry of the early 1990s. There
69 Chapter 3. Learning in the International PV Manufacturing Industry
are differences in the extent to which market support measures in the different countries have provided opportunities for local manufacture. In general measures which do not have long term political certainty, such as those employed in the US and Australia are less likely to encourage local investment. Short-term market opportunities will instead be taken up by existing firms, often importers. Some market development programs explicitly favour locally-owned firms. Local content rules, for example, can mandate a certain percentage of locally manufactured content in PV projects. This may cover all PV projects, government procurements, or only particular projects. Firms that wish to sell their product in such a market must transfer some of their manufacturing to the country. This has happened in the case of Germany. Unless the rule is specifically targeted at locally owned companies, it will equally encourage investment in local manufacturing by international and domestic firms. If local content requirements are very high, the technological system may not be able to successfully learn to supply the product and poor quality products or higher prices may result. Local PV content rules in Thailand have had this effect.
Clusters The entry of more actors in response to growing markets can increase the opportunities for learning. Networks between firms, customers and suppliers in close proximity (clusters) have intensified interactions which enhance learning benefits, and are considered to be of substantial importance. The centralisation of the PV industry in a few countries including Japan, Germany, the US and more recently China provides opportunities for a high level of interactions in the domestic industries of these countries and the emergence of cluster effects. The existence of many PV manufacturers and researchers in Germany and Japan has been found to create a high level of diversity and a competitive environment, leading to accumulated experience and market leadership (Balaguer & Marinova, 2006). The US manufacturers ASE Americas, Evergreen Solar and Ascension Solar, for example are located in Boston, close to Massachusetts Institute of Technology. In Germany, many firms in the PV value chain (including equipment manufacturers) and the Fraunhofer Institute for Solar Energy are clustered around the central-east of the country where capital investment subsidies are provided. These clusters increase the potential for profitability, since transactions may be realised more quickly and at lower cost, while R&D cooperation and learning opportunities are also increased.
3.2.5.4. Other Incentives Other financial inducements to locate manufacturing in a particular place have included direct assistance for new capacity, tax concessions, R&D funding, finance for commercialisation, export assistance, import tariffs and local content rules. In East German
70 Chapter 3. Learning in the International PV Manufacturing Industry
economic development zones, the state governments offer investment incentives of up to 50% of capital expenditure for new manufacturing facilities. The existence of strong advocacy for PV is also likely to ensure the continuity of political support for the technology and hence funding.
3.2.5.5. Comparing Government R&D Spending During the oil crisis of the 1970s, the US led the way in photovoltaics R&D, spending up to 300 million US dollars per year, according to the IEA R&D database (IEA, 2006a). The relevant data from the database is displayed in Figure 3-18 Since the late 1980s, the US has spent between US$50 and 100 million per year. Germany spent close to US$100 million in some years in the late 1970s and early 1980s, before settling to funding closer to US$50 million annually in the late 1980s. Japan’s R&D funding has increased since the late 1990s to become the most substantial worldwide. While the R&D spending of Japan and Germany have increased, that of the US still remains high. In general public spending in IEA countries on PV has transitioned from an R&D focus to a market development focus over the last decade (IEA, 2006b), as illustrated in Table 3-4, most significantly in Germany.
Table 3-4: Public R&D, Demonstration and Market Development Spending in Australia, Germany, Japan and the US, 2005
Country R&D Demonstration Market Total AUS 3.1 0 11.9 15 DEU 30.3 0 308.6 338.9 JPN (METI) 60.5 102.8 48.6 211.9 USA 86 10.5 180 276.5 Source: (IEA PVPS, 2006)
3.2.5.6. The Role of Government R&D Support Many PV firms have relied heavily on the knowledge stock held by publicly funded research institutes or universities. Support for R&D is likely to be of great importance to cell manufacturers, due to the complex nature of the knowledge, and the inability of users to judge and give feedback on the performance of the product.
Encouraging Innovation R&D programs influence the learning rate by directly increasing innovative effort. They also have the potential to promote private investment, influence the direction of search, and support collaborations between research organisations and industry. Shum & Watanabe (2006) demonstrate via regression analysis of Japanese expenditure (including R&D and production investment) against PV costs that cost reductions have been more sensitive to expenditure (a 54% drop for doubling of expenditure) than production volume (a 36% drop for doubling of production), suggesting that R&D and learning by doing both make an important contribution to learning.
71 Chapter 3. Learning in the International PV Manufacturing Industry
The number of patent applications by a firm (or country) is expected to reflect the innovative activity. Watanabe et al. (2000) have found that R&D expenditure is the most significant factor influencing the number of patent applications in PV in Japan, followed by technology knowledge stock, but that knowledge assimilated from external sources also has a significant impact. Andersson & Jacobsson (2000) also find correlation between the peak OECD government R&D expenditure in photovoltaics in 1980, and the peak number of patents registered at the USPTO (Figure 3-18), which lags by a few years. The expenditure was 400 million US dollar in 1980 and 250 million in 1995.
Figure 3-18: R&D Funding and Number of USPTO Patents Granted 1974-2005
350 120
300 100
250 80
200 60 150
40 R&D Budget $US Millions
100 Patents USPTO of Number
20 50
0 0
6 6 6 74 77 80 81 83 84 87 88 90 91 93 94 95 97 98 00 01 03 04 05 19 1975197 19 1978197919 19 198219 19 1985198 19 19 198919 19 199219 19 19 199 19 19 199920 20 200220 20 20
R&D Budget Germany R&D Budget Japan R&D Budget United States Patents Germany Patents Japan Patents United States
Source: (Andersson & Jacobsson, 2000; IEA, 2006a)
It was previously noted that private Japanese investment in R&D followed government R&D expenditure. Surek (2003) notes that private investments in the US have also mirrored those by the government, especially when government support rapidly increased or decreased. Public R&D expenditure is clearly an incentive to invest in innovation, increases the technological knowledge stock within the industry, and access to this public R&D typically favours locally owned over foreign firms.
Influencing the Direction for Search R&D support tends to influence the direction of search, since funding, even outside of public research institutes, is usually allocated according to a national strategy, or roadmap, for the technology. R&D funding has at times been focused on particular technologies that were expected to emerge as the lowest cost per Wp design. For example, the commercialisation of a-
72 Chapter 3. Learning in the International PV Manufacturing Industry
Si solar cells by Sanyo in 1980 for consumer electronics led to an interest in low cost a-Si modules as an alternative to c-Si (Jacobsson et al., 2002). The US, Japan and the EU subsequently focused on a-Si thin film R&D and manufacturing (Andersson & Jacobsson, 2000; Jacobsson et al., 2002). Jacobsson et al.(2002) believe that some German enterprises were ‘guided in their search process’ towards a-Si technology by the German R&D programmes. Despite significant experience in production, the promise of a-Si technology has not been realised and the cost of production ($/W) is still higher than the dominant wafer-based technologies. Since the 1990s, funding has been spread among a variety of technologies, as it has been historically difficult to predict the direction that technological breakthroughs will lead the industry. In Germany, four thin film technologies have received equal funding (Jacobsson et al., 2002). Historically, R&D funding has enabled knowledge creation in many alternative designs, as German, US and Japanese programs have, to varying degrees over time, allocated funds to non-dominant technologies. Government demonstration projects may also influence the direction of search. Investments in R&D for BIPV have been strong in Germany, where there have been a number of BIPV demonstrations, in addition to enhanced tariff incentives.
Encouraging Firm-Research Interactions Governments have encouraged innovative interactions by funding collaborative research. University-industry network formation has been encouraged by funding tied to these collaborations, as well as innovative arrangements for sharing public funding, facilities, and intellectual property in Germany, Japan and the US (Balaguer & Marinova, 2006). Japan intentionally encouraged inter-firm collaboration through the Sunshine programmes. The US also encouraged inter-firm and firm-research organisation collaborations. Germany also specifically encouraged university-firm research collaborations. In Australia, conversely, the commercialisation of technology has occurred mainly through licensing, since the government funding has been focused on university research. Commercialisation has been hampered by high transaction costs in IP transfers, concerns over the transparency of government funded research, and lack of funding for the development and commercialisation phases. Technology transfer from research to industry, which is one of the most difficult stages in the innovation cycle has been held back.
Encouraging Spillovers Watanabe et al. (2003) suggest that photovoltaics is an interdisciplinary technology, and as such, the importance of and benefits of inter-industry technology spillovers are substantial. In particular, photovoltaics manufacture is closely related to microelectronics semiconductor technology. The choice of silicon as the dominant material for the solar cell industry was
73 Chapter 3. Learning in the International PV Manufacturing Industry
heavily influenced by the potential to use by-products and the existing knowledge about silicon existent in the electronics industry (Menanteau, 2000). Much of the equipment, materials and many of the techniques used for standard silicon solar cell manufacturing are similar to those used in the microelectronics industry (Green, 2000). As the PV industry matures, however, the production processes for photovoltaics and for microelectronics often diverge. Reductions in the cost of silicon material have historically been due to progress in silicon production within the microelectronics industry, but with the advent of solar grade silicon producers, the PV industry is expanding vertically to produce its own material, and its own learning associated with this production. Although PV manufacturers initially used photolithography methods borrowed from the microelectronics industry, microelectronics manufacturers require increasingly fine lines, whereas photovoltaic manufacturers have moved from photolithography to screen printing, a less precise but cheaper method of metallization (Linden et al., 1977). Spillovers from the microelectronics industry have been an important source of learning, but as the photovoltaics industry matures, more technologies that are specific to the industry are being developed. Since the semiconductor industry requires high-purity silicon wafers, it has no interest in developing lower grade silicon, thin film materials, or other novel technologies which are of interest to the PV industry. If the photovoltaics industry, as is expected, moves towards thin-film devices, the similarities with the microelectronics industry will be reduced, and the technological spillovers will be less direct. Progress in other industries is also of relevance to the PV industry and spillovers may be obtained. For example, conductive metal pastes are used for car rear windows. Lamination materials used in glass have formed the basis of the development of materials for PV module encapsulation. Japan has supported the involvement of a wide range of industries in PV R&D, encouraging the spillover benefits that may be obtained through the entry of these actors. Balaguer & Marinova (2006) believe that the German PV industry has also benefited from the existence of mass production expertise in a variety of other industries. Funding has been used to intentionally bring a diverse set of industries into the technological system in the case of Japan and in the early days of the US industry, increasing opportunities for spillovers and for accessing the supply chains of existing industries. Japan’s PV industry has largely been built through diversification from a strong electronics industry. In both Japan and Germany, downstream interactions have led to learning and have enabled firms to access the marketing channels of exiting industries. Some Japanese PV firms have either purchased housing or construction companies or collaborated with existing firms to access the grid connected solar home system market. German firms have invested heavily in BIPV R&D and commercialisation. The historical regulation of foreign direct investment in Japan, the tendency
74 Chapter 3. Learning in the International PV Manufacturing Industry
towards lifetime employment in large Japanese firms, the high cost of doing business in Japan (Paprzycki, 2004) as well as the selection of firms for R&D support has created a somewhat oligopolistic market. The strong industrial base in Germany has enabled firms to obtain the capabilities for production in all parts of the PV manufacturing value chain, from silicon feedstock and equipment manufacture to module assembly (Balaguer & Marinova, 2006). Japan has had less success in developing equipment manufacture, and their cell and module manufacture is concentrated within a small number of large firms.
3.2.6. Virtuous Circles of Policy Induced Development Many of the leading PV firms based their initial success upon national R&D programmes in the 1970s and 1980s. These firms benefit from a first-mover advantage. The experiences of Germany and Japan have shown, however, that stable and supportive programs for PV have created strong markets and provided an environment for the emergence of successful local manufacturers, while the early dominance of the US firms has waned. Germany has been able to catch up by providing strong incentives for manufacturers to locate domestically. In Germany and Japan the market expansion facilitated learning by doing in manufacture, investment in upstream and downstream parts of the value chain and learning by interacting. A ‘virtuous circle’ of learning was observed, as market growth encouraged learning investments independently of publicly funded R&D, further bringing down costs and expanding markets. New entrants were also encouraged by the market growth, bringing further knowledge, resources and links to new market segments, such as BIPV. As the number of entrants has grown, the legitimisation of the technology and political pressure has further increased. The actors and their networks, and the virtuous economic circles, work for the survival of the technological system, and the government support it depends on. Learning occurs not only in manufacturing, but also by strengthening the supporting infrastructure, networks and institutions. More information about the technology is disseminated and the technology is legitimised. Standards and testing facilities establish the quality and reliability of the product and build consumer confidence, since customers can differentiate between good and poor quality products. Customers also benefit from increasing returns to adoption, such as better system maintenance as more people use PV, potentially increasing the price people may be willing to pay. Banks in Germany, for example now readily offer finance for PV investments, since the return on the investment is predictable. In both Germany and Japan, the legitimisation of PV technology has encouraged investment in production. In Germany, legitimisation occurred through political pressure from
75 Chapter 3. Learning in the International PV Manufacturing Industry
green groups and community-wide support for renewable energy deployment, whereas in Japan, the government emphasised the strategic importance of the industry, giving the industry a “key industry” status. In both cases, support programs have been long term, allowing firms to plan investments. In Japan the support has also been continuous. The formation of standards and the diffusion of technical information about photovoltaics applications were promoted in Germany by industry and political lobbying organisations. These organisations have also strongly influenced political support and the design of programs for the technology, with some of the organisations, such as Eurosolar having their members embedded in the political structure (Jacobsson et al., 2002). Although a self sustaining market was not created by the early market incentives in Germany, green groups that supported the program gained strength and eventually larger market programs were implemented, such as the 100 000 roofs program in 1998 (Balaguer & Marinova, 2006). In the US, despite large R&D investments and one of the largest markets in the world for many years, the technology has not been accepted politically as a viable future option. In Australia, where one of the world’s leading research organisations is based, the same applies. Neither of these countries has been able to build the confidence of manufacturers to invest heavily in the past decade, since market support has been piecemeal, inadequate, and uncertain, and without an observable political commitment to long term renewable energy development. The low cost of electricity from fossil fuels in Australia and most US states, and the strong political support for economically powerful fossil fuel industries has also hampered the perception of the renewable energy industries. The retail price for electricity in Japan is three times Australia’s, while Germany’s is twice Australia’s. Lack of long term commitment has sent negative signals to the industry and financiers, and the public conversation about energy in Australia and the US has been one of climate change denial and emphasis on the importance of keeping energy costs low. There is little awareness or information about the technology and sales and service networks are underdeveloped. The lobby in these countries is small and weak relative to the fossil fuel lobby, since there are only a few relatively small actors.
76 Chapter 3. Learning in the International PV Manufacturing Industry
3.3. Conclusion
Nagamatsu et al.(2006) regard photovoltaics (PV) as a ‘footloose’ technology, since it does not rely not on national advantages in factors of production, but on advantages that are not location-specific, such as technological learning and economies of scale. The technological capabilities of PV firms, the appropriateness of their technology strategies and the ability of the technological system to provide them with the resources, incentives and interactions they need will largely determine their success. This chapter has reviewed the literature on learning in the PV industry and the experiences of different countries in establishing a successful PV industry. Learning in the PV industry is characterised by a fast learning rate. Firms will need to continually improve and innovate in order to remain competitive. Technological improvements that have been and are expected to continue to be important sources of cost reductions in the industry have been established and comprise: Scale increases Efficiency improvements Material cost reductions Manufacturing cost reductions
Even if a local PV manufacturing industry is established, there are variations in the extent to which a vibrant national technological system that fosters learning and expansion throughout the value chain is achieved. Multinational PV firms may set up manufacturing of established technology in a developing country, or at the other end of the spectrum, new technology may be developed through national R&D, within domestic firms or through collaborations. Intermediate is the licensing and improvement of existing technology. Different policies will be needed to encourage movement to higher levels of technology ownership and innovation. The cases of Australia, Japan, Germany and the US revealed that countries have different abilities to support PV manufacturing firms. The presence of long term and predictable domestic markets within a supportive energy policy environment has been the most important factor in the establishment of a domestic cell manufacturing industry. Financial incentives for local manufacture and access to commercialisation of new technology via R&D collaborations have also been important in enabling firms to take advantage of these markets. Public R&D spending has been less important than the presence of local markets in fostering a domestic manufacturing industry, but R&D activities develop the stock of human capital and can also be used to encourage private investment in R&D. The R&D may be of little value to firms without
77 Chapter 3. Learning in the International PV Manufacturing Industry
the right kind of IP arrangements and specific funding for commercialisation. Local downstream value chain interactions have resulted in cost reductions and product improvements in Japan and Germany. Japan has few equipment suppliers for cell manufacture, whereas Germany has many, but from the literature, it is not possible to tell whether Japanese cell manufacturers have been constrained by the lack of local interactions. It is possible that spillovers from related industries in Japan and the US have provided related expertise that has compensated for the lack of interactive learning with suppliers. The literature reviewed provides an industry wide view of the learning rate and sources of cost reductions, and indicates some of the national-level factors that influence the success of PV firms within countries, but does not explain how learning occurs within firms and does not provide any information about the particular case of learning in PV firms in developing countries. The following chapter establishes a role for small scale PV manufacture in developing countries and demonstrates that it has also been difficult to achieve sufficient technological capabilities in that context. The remainder of the thesis uses a case study approach to contribute to the knowledge about suitable capability building strategies, environments and interventions for PV manufacturers in developing countries.
78 Chapter 3. Learning in the International PV Manufacturing Industry
References
AGO (2007), Mandatory Renewable Energy Target, Australian Greenhouse Office, Accessed from: http://www.greenhouse.gov.au/markets/mret/, on: March 2007. Anderson, J., Bassi, S., Stantcheva, E. and Brink, P.t. (2006), Innvovation Case Study: Photovoltaics, Assessing Innovation Dynamics Induced By Environment Policy, Institute for European Environmental Policy. Andersson, B.A. and Jacobsson, S. (2000), Monitoring and assessing technology choice: the case of solar cells, Energy Policy, 28 (14), p 1037. Argote, L. and Epple, D. (1990), Learning Curves in Manufacturing, Science, 247 (4945), p 920. Arrow, K. (1962), The economic implications of learning by doing, The Review of Economic Studies, 29 (3), pp 155-173. Balaguer, A. and Marinova, D. (2006), Sectoral Transformation in the Photovoltaics Industry in Australia, Germany and Japan: Contrasting the Co-evolution of Actors, Knowledge, Institutions and Markets, Prometheus, 24 (3). BCSE (2004), The Australian Photovoltaic Industry Roadmap, Australian Business Council for Sustainable Energy, Melbourne, Australia. Bruton, T.M. (2002), Music FM - Five Years On Fantasy or Reality?, 18th European Photovoltaic Solar Energy Conference, Rome, Italy, 7-11 October, 2002. Bruton, T.M. (2005), The Impact of high Volume Production on Cost of Silicon, Solar Cells and Modules, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. de Moor, H., Hagemann, I., Herrero, J., Malbranche, P., Siebentritt, S., Schedel-Niedrig, T., Riedle, T., Poortmans, J., Nijs, J., Szlufcik, J., Jäger-Waldau, A., Karg, F., Teksolar, E.Ö., Pietruszko, S., Haywood, S., Topic, M., Schropp, R., Rath, J., Dimmler, B., Powalla, M. and Springer, J. (2004), PVNET European Roadmap for PV R&D, R&D for PV Products Generating Clean Electricity Contract Reference: ENK6-CT2001- 20400, Jäger-Waldau, A. (ed), European Communities, Luxembourg. Eberhardt, K. (2005), PV Production Facilities adapted to Technology Requirements - a Prerequisite for cost effective Mass Production, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. EIA (2006), Solar Thermal and Photovoltaic Collector Manufacturing Activities 2005, US Energy Information Administration, Washington DC, USA. EPIA (2004), European Photovoltaics Industry Association Roadmap, European Photovoltaics Industry Association. Foster, R. (2005), Japan Photovoltaics Market Overview, Sandia National Laboratories, U.S. Department of Energy. Frantzis, L., Jones, E., Lee, C., Wood, M. and Wormser, P. (2000), Opportunities for Cost Reductions in Photovoltaic Modules, 16th European Photovoltaic Solar Energy Conference, Glasgow, U.K., 1-5 May, 2000. Ghannam, M., Sivoththaman, S., Poortmans, J., Szlufcik, J., Nijs, J., Mertens, R. and Van Overstraeten, R. (1997), Trends in industrial silicon solar cell processes, Solar Energy, 59 (1-3), p 101. Goto, S., Kawakami, K., Nishimura, T., Uda, K., Ishiyama, K. and Aratani, H. (2004), Present Status of Research and Development on Solar Cells in Japan, 19th European Photovoltaic Solar Energy Conference, Paris, France, 7-11 June 2004. Green, M.A. (2000), Photovoltaics: technology overview, Energy Policy, 28 (14), pp 989-998. Green, M.A. (2003), Crystalline and thin-film silicon solar cells: state of the art and future potential, Solar Energy, 74 (3), pp 181-192. Haas, R. (2003), Market deployment strategies for photovoltaics: an international review, Renewable and Sustainable Energy Reviews, 7 (4), pp 271-315.
79 Chapter 3. Learning in the International PV Manufacturing Industry
Haase, J. (2005), High Volume Solar Cell Production Lines - Concepts and Cost Aspects, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. Harmon, C. (2000), Experience Curves of Photovoltaic Technology, Interim Report IR-00-014, International Institute for Applied Systems Analysis, Laxenburg, Austria. Hartley, O. (2006), Driving the PV Industry Towards Competitiveness, 3rd EPIA PV Industry Forum, in the Frame of the 21st EUPVSEC, Dresden, Germany. Hegedus, S.S. and Luque, A. (2003), Status, Trends, Challenges and the Bright Future of Solar Electricity from Photovoltaics, in Luque, A. & Hegedus, S.S. (eds), "Handbook of Photovoltaic Science and Engineering", John Wiley & Sons. Hoffmann, W. (2004a), PV Solar Electricity Industry: Market growth and perspectives, 14th International PVSEC, Bangkok, Thailand, 26-30 January, 2004. Hoffmann, W. (2004b), A Vision for PV Technology up to 2030 and beyond: An industry view, Future Vision for PV, Brussels, 28 September 2004. Hoffmann, W. (2006), PV solar electricity on-grid market: Policy driven next decade, an industry point of view, 3rd EPIA PV Industry Forum, in the Frame of the 21st EUPVSEC, Dresden, Germany. IEA (2006a), IEA Energy Technology R&D Statistics Service, International Energy Agency. IEA (2006b), Trends in photovoltaic applications. Survey report of selected IEA countries between 1992 and 2005, Report IEA-PVPS T1-15:2006, International Energy Agency PVPS Task 1. IEA PVPS (2006), IEA Photovoltaic Power Systems Programme International Statistics, Accessed from: http://www.iea-pvps.org, on: February 2007. IEA PVPS (2007a), Australia Country Information, Accessed from: http://www.iea-pvps.org, on: March 2007. IEA PVPS (2007b), Germany Country Information, Accessed from: http://www.iea-pvps.org, on: March 2007. Ikki, O., Ohigashi, T., Kaizuka, I. and Matsukawa, H. (2004), Overview of PV Activities in Japan: Current status and future prospects, 19th European Photovoltaic Solar Energy Conference, Paris, France, 7-11 June 2004. Jacobsson, S., Andersson, B.A. and Bångens, L. (2002), Transforming the energy system - the evolution of the German technological system for solar cells, SPRU Electronic Working Paper Series Paper No 84, Science and Technology Research, University of Sussex, Brighton, U.K. Jacobsson, S. and Lauber, V. (2006), The politics and policy of energy system transformation-- explaining the German diffusion of renewable energy technology, Energy Policy, 34 (3), pp 256-276. Jäger-Waldau, A. (2004), R&D roadmap for PV, Thin Solid Films, 451-452, p 448. Jäger-Waldau, A. (2005), Research, Solar Cell Production and Market Implementation of Photovoltaics, PV Status Report, European Commission. Jäger-Waldau, A. (2006), Research, Solar Cell Production and Market Implementation of Photovoltaics, PV Status Report, European Commission. Jester, T.L. (2002), Crystalline Silicon Manufacturing Progress, Progress in Photovoltaics: Research and Applications, 10 (2), pp 99-106. Koch, W., Lucas, P. and Viaud, M. (2005), Cooperation Opportunities Between the Semiconductor and the Photovoltaic Industries, 20th European Photovoltaic Solar Energy Conference, Barcelona, Spain, 6-10 June. Lewis, J. and Wiser, R. (2005), Fostering a Renewable Energy Technology Industry: An International Comparison of Wind Industry Policy Support Mechanisms, Environmental Energy Technologies Division, Ernest Orlando Lawrence Berkeley National Laboratory. Linden, L.H., Bottaro, D., Moskowitz, J. and Ocasio, W. (1977), The Solar Photovoltaics Industry: The status and evolution of the technology and the institutions, MIT Energy Laboratory Report - MIT-EL-77-021, US Department of Energy.
80 Chapter 3. Learning in the International PV Manufacturing Industry
Lüdemann, R. (2005), Experience and Expectation of Silicon Solar Cell Mass Production - Requirements for Next Generation Equipment, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. Margolis, R.M. (2003), Photovoltaic Technology Experience Curves and Markets, NCPV and Solar Program Review Meeting, Denver, Colorado, March 24, 2003. Masini, A. and Frankl, P. (2003), Forecasting the diffusion of photovoltaic systems in southern Europe: A learning curve approach, Technological Forecasting and Social Change, 70 (1), p 39. Maycock, P.D. (1997), Cost reduction in PV manufacturing impact on grid-connected and building-integrated markets, Solar Energy Materials and Solar Cells, 47 (1-4), pp 37- 45. Maycock, P.D. (2006), PV cell production data, PV News (March). Menanteau, P. (2000), Learning from Variety and Competition Between Technological Options for Generating Photovoltaic Electricity, Technological Forecasting and Social Change, 63 (1), p 63. Milner, A. (2006), Industry Competitiveness and PV R&D: An Industry point of view, 3rd EPIA PV Industry Forum, in the Frame of the 21st EUPVSEC, Dresden, Germany. Mints, P. (2006a), PV - The story so far, Refocus, November/December 2006, pp 32-36. Mints, P. (2006b), PV in the US: Where is the market going and how will it get there?, Renewable Energy World (September 2006). Nagamatsu, A., Watanabe, C. and Shum, K.L. (2006), Diffusion trajectory of self-propagating innovations interacting with institutions--incorporation of multi-factors learning function to model PV diffusion in Japan, Energy Policy, 34 (4), pp 411-421. NCPV (1999), Report of the Photovoltaic (PV) Industry Roadmap Workshop, NCPV, Chicago, Illinois. NEDO (2004), Overview of “PV Roadmap Toward 2030" (PV2030), New Energy and Industrial Technology Development Organization. Nemet, G.F. (2005), Technical Change in Photovoltaics and the Applicability of the Learning Curve Model, Interim Report 05-029, International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria. Nemet, G.F. (2006a), Beyond the learning curve: factors influencing cost reductions in photovoltaics, Energy Policy, In Press, Corrected Proof. Nemet, G.F. (2006b), How well does Learning-by-doing Explain Cost Reductions in a Carbon- free Energy Technology?, Working Paper: CCMP – Climate Change Modelling and Policy, Fondazione Eni Enrico Mattei, Milano, Italia. Norberg-Bohm, V. (2000), Creating Incentives for Environmentally Enhancing Technological Change: Lessons from 30 Years of U.S. Energy Technology Policy, Technological Forecasting and Social Change, 65, pp 125-148. Paprzycki, R. (2004), What Caused the Recent Surge of FDI into Japan?, Hi-Stat Discussion Paper Series No31, Hitotsubashi University Research Unit for Statistical Analysis in Social Sciences. Photon (2006), Worldwide Market Survey - Solar Modules 2006, Photon International CD- ROM. Podewils, C. (2007), Dreaming of Dinosaurs, Photon International (March 2007), pp 114-123. Ristau, O. (2001), The Photovoltaic Market in Japan: Unquestioned leadership of world market, Solar Reports, The Solarserver. Rohatgi, A. (2003), Road to cost-effective crystalline silicon photovoltaics, 3rd World Conference on Photovoltaic Solar Energy Conversion, Osaka, Japan, 11-18 May 2003. Rohatgi, A., Ristow, A. and Yelundur, V. (2003), Cost and Technology Roadmaps for Cost- Effective Silicon Photovoltaics, 12th International Workshop on the Physics of Semiconductor Devices (IWPSD-2003), New Delhi, India, December 16-20, 2003. Schaeffer, G.J., Alsema, E., Seebregts, A., Beurskens, L., de Moor, H., van Sark, W., Durstewitz, M., Perrin, M., Boulanger, P., Laukamp, H. and Zuccaro, C. (2004), Learning from the Sun Analysis of the use of experience curves for energy policy
81 Chapter 3. Learning in the International PV Manufacturing Industry
purposes: The case of photovoltaic power, Final report of the Photex project: Report ECN-C-04-035, ECN Renewable Energy in the Built Environment. Schmela, M. (2005), Super Sonic Solar Market - Worldwide Market Survey - cell & module production, Photon International (March 2005), pp 66-82. Schmela, M. (2006), Silicon Shortage - so what! Market survey on cell & module production 2005, Photon International (March 2006), pp 100-124. Shum, K.L. and Watanabe, C. (2006), Photovoltaic deployment strategy in Japan and the USA-- an institutional appraisal, Energy Policy, In Press, Corrected Proof. Siemer, J. (2005), Saving intellectual property: Patents for photovoltaic applications, Photon International, November 2005, pp 36-40. Sinke, W., del Coso, G. and del Cañizo, C. (2006), Crystalline Silicon PV Technology Roadmapping in the Crystalclear Integrated Project, 3rd EPIA PV Industry Forum, in the Frame of the 21st EUPVSEC, Dresden, Germany. Stryi-Hipp, G. (2006), Photovoltaics in Germany: Market and Industry Development, German Special Renewable Energy Day, Solar Power 2006, San José, California / U.S.A, 17 October 2006. Surek, T. (2003), Progress in U.S. photovoltaics: looking back 30 years and looking ahead 20, pp 2507-2512 Vol.2503. Swanson, R.M. (2004), A Vision for Crystalline Silicon Solar Cells, Sunpower White Papers (Last Updated: June 2004), Accessed from: http://www.sunpowercorp.com/html/Resources/TP_index.html, on: November 2005. UNSW (2006), School of Photovoltaics and Renewable Energy Engineering, The University of New South Wales, Accessed from: http://www.pv.unsw.edu.au/Research/Researchindex.asp, on: March 2007. van der Zwaan, B. and Rabl, A. (2004), The learning potential of photovoltaics: implications for energy policy, Energy Policy, 32 (13), p 1545. Visentin, A., Voignier, V. and Königstein, C. (2005), Technology Trends in Photovoltaics for the Last Decade: An analysis on the basis of patent application filing data, 20th European Photovoltaic Solar Energy Conference and Exhibition, Barcelona, 6.- 10 June 2005. Watanabe, C., Nagamatsu, A. and Griffy-Brown, C. (2003), Behavior of technology in reducing prices of innovative goods--an analysis of the governing factors of variance of PV module prices, Technovation, 23 (5), p 423. Watanabe, C., Wakabayashi, K. and Miyazawa, T. (2000), Industrial dynamism and the creation of a "virtuous cycle" between R&D, market growth and price reduction: The case of photovoltaic power generation (PV) development in Japan, Technovation, 20 (6), p 299. Watt, M. (2003), The Commercialisation of Photovoltaics Research in Australia, A report for Science and Innovation Mapping, Department of Education Science and Training, Canberra, Australia. Watt, M., Passey, R. and MacGill, I. (2006), Photovoltaics research and development in Australia, International Journal of Environmental Studies, 63 (6), pp 777-790. Wenham, S.R. and Bruce, A. (2002), The Virtual Production Line for the Manufacturing of Screen-Printed Solar Cells, CD-Rom published by the Key Centre for PV Engineering, The University of NSW, Sydney, Australia. Witt, E.C., Mitchell, R.L., Symko-Davies, M., Thomas, H.P., King, R. and Ruby, D.S. (2001), Current status and future prospects for the PVMaT project, Solar Energy Materials and Solar Cells, 67 (1-4), pp 355-362.
82 CChhaapptteerr 44.. TThhee PPVV IInndduussttrryy aanndd SSmmaallll SSccaallee MMaannuuffaaccttuurree iinn DDeevveellooppiinngg CCoouunnttrriieess
The purpose of this chapter is to establish a role for the small scale manufacture of PV system components and to identify the advantages and barriers to commercialisation of the technology in developing countries. The small scale manufacture of PV system components in developing countries is of importance because photovoltaics has a significant role to play in the provision of electricity to the rural poor, as described in section 4.1; while small scale enterprises are an essential element in rural development, as described in section 4.2. In section 4.3, typical PV diffusion modes, and actors and relationships in PV value chains in developing countries are described, identifying the interactions of manufacturers in these value chains. The technology for balance of system components (BOS) and module manufacture and typical failure modes are described in section 4.4, identifying technical challenges for small scale manufacturers. In section 4.5, accounts of small scale manufacture of PV system components from the literature are summarised. In section 4.6, through the analysis of the documented cases of small scale manufacture and the literature on the dissemination and use of PV technology for rural electrification; barriers and advantages for small scale manufacturers and suitable interventions to support these enterprises are identified.
83 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.1. The Role of Electricity from Photovoltaics in Development
4.1.1. Energy and Poverty Since the 1960s, when integrated rural development projects usually contained an energy component, energy has not been widely recognised as a “basic need” in development circles, with the focus being, understandably, on issues such as clean water, food and health care (Cecelski, 2003). There has, however, recently been more recognition of the role of energy in fulfilling development targets. Energy, even if not defined as one of the basic human needs, is now seen as crucial to the fulfilment of these needs, and access to modern energy, in addition to or instead of traditional biofuels, is now widely assumed to be vital in poverty reduction. The UN Millennium Development Project recommends that energy services, and electricity in particular, are vital to achieving the Millennium Development Goals (Box 4-1).
Box 4-1: UN Millennium Development Project Recommendations on Energy Goals
Improved energy services—including modern cooking fuels, access to electricity, and motive power—are necessary for meeting almost all the Goals. They can reduce child mortality rates and improve maternal health by lowering indoor air pollution. They can reduce the time and transport burden of women and young girls by reducing the need to collect biomass. And they can lessen the pressure on fragile ecosystems. Electricity is critical for providing basic social services, including health and education, and for powering machines that support income-generating opportunities, such as food processing, apparel production, and light manufacturing.
The UN Millennium Project proposes that countries adopt the following specific targets for energy services to help achieve the Goals by 2015: Reduce the number of people without effective access to modern cooking fuels by 50 percent and make improved cook-stoves widely available. Provide access to electricity for all schools, health facilities, and other key community facilities. Ensure access to motive power in each community. Provide access to electricity and modern energy services for all urban and periurban poor.
Source: (UN, 2005, p 30)
The impact of energy poverty cannot be isolated from the poor’s broader lack of access to livelihood assets and capabilities. The poor are unable to afford good quality energy services, which constrains their livelihood activities, thus reducing their productivity and profitability. Low income results and the poor continue to be unable to afford to pay for improved energy services. A vicious circle of energy poverty has been identified by IDS (2003), as illustrated in Figure 4-1.
84 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
Figure 4-1: The Vicious Circle of Energy Poverty
1. No energy to run machines, 2. Low productivity, low low productivity, surplus, poor quality, range of little cash output & time poverty
vicious circle
3. No money to buy improved energy supplies or energy conversion equipment
Source: (IDS, 2003)
IDS has highlighted the importance of improved energy supply as an intervention that can increase the productivity and sustainability of livelihoods. A virtuous cycle to break out of energy poverty is identified (Figure 4-2), whereby increased productivity, as a result of energy, increases income and therefore enables the poor to purchase improved energy services. This paradigm emphasises the importance of productive end uses, particularly those that generate cash income. Building assets, including human capital via health and education, however, can also increase people’s opportunities for productive livelihood options (Barnett, 2001, 2005; Wilkinson, 2002).
Figure 4-2: A Virtuous Circle to break out of Energy Poverty
2. Increased 1. Increased access to productivity. improved energy Women gain time for services economic activity
virtuous 5. Money to buy circle improved energy 3. Increased sales, supplies or energy surplus and profit equipment
4. Increased income
Source: (IDS, 2003)
85 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.1.2. The Impacts of Electricity on Poverty An extensive study of the linkages between energy, poverty and gender has been undertaken by the Energy, Poverty and Gender (EnPoGen) project of the World Bank (Ramani & Heijndermans, 2003), revealing more accurately the nature of the relationship. Box 4-2 summarises the findings as they relate to electricity.
Box 4-2: Energy Poverty Gender Project Findings
The study has verified that electricity does impact the lifestyles and living conditions of the poor, primarily through electric lighting. Access to modern communications technologies, such as television and radio, has also been found to alleviate isolation. Other lifestyle improvements may be obtained by the use of appliances such as water heaters, clothes irons and grinders. What is less certain is the impact of electricity on health and education. Electricity only has an impact where health and education facilities exist. Electricity does not significantly improve air-quality fouled by cooking fuels, which is the major cause of indoor air pollution. Nor does electricity usually impact children’s study time when a dwelling is electrified. However, other significant time savings result from electrification, although women tend to use the additional time to do more household chores, whereas men have additional leisure time. Interactions between households and communities also result from electrification, enhancing opportunities for development.
Livelihoods and income are impacted by increases in agricultural productivity where technologies such as water pumping and product processing are employed. Only a small proportion of electrified households have been found to engage in home-based microenterprise. Enterprises and businesses can benefit significantly from electrification, depending on the quantity of electricity available and conditional on the satisfaction of other enabling factors such as available capital and access to markets. The amount of money spent on fuels for lighting is on average decreased by 30-40%, but increased use of appliances normally increases electricity use to the same level of expenditure. In general communities can be enhanced by the presence of electricity, increasing the value of land and property.
Source: (Ramani & Heijndermans, 2003)
Electricity clearly provides some direct benefits for enhanced living conditions and lifestyles, and provides opportunities for improved education, health services, increased income and livelihood opportunities, provided the right conditions and complementary inputs are in place. Electricity is of special importance as an energy source, because it is the only energy source that can drive many modern appliances, including communications technologies. Electricity can conveniently be used to supply almost all energy end uses, is equated with modernisation and is highly desired by those who don’t have it.
4.1.3. Decentralised PV Systems and Development The cost of delivery of services such as health, education and energy to rural areas is inversely proportional to the population density (James, 1989). There is consensus that centralised electricity grids will not reach many of those without access to electricity due to (IEA, 2003b): long distances, dispersed population,
86 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
low demand, low ability to pay for electricity or appliances.
Electricity supply often does not reach poor villages, and even where a village is ‘electrified’, it is common that many buildings are not connected to the grid. Even where a grid connection or other electricity service is available, the poor often cannot afford to purchase them. Decentralised (off-grid) supply has the potential to provide greater access to and availability of electricity for those rural communities for whom a grid connection is not available, and is seen to be the most probable way large numbers of rural poor people will be able to avail themselves of modern energy services (Foley, 1990; Huacuz & Gunarante, 2003). It has also been suggested that decentralized planning processes and delivery of electricity supply empower the poor and more accurately inform decision making, improving acceptability (Ramani & Heijndermans, 2003). Decentralised supply may involve mini-grids, which are isolated from the main electricity networks, or stand-alone systems, which supply a single building or application. Decentralised generation options include fossil-fuel generators, small or micro-hydro systems, biomass generators, wind turbines or photovoltaics. Many locations do not have suitable hydro or wind resources, but solar resources are universally available, and solar irradiation is most abundant close to the equator, where most of the world’s poor reside. Stand alone photovoltaic systems for small scale remote applications are often less expensive over the life cycle of a system than grid or diesel generators (Barnes & Floor, 1996), and in many instances cheaper than extension centralised grid services (Cabraal et al., 1998). The quality of supply of electricity from remote PV systems is usually superior to remote grids, and the requirements for maintenance and spare parts required are low compared to alternatives such as diesel generators (Foley, 1990; Huacuz & Gunarante, 2003). The FAO has detailed the potential uses for PV electricity in development (van Campen et al., 2000) including an extensive list of productive uses for photovoltaics, which are seen to be economically sustainable. It should be noted at this point that despite the advantages of PV relative to other sources of electricity in remote areas, there are many barriers to successful use of the technology, as discussed in section 4.4. Having established the place for PV in rural electrification, the following section describes the role of small-scale enterprises in general terms, and the remainder of this chapter looks specifically at the role of and barriers to small-scale PV manufacture.
87 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.2. The Role of Small Scale Enterprises in Developing Countries
While growth via industrialisation has a role to play in development, there is also increasing recognition of the importance of targeted poverty reduction strategies. Economic growth via industrialisation as a focus for development has been criticised for its tendency to create a dual society (Schumacher, 1973), since it tends to centralise investment and control over assets and hence create inequalities or allow foreign interests to acquire ownership. Some countries, such as Korea and Taiwan, have achieved growth and stable or increased equity (Todaro, 1982), but it is more common for inequality to increase if no concerted effort is made to avoid it, as has been the case in most of the transition economies (UNDP, 1999). The poor often lose out economically as a result of industrialisation, since factors such as low wages, concentration of capital in urban areas, exploitation of rural resources, debt, unfair trade and expensive imported inputs are all factors that may enhance the industrial growth of a country, but exacerbate the problems of rural poverty (Medellin-Erdmann, 1992). In addition, industrialisation can be a socio-culturally destructive process. Traditional agricultural methods, roles, social ties and cultural values may be eroded in favour of individualism, rationality and economising (Todaro, 1982). Urbanisation will certainly occur, and industrialisation usually involves environmental degradation (Bruntland, 1987). For these reasons, there is increasing recognition of the importance of targeted poverty reduction strategies and the development of small enterprises and informal sectors of the economy. Micro, small and medium-sized enterprises, often as part of the informal economy, provide the majority of employment in developing countries (Overseas Development Institute, 2002). While modern sector firms have business objectives such as profit-making and market expansion as their primary purpose, the importance of small scale enterprises lies in their ability to enhance the livelihoods of the poor, who are often isolated from the processes of modernisation and industrialisation (Khosla, 1994). The social objectives of small manufacturing enterprises may include: Employment, income and livelihood diversification, Resource use and keeping money in the local economy, Providing technology for the rural market, The development of local capabilities. The following sections elaborate on these objectives.
88 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
Employment, Income and Livelihood Diversification Employment is identified by (Scoones, 1998) as one of the key elements of a livelihood. He notes the importance of ‘recognition for being engaged in something worthwhile’ as well as income and production. Small scale industries are generally more labour intensive than larger enterprises (Albu, 2001), generating more employment per unit of output and per unit of energy (Fernando, 1992). They are also able to generate employment more suitable to women (Islam, 1992). Rural enterprises can provide employment in rural villages, preventing migration to urban areas (Panditrao, 1994). It is also believed that the existence of small enterprises creates a more equitable income distribution (Aftab & Rahim, 1989). Income from non-farm sources can keep help to rural wages higher, and reduce livelihood vulnerability to seasonal or disaster-related variations in agricultural output or employment opportunities. People or households can spread their risk by working in both sectors, and non-farm employment can provide a means for survival if farm work cannot. The importance of rural livelihood diversification away from agriculture has been observed in Africa, where the less poor have been seen to have more diversified incomes, and the poor are engaged mainly in low return farming activities (Dorward et al., 2003). Rural non-farm income makes an increasingly important contribution to household income: 40-45% in sub-Saharan Africa, Latin America and South East Asia and 30-40% in South Asia (Davis, 2004). Livelihood diversification away from agriculture can be driven by two processes: distress-push where the poor are driven to seek non-farm employment due to lack of adequate farm employment opportunities; and demand-pull, where rural people take advantage of new opportunities. In the distress-push situation, people may be drawn into poorly remunerated activities, but ones which have low barriers to entry; whereas demand-pull opportunities are more likely to lead to improved livelihoods (Davis, 2004). The entrepreneur is a key figure in exploiting demand-pull diversification activities. In the case of distress push diversification, people may be forced into entrepreneurial activity, rather than choosing it (Dunnett, 2001).
Resource Use and Keeping Money in the Local Economy Small enterprises may produce goods or services for the local economy, or for export to markets outside the local economy (Davis, 2004). Small enterprises employ entrepreneurial talent and savings that would not have otherwise been mobilized (Aftab & Rahim, 1987), and may employ labour more productively than agricultural activities. Small enterprises may therefore be drivers of economic growth. Dorward (2001) highlights the importance of linkages and leakages in a local economy. Activities producing goods for external markets may bring money into the local economy, while the provision of goods and services for local use prevents leakages to pay for externally produced goods. Entrepreneurial activities that rely on local resources are preferable, since they
89 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
do not require money leaving the local economy to pay for capital goods and materials (Masum, 1992). Activities that generate cash income, rather than goods that are exchanged locally are likely to result in demand for externally produced goods which can be purchased with cash. Dorward (2001) recommends that production and market opportunities should be analysed for their potential to provide local jobs and demand for locally produced goods. Some small enterprises may contribute to the regional dispersal of industries, and in some cases, can spread the linkage effects of large scale industries (Islam, 1992). However, support for such often high-return non-farm enterprises may ignore the ultimate dependence of the rural economy on agriculture and the constraints that prevent the poor from participating in such activities. They may bring in cash but therefore increase demand for externally produced goods, rather than increasing local economic activity.
Appropriate Technology for the Rural Market Since the provision of goods and services for the poor is not highly profitable, the modern sector may not receive the necessary market signals from the poor to induce the development of tools and technologies suitable for the poor. Small scale manufacturers therefore have a role to play in providing these goods and services. Poor people in rural communities need low cost tools and technology to carry out small scale agricultural and other productive work. The technologies required may be different from those required by other sectors and hence may not be provided for in urban markets. “Technology is created in response to market pressures – not the needs of poor people, who have little purchasing power” (UNDP, 2001). Small scale manufacturers can provide technology that is appropriate for the poor, in terms of the capabilities of the users, their ability to use and maintain the technology; and the financial resources, materials and parts available. Locally adapted technologies may also be more compatible with the aspirations of the community, the culture and the society. Improved production techniques, whether agricultural or non-agricultural, may increase productivity; while access to a new product may enhance lifestyles. Decentralised small scale manufacture, which does not rely on the advantages of economies of scale in capital goods production and manufacturing, requires less standardisation in product and process design (Krishnaswamy & Reddy, 1994). Where technologies need to be adapted, repaired or maintained locally to suit local conditions and activities, local manufacturers are therefore well placed to provide them in a cost effective and sustainable manner.
90 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
The Development of Local Capabilities Technologies require maintenance skills and infrastructure in order to be used effectively. Local manufacture may help to develop and improve the capabilities for after-sales service (Krishnaswamy & Reddy, 1994). Even where technology is relatively complex, centralised manufacturing of the components and small-scale assembly is feasible, and still has the potential to impact maintenance capabilities. Secondary benefits to the adoption of new technology may be related to the acquisition of skills that may have application in other industries or areas of life. The use of technology requires, but also enhances the technological capabilities of entrepreneurs, their employees and users of the technology. Through enterprises, material and social assets may be built, reducing vulnerability and increasing livelihood options for poor communities.
91 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.3. Markets and Actors in the PV Industry in Developing Countries
Having established roles for electricity from photovoltaics and for small enterprises in rural development, this section describes the most common types of markets, actors and value chain relationships in the PV industry in developing countries, providing a background for identification of the barriers to successful small scale manufacture and commercialisation of PV in these markets.
4.3.1. PV Markets in Developing Countries PV diffusion is generally through cash sales or projects or programmes that may be either government or NGO driven. About two thirds of PV systems distributed in developing countries have been sold through private cash sales (Nieuwenhout et al., 2001). Cash sales of photovoltaic systems through commercial dealers have been particularly widespread in Kenya and China. Because of the importance of energy to basic needs, and as an input to productive activities; where the market cannot supply energy to rural communities, governments and development organisations frequently support subsidised or welfare driven delivery of rural electricity services, including photovoltaic systems. Large government programmes have been implemented in India, Mexico and Indonesia; while there have been numerous NGO projects in many developing countries. In these non-market diffusion models, the institutional arrangements must perform all the functions that are normally carried out by the market, such as market (needs) surveys, demand forecasting, pilot testing, marketing, provision of credit, after- sales service etc (Krishnaswamy & Reddy, 1994). Many projects or programmes have included subsidies and / or loans and, more recently have included provision for maintenance, spare parts and fee collection. Fee-for service models have also been demonstrated recently whereby the utility or service provider continues to own the hardware and services the system, while the user pays a fee for the energy service (EDRC, 2003).
4.3.2. Actors in the Small Scale PV Industry The value chain participants and interactions in the PV industry of developing countries generally take a form similar to that represented in Figure 4-3.
92 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
Figure 4-3: Photovoltaics Industry Actors and Interactions in Developing Countries
IMPORTER RETAILER SYSTEM ASSEMBLER DISTRIBUTOR
System Components Sytems Systems Installation LOCAL ORGANISATION USER After Sales Service Coordination
SERVICE PROVIDER / UTILITY SMALL-SCALE URBAN MANUFACTURER Sytems Installation INTERNATIONAL / After SalesService DOMESTIC SUPPLIER
Parts, Materials SMALL-SCALE System Components RURAL MANUFACTURER
PROJECT DEVELOPER / INTERNATIONAL / IMPELEMTER / DOMESTIC TECHNOLOGY CONSULTANT DEVELOPMENT
Designs
International CentralRegional Local
The figure shows the actors in the small scale supply chains, as well as the competing channels of production and distribution. International, centralised, regional and local phases in the supply chain are identified. Some functions of these actors and their interactions that impact small scale manufacturers are noted in the following sections. As illustrated in Table 4-1, although primary roles are identified for different types of actors, they may perform a number of additional functions.
Table 4-1: The Roles of Actors in the PV Value Chain in Developing Countries Actor Operation Maintenance Installation System Manufacture and Use Design User Local organisation Rural Manufacturer Retailer / Distributor Service provider / Utility Project developer Urban Manufacturer Importer / Assembler Supplier
Primary Role Additional Functions
Individual Users The market for rural PV systems is generally comprised of the poor, who, when they are faced with livelihood stresses and shocks, are less able to make payments for non-essential services and often adjust their spending on other things in order to be able to afford food. These users are unlikely to be able to carry out basic maintenance, unless they are specifically trained. User maintenance has been found to be partially successful in community buildings, but less successful in solar home systems (SHS), where users often do not have the capability (Huacuz & Gunarante, 2003). They are also unlikely to have sufficient design capabilities to put together
93 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
a functioning system from available components. For instance, if users purchase components and assemble their own system, they are likely to have mismatched panels, batteries, wiring and loads.
Local Organisations Local groups are often formed within projects or programmes. Pre-existing groups are sometimes utilised. These groups perform various tasks that require a local presence and understanding of local cultural factors, such as local management, monitoring, basic maintenance and fee collection. These groups, because of their local social capital, can assist in ensuring that payments are made and may guarantee the debt of members, giving them better access to finance.
Rural Manufacturers Rural manufacturers are likely to sell to local organisations that provide PV services, or directly to rural users, sometimes competing with utilities and retailers. They may also access larger markets through involvement in projects or programmes, including fee-for service arrangements. Manufacturers will need to purchase from urban or international suppliers of equipment and materials and will need to acquire product and process technology externally.
Urban Manufacturers Domestic manufacturers located in urban areas compete with businesses that import and assemble systems. They may in some cases sell their products directly to users, but will usually sell to retailers/distributors, utilities or project developers. They also need to acquire materials and equipment; and may either develop technology internally or acquire it externally. Urban manufacturers are likely to have more difficulty than local manufacturers in linking with rural markets, but may have better access to input markets, technology and skilled personnel.
Retailers, Distributors Retailers and distributors sell complete systems or components to users or local organisations. They are sometimes contracted to carry out the distribution for a project or programme. They may or may not provide after sales service. PV systems are often just one of many items they sell. Retailers and distributors are primarily interested in making a profit, but may also have some social objectives if they are located rurally.
Utilities PV utilities (service providers) sell electricity to users for a regular fee. They may be private, often under an exclusive concession arrangement, or may be provided through a project or by the state. Since users will cease payments if the system ceases to supply electricity, service providers generally have some kind of maintenance contract with the users; which includes repairs and replacements. Service providers require technical and managerial 94 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
capabilities, including those for maintenance planning, management of remote staff, recruitment, training, inventory and supply of spare parts.
Project Developers and Implementers PV electrification projects or programmes may be executed by NGOs, government departments or utilities. The project developer/implementer will arrange for the procurement and installation of systems. Projects and programmes also often include provision for credit, maintenance and supply of replacement parts. These functions and associated fee collection may be carried out by service providers, private retailers or local organisations. Technical and management training for the staff of these agencies is often supported through projects or programmes.
Importers, Assemblers Importers and system assemblers bring in system components and may sell them on or assemble them into systems. They may be in competition with local producers.
Suppliers Suppliers provide parts and materials to manufacturers. Manufacturers may be able to find some suppliers located domestically, but in many countries will need to import materials and parts from international suppliers. International suppliers also provide system components to importers and system assemblers.
Technology Developers Sophisticated manufacturers may design their own equipment, but many will acquire designs through consultants or development organisations involved in projects or programmes. Information about user’s needs may not be available to international technology developers, while local developers may not have access to technology or information about quality and certifications.
4.3.3. Government Ministries Government ministries are often important actors in small scale PV technological systems, even when they don’t participate in the supply chain directly. A variety of ministries such as agriculture, education, energy, environment, health, industry, public works and water may have an interest and involvement in PV implementation. Ministries may influence project designs and allocation of funding. They may also create regulatory or quality control infrastructure. However, the background knowledge and expertise within ministries is likely to be oriented towards grid extension, rather than decentralised supply (IEA, 2003b). Awareness-raising about the suitability and requirements of PV technology may therefore be necessary. Because of the importance of energy to a variety of
95 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
ministries, it would be optimal for rural electrification to be viewed from an integrated development perspective, resulting in better coordination and linkages between the ministries.
4.3.4. Financiers Financiers may provide credit to users via local organisations, retailers or utilities. Project or programme developers often negotiate with private sector financiers over financing arrangements, fee collection and risk bearing. Retailers or manufacturers may also be able to arrange credit for their customers. Finance for PV systems is usually unattractive to private financiers because of the slow return on investment, high risk and low income of users, but they are starting to enter the PV market in some countries. In some countries, national banks may also provide credit to rural areas and often have rural branches. Alternatively, local organisations may organise their own credit system, usually based around a revolving fund. The participants contribute to the seed capital, which can then be borrowed by members and repaid.
96 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.4. Technology for Small Scale BOS and Module Manufacture in Developing Countries
In this section, the function and design of balance of system (BOS) components and modules are described. Typical failures modes are also discussed, since these are likely to cause the most problems for small scale manufacturers in developing countries. The most common balance of system components for PV systems include: Charge controllers, which regulate the charging of the battery from the PV module, protecting against overcharging or undercharging, which can reduce the life of the battery; A number of types of lamps, including fluorescent and LED lamps; DC/AC converters, which are required for the conversion of DC current into AC, required for running appliances such as colour televisions; and Batteries for energy storage.
This section deals only with controllers and florescent lamps, since they are among the most common BOS components, and, along with modules, are manufactured by the subjects of the case studies in this research. While batteries are equally common in PV systems, they are not the focus of this research, because they are not specific to PV systems. They are also considered to be of less interest to this research because their manufacture is less likely to contribute to better capabilities for maintenance, since they are less likely to be repaired in the field.
4.4.1. Charge Controllers The main function of a charge controller is to regulate the charge to the battery, which in solar home systems is usually either an open or vented lead acid battery. The charge controller limits the voltage and current supplied to the battery in order to protect the batteries from damage due to overcharging. Most controllers for non-critical loads also have a low voltage disconnect to prevent batteries from being over discharged. Most small controllers used in developing countries are simple on/off charge regulators, which require a relatively small amount of electronics. On-off regulators disconnect the PV array from the load when the high voltage set point is reached and then reconnect it when the voltage has dropped to a lower reconnect set point (Usher & Ross, 1998). Because solar home systems need to be as cheap as possible, more advanced and expensive charge regulators are not common. In addition, PV applications and marketing channels are both highly diverse, so it is
97 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
difficult to supply ‘standard’ small PV controllers. Charge controllers are therefore often hand- built by small-scale local manufacturers (Vervaart & Nieuwenhout, 2000). More sophisticated controllers can limit the current supplied to the battery linearly, so that a high voltage can be supplied without damage to the battery from high current. They can therefore more effectively fully charge the battery (Dunlop, 1997). This may be achieved by dissipating the additional power from the array, or by using a pulse-width modulator, which limits the current by breaking it into high frequency pulses (Usher & Ross, 1998). When more current is required, the pulses are on for longer and as the battery voltage rises, the pulse width is decreased. Some controllers can perform additional functions such as surge protection and temperature compensation to protect the battery from charging at high currents when the temperature is high. They may also provide indication of the operation of the system and may have different modes for different types of batteries. Sophisticated controllers can perform a number of charging modes, which more effectively charge and increase the life of the battery. The main battery charging modes employed by these controllers are summarised in Box 4-3.
Box 4-3: Charging Modes of Photovoltaic Battery Charge Controllers
Main charge, used for charging the battery up to a level when gassing starts and the voltage rises. This mode is used up to a voltage limit of around 2.39 V at 25 °C, or 2.33 V at 40 °C. Top-up charge, used to reach a 100% state of charge from a level of 90 - 95 % state of charge. In this mode, the battery voltage limit is maintained by decreasing the current supplied to the battery. Equalisation charge, which is used for equalising the charge of the individual cells in the battery. This is an important issue for extending the life of the battery, but requires a special controller mode. In this mode, the voltage is increased to around 2.5 - 2.6 V/cell for around half to 1 hour. Equalisation charging is recommended at regular intervals, once a week. Maintenance charge, used for maintaining full capacity in a battery that is already fully charged but not frequently used for some period. In this mode, the controller supplies approximately 2.20 - 2.25 V/cell.
Source: (Vervaart & Nieuwenhout, 2000)
Most charge regulators are designed with analogue electronics, which have historically been cheaper for such simple devices. Several features can be added to the system using analogue electronics, such as PWM, over-current protection and state of charge indication to the user (Vervaart & Nieuwenhout, 2000). Digital electronics (microprocessors) can reduce the amount of electronic components needed, and are therefore cheaper when the controller becomes complex. They can easily perform temperature compensation calculations, and calculate state of charge accurately using complicated algorithms. Extra functions or changes in functions can also be edited in digital controllers. Standard processors are available for around $US2.50 for large production quantities (Vervaart & Nieuwenhout, 2000). Microprocessors which contain extra functions such as analogue-digital conversion, PWM generators and memory which can be programmed in the field are now available, but add an extra $US10 to the
98 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
price. Only a few analogue inputs and steering outputs are then required in the regulating part of the circuit. However, most simple charge controllers continue to be made more cheaply with analogue components, particularly in developing countries.
4.4.2. Failures of Charge Controllers Charge controllers have been found to have the following problems related to design (IEA, 2003a; Katic, 2002; Kumar et al., 2000; Mulugetta et al., 2000): also see (Varadi et al., 2003) Inappropriate settings for low voltage and high voltage disconnect, Frequent fuse breaking or insufficient fusing and polarity protection, causing failure of the product with connection of incorrect polarity, Consumption of too much power, Inability to carry the rated current, Deep hysteresis (cycling between disconnect and reconnect).
Failures of charge controllers may also be caused by manufacturing problems and by improper use. These problems are common to all electronic BOS components and will be discussed in section 4.4.5.
4.4.3. Fluorescent Lamps Fluorescent lamp ballasts provide a very high frequency waveform of around 20kHz to a tube filled with low pressure mercury vapour in an inert gas. The electrical energy releases electrons from mercury vapour. When the atom recombines with the electron, the excess energy in the electron is released as a plasma of ultraviolet electromagnetic radiation, which is converted into the visible spectrum by a coating on the tube. The circuit for a lamp ballast must convert the DC electricity provided by the PV system to AC and therefore requires an inverter circuit. A transformer is also required to achieve the correct voltage. The waveform provided to the tube must have a symmetric shape, maintain the correct voltage, and avoid voltage spikes. If these conditions are not met, the ends of the tube will blacken and its life will be reduced (Vervaart & Nieuwenhout, 2000). The voltage supplied to the lamp will vary with the input voltage from the system without additional circuitry to keep it constant. The interactions between transformer coils, inverter circuits and lamp impedance, and the small differences between transistors with the same specifications can cause asymmetry and unexpected spikes in waveforms. Circuits must be carefully designed to mitigate these effects. Field Effect Transistors (FETs) are more efficient than conventional transistors and reduce voltage peaks in switching. They have until recently been more expensive
99 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
and require a different circuit, so are not compatible with the older circuit designs used by many manufacturers.
4.4.4. Failures of Fluorescent Lamps The fluorescent light is one of the weakest components in a solar home system. In many cases the tube starts blackening, or totally stops working very soon after installation. In addition to tube blackening, lamps have been found to have the following problems related to unsuitable design (Corkish et al., 2004; Katic, 2002; Kumar et al., 2000): Still drawing current when the luminaire has failed, No reverse polarity or short circuit protection, Radio frequency interference, High power consumption, Inappropriate frequency, leading to short tube life.
Failures may also be cased by problems related to the manufacturing or use of lamps, which will be discussed in the following section. Another common problem in the field is breakage of tubes during transport to rural areas. The transformer in the ballast can also be very fragile. If the secondary coil wire is dimensioned too thin or not mounted securely on the circuit board, the likelihood of failure in the field is increased (Vervaart & Nieuwenhout, 2000).
4.4.5. Problems in the Manufacture and Use of Electronic BOS Components The quality of the design of lamps and charge controllers is critical to ensuring good performance and long life. Small manufacturers are unlikely to have the capabilities to design their own circuits, but may be able to access the latest designs from universities or NGOs. In addition to the problems related to poor design described in previous sections, many failures of electronic BOS components can be attributed to one or more of the following factors (Vervaart & Nieuwenhout, 2000): Lack of availability of appropriate and reliable components, Insufficient quality control during manufacture, Poor workmanship, leading to risk of short circuits, Variation of operating conditions compared to the rated design conditions, Misuse by the user.
Manufacturers therefore need to have good linkages with retailers and users of systems to ensure that transport, installation, maintenance and operation are carried out appropriately. They also need good linkages with component suppliers to ensure reliable access to quality
100 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
components. Manufacturers may have difficulty sourcing the components specified if they are using designs from international organisations. The processes involved in BOS manufacture are simple, including identifying electronic components (resistors, capacitors, ICs, diodes, transistors), soldering them onto PCBs, winding transformers, constructing the housings for the electronics. Quality control for BOS is more challenging, involving operating multimeters and power supplies, understanding electrical quantities, the function of different components and the correct operation of the circuits and being able to predict which components are likely to cause different faults. Although small scale manufacturers are unlikely to be able to design their own circuits, they must have a detailed understanding of the operation of the circuit and quality control procedures.
4.4.6. Modules Commercial photovoltaic module assembly typically involves testing and sorting the cells into similar lots, soldering the cells into strings, forming a stack of (typically) glass- encapsulant-cells-encapsulant-Tedlar™ (shown in Figure 4-4), and curing the encapsulant by placing the stack in a vacuum lamination machine that carefully controls the temperature and pressure profile. The conventional technology available for module assembly using commercial techniques is capital intensive and not appropriate for the resources of small scale enterprises. Enterprises manufacturing modules at a small scale will therefore have to either use expensive modern sector equipment, or develop processes appropriate to small scale manufacture.
Figure 4-4: Typical PV Module Construction
Figure has been removed due to copyright restrictions.
Source: (Wenham et al., 2006)
A transparent adhesive such as ethylene vinyl acetate (EVA) holds the glass to the cells and surrounds them, creating a sealed watertight unit. The encapsulation material must keep out water, water vapour and gaseous pollutants that could degrade the cell, or corrode the metal contacts. EVA has been specially developed for use in PV modules and has high optical transmittance, good adhesion to different module materials, can accommodate stresses induced
101 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
by differences in thermal expansion coefficients between glass and cells and provides good electrical insulation (Czanderna & Pern, 1996). Glass is usually used as the top layer of PV modules since it is highly transparent to light between 350nm and 1200nm, does not stain easily, and is easy to keep free of dirt (Sandia, 1999). The glass is usually low-iron and often hardened (tempered) to prevent damage from hail or wind. It also contains cerium, which protects the EVA from discolouration (King et al., 1997). The function of the Tedlar at the back of the module is to protect the EVA from damage, hence ensuring the integrity of the seal.
4.4.7. Module Failure Modes Since there are no moving parts in PV modules, they are very robust, and manufacturers currently guarantee them for around 20-30 years. The PV cells themselves are the most reliable part of the module. Catastrophic failures in modules may result from breakage of the module caused by hail, poor handling of the tempered glass or thermal cycling (Wenham et al., 2006). Any breach of the encapsulation can cause the module to fail, since it can allow moisture to enter, resulting in corrosion, which is the cause of 45% of field failures of PV modules (Wohlgemuth, 2003). The existence of bubbles in the encapsulant material can lead to water ingress, as thermal cycling will cause the bubbles to expand and contract. Over an extended period, delamination of the encapsulant at the interface to the cell, glass or backsheet will occur. Problems were initially encountered with EVA discolouring or ‘browning’ after long- term exposure to UV sunlight at high operating temperatures (near 50˚C) (Czanderna & Pern, 1996). Moisture migration through module backing materials and through encapsulants is also associated with encapsulant browning and with the chemistry associated with module delamination (Pern & Glick, 2000). Discolouration of the encapsulant leads to reduced light entering the solar cells, reducing output. New formulations of EVA and the addition of cerium to glass has largely eliminated this problem (Sandia, 1999). Gradual increases in the series resistance of PV modules typically result in a decline in power output of 0.5% per year, which contributes a significant fraction of the approximately 1±2% per year losses measured in PV modules in the field (Sandia, 1999). Series resistance increases are caused by daily thermal cycling, fatigue of tin/lead solder bonds (Wohlgemuth, 2003) and corrosion, which can sometimes lead to short circuiting of cells. Short circuits can also be caused by manufacturing defects, if top and rear contacts are shunted by the interconnect tape (Figure 4-5).
102 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
Figure 4-5: Cells Shunted by Interconnect Tape
Figure has been removed due to copyright restrictions.
Source: (Honsberg & Bowden, 1999)
Breakage of interconnects can occur due to thermal stress (if there is not sufficient excess interconnect tape to cope with the shrinking of the tape at low temperatures), or corrosion, and can also lead to open circuit conditions.
Figure 4-6: Interconnect with Stress-Relief Loop
Figure has been removed due to copyright restrictions.
Source: (Honsberg & Bowden, 1999)
When mismatched cells are at or near short-circuit, power from good cells is dissipated in poorly performing cells and localised heating occurs in the bad cell (hot-spot heating). Either cells or encapsulation materials can be damaged by hot spot heating (Wenham et al., 2006). Hot spot heating can be caused by the failure or shading of one cell in the module. This problem can be mitigated by the use of bypass diodes for groups of cells in the module, so that current will bypass the group through the diode, rather than being dissipated in the bad cell. Other problems include open circuit failures in the bus wiring or junction box of the module and poor soldering of the interconnects to the busbars, resulting in series resistance losses. In summary, manufacturers need to ensure: The use of bypass diodes and high quality encapsulant and glass, The correct application of encapsulant and backsheet to avoid moisture ingress, Careful soldering and wiring of the junction box and good quality control,
103 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.5. Experiences with Small Scale BOS and Module Manufacture in Developing Countries
The formation of a local PV industry consisting of small enterprises which retail and service systems is seen to be a sustainable path to building local infrastructure and skills (Acker & Kammen, 1996; UNEP, 2003; Wilkins, 2002). It has also been suggested that the local manufacture of PV system components can contribute to a more sustainable use of the technology via improved technical knowledge and skills, availability of spare parts, access to information and potentially lower cost (Acker & Kammen, 1996; Gillett & Wilkins, 1999; Green, 2004; Mulugetta et al., 2000) However, while the manufacture of BOS for remote applications in developing countries is primarily carried out by small enterprises, often in developing countries (Vervaart & Nieuwenhout, 2000), there is very little documentation of the nature and extent of these small scale manufacturers. There are also no readily available statistics on the global distribution of BOS manufacture. Almost 40% of the world’s module assembly capacity is located in developing nations (Hirshman et al., 2007), while lower cost manufacturing, and expectations of a growing share of the global market in countries such as China are increasing this share (Li, 2004). The small scale manufacture of modules, however, is virtually unheard of apart from the case study of Grupo Fénix in chapter 8. The PV literature has been surveyed for any description of local manufacture, and experiences from Bolivia, Brazil, China, Indonesia, Kenya, Kiribati and Tuvalu, Nepal, Sri Lanka, Zambia and Zimbabwe are summarised in this section. In addition to these countries, manufacture of BOS takes place in India, Bangladesh, the Philippines (Kumar et al., 2000), Mexico (Huacuz & Agredano, 1998), Ghana, Morocco, Namibia and South Africa (Moner- Girona et al., 2006), and certainly many others.
4.5.1. Bolivia A Spanish PV project carried out in the Bolivian altiplano from 1988-1993 included the establishment of rurally-based manufacturing of balance of systems components in a small town of 50 families, 45 km from La Paz (Aguilera & Lorenzo, 1996). The circuits and casings for simple series charge regulators with MOSFET switches and 18W fluorescent lamps were manufactured on the basis of designs from the Institute of Solar Energy in Madrid, Spain. The charge regulators were designed specifically for the project, and had constant voltage set points, except for on the overcharge set point, which was temperature
104 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
corrected. The lamp ballast design was already proven in the Spanish solar home system market. The manufacturing facility occupied 120 m2, and was divided into two parts: mechanical and electrical. The workshop was equipped with metalworking and electronics equipment, at a total cost of US$30 000. Since Bolivia has no electronics industry, the enterprise was the primary importer of electronic components in Bolivia. The enterprise employed seven local workers, who manufactured the BOS components and installed and maintained the systems. 70% of their time was occupied with manufacture, 25% by installations, and 5% for system maintenance. A user’s association was formed to manage and maintain the systems, but financial management problems caused the Spanish agents to withdraw and the project stopped (Huacuz & Gunarante, 2003). Lamps and charge controllers produced in Bolivia were cheaper than those produced in Spain, as illustrated in Table 4-2. On the basis of maintenance data from the Bolivian project, the locally produced equipment performed favourably compared with imported Spanish components. The first battery replacement took place 5 years after the system installation. The steady growth in demand experienced for the products up until 1992 is also an indication of quality.
Table 4-2: Summary of Rural Bolivian BOS Production Equipment Total Units Units Price on Spanish Cost in Bolivia Produced by Manufactured per Market ($US) ($US) 1996 Man per Week Electronic Ballast 7000 40 10.3 9 Charge Controller 1500 3 115 90 Module Support 1400 9 40 33 Structure Reflector and 7000 100 5 3 Housing for Lamp Source: (Aguilera & Lorenzo, 1996) According to Aguilera & Lorenzo (1996), the establishment of local manufacture has improved the local technical skills for maintenance and repair. The ability to repair damaged components, in particular those damaged by lightning strikes, which are the most common form of system failure on the altiplano, has also reduced the average life cycle cost of the systems.
4.5.2. Brazil In Brazil, local lamps designed for use in buses were cheaper, locally available and more reliable than imported CFLs from international manufacturers with a good reputation (Zilles et al., 2000). Imported lamps from the 15 SHSs in a small project all failed within 6 months. Simple, locally produced on-off charge controllers have also been used effectively in Brazil. High set points, however, have been shown to increase water consumption and reduce battery life (Zilles et al., 2000).
105 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.5.3. China In a western China NREL project, charge controllers, batteries and lights were manufactured locally (Stone et al., 1998). Imported components were often preferred by suppliers due to their perceived higher quality and reliability. VAT and tariffs, however, make imported equipment expensive in China. The PV component of the GEF/World Bank assisted China Renewable Energy Development Project (CREDP) (Cabraal, 2000, 2004; Stone et al., 1998; ter Horst & Zhang, 2005) has developed technical specifications for modules, charge controllers, inverters and DC lights during the project preparation stage. National PV testing facilities have been established and national component and system standards adopted. 50% of the cost of testing was covered by CREDP for businesses whose components complied. The project management office has developed a list of qualified component suppliers and product descriptions. Prior to the project, Chinese markets usually opted for the most price-competitive products, resulting in low quality and loss of consumer confidence. The Hefei University of Technology was contracted to provide design assistance to Chinese manufacturers to improve the quality and performance of their products to meet the standards (Li, 2003), resulting in recommendations for changes in charge controllers subsequent to testing. Assistance in the form of grants was provided to manufacturers for research and development to improve products and form technical collaborations. The commercial capability of PV enterprises has been improved via business development assistance (training, workshops and tailored services). There is little activity in module assembly in the least developed areas of China. Some rural Chinese enterprises that manufacture BOS are reportedly considering module manufacture (Ling et al., 2002). The CREDP project achieved only 2% of the 10 MW target by 2002. The qualified local retailers, who were eligible for US$1.50 /Wp subsidies now face unpredictable competition from the government Brightness Programme (Li, 2004).
4.5.4. Indonesia Results of an investigation of the Indonesian BANPRES-LTSMD projects revealed that the technical components used were not of sufficient quality. Nine charge controllers produced in Indonesia were investigated as part of a study of the socio-technical aspects of the Indonesian BANPRES-LTSMD projects (Fitriana et al., 1998). The controllers were found to have inappropriate voltage thresholds (especially under-voltage protection) for the lead-acid batteries in use. The authors of the study suggest that the manufacturers based their designs on literature, rather than the real batteries and operating conditions, and recommend random charge and discharge tests with the batteries concerned in order to set appropriate thresholds. An alternative solution suggested is the use of charge controllers based on state of charge. This type of charge
106 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
controller, however, is not yet available even in industrialised countries, and is likely to be too technologically complex and expensive. Six automotive (SLI) batteries produced in Indonesia were also characterised by performing ten charge and discharge cycles, revealing that none of the batteries achieved the nominal capacity of 70Ah, with four of the six batteries remaining below 80% of nominal capacity (normally defined as the end of lifetime for a lead-acid battery). After the cycling, first ageing effects have also been detected. The Indonesian Ministry for Research and Technology BPPT and its energy laboratory LSDE worked with German institutions Fraunhofer ISE and TüV Rheinland to establish and accredit a test centre for the certification of PV components and systems to be installed in a collection of projects under a national 50MW programme (Fitriana et al., 1998). A World Bank and GEF renewable energy project was part of the Indonesian 50MW programme and was implemented in the 1990s (Nieuwenhout et al., 2004). Technical specifications were developed for the project and local manufacturers were supported in achieving certification. Of four fluorescent light ballasts submitted for certification, one passed immediately and the others were assisted by ECN in the Netherlands to improve their designs. One of the manufacturers required a complete new design, while another required the inclusion of a much more expensive component in order to pass the tests. The manufacturers did not implement the changes straight away due to low demand at the time. The other manufacturer later replaced an essentially good design with a bad imitation of an expensive high-quality design, which has caused many failures due to component ratings being too close to the current carried in the circuit. The introduction of standards in Indonesia has therefore not led to quality improvements. The ECN authors of the case study believe that field testing is required to ensure certified products perform over the long term.
4.5.5. Kenya There are as many as 20 companies involved in the manufacture of 12V lamps and charge controllers and three battery manufacturers in Kenya (Moner-Girona et al., 2006). In 2000 more than 90 percent of the batteries, 30–50 percent of the lamps, and 10 percent of the charge regulators used in local solar home systems were manufactured locally (Hankins, 2000). In the Kenyan PV industry, avoidance of transport costs and import duties has made domestically made batteries and lamps cost one third to half as much as imported components (Acker & Kammen, 1996). Import duties have added up to 35% to the price of imported components (Hankins, 2000). The Kenyan PV industry has been able to provide smaller and lower cost systems in response to market demand. This has occurred both through local manufacture and the importation of modules and BOS, particularly from China.
107 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.5.6. Kiribati and Tuvalu Early Pacific island PV experiences were poor, due to extreme remoteness, harsh environmental conditions, lack of training and social issues. The Kiribati and Tuvalu utility model was the first to be at all sustainable in maintenance and operation (Wade, 2003). Local utilities have been established with the support of the EU in Kiribati (SEC) and Tuvalu (TSECS) (Gillett & Wilkins, 1999; Tani, 2003b; Wade, 2003). 250 lighting systems were installed in Kiribati and 50 in Tuvalu by 1995. The cost of the hardware was covered by the EC, and the Kiribati utility is financially sustainable in maintaining and operating the systems, as well as systems installed under previous projects on the islands. Poor management has, however, led to problems in Tuvalu. The utilities own the panel, battery and charge controller; while the user owns the appliances and wiring. The utilities maintain the system and the user is responsible for the appliances and lights after a 1 year warranty period. Field technicians from the villages collect fees and perform maintenance, checking the systems at least once a month. These local technicians are trained and paid by the utility using the monthly fee. A senior technician in the SEC head office manages the field technicians and visits each site twice a year to deal with electronics problems. User committees provide communication between management and the users to clarify rights and obligations and resolve disputes (Wade, 2003). SEC senior technicians manufacture charge controllers, DC-DC converters and night lights locally using designs provided by EU. The controller has a good reputation and is sometimes exported. The controller is designed using an open circuit board to be repaired or refurbished in the field. In contrast, most controllers on the international market use microelectronics and sealed units, which cannot be repaired locally. The design for the BOS connections and box have been improved over time SEC had created 13 full time and 14 part time jobs in Kiribati and TSECS had created 10 full time jobs in Tuvalu by 1998. In Kiribati, maintenance duties have been performed well by the technicians, and it is expected that the batteries may last beyond their design life of 5 years. The utilities gained skills for management and local manufacture through the initial tendering process for the project. In Kiribati, both external and internal training courses and on the job training has built technical and management capacity through the project. Although some trained employees have left the utilities, their training will increase PV capabilities in the region. The capabilities of the Kiribati senior technicians to understand and maintain the BOS are enhanced by their manufacturing capabilities. In Tuvalu however, the technicians did not fully understand how the controllers worked and were unable to carry out equalisation charging on the batteries. There had been insufficient training of TSECS technicians since the start of the project and it was concluded that they needed follow up training.
108 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
A review of the programme carried out in 1998-1999 (Gillett & Wilkins, 1999) concluded that the senior technicians require more advanced training to be able to solve the complex problems which arise in the field, and that the utility staff require better business skills to manage procurement, stock control and management functions. It was also recommended that the successful manufacture of charge controllers should be expanded.
4.5.7. Nepal In 1998, there were three manufacturers of PV products, including charge controllers, lamp ballasts and DC-DC converters (Shestha et al., 2000). One Nepalese company had a module lamination machine by 2002, but was unable to operate it due to unstable power supply (Katic, 2002). With the support of SIDA (Swedish International Development Cooperation Agency), a Swedish NGO, laboratory and field testing was carried out by CRE (Centre for Renewable Energy), a Nepalese NGO, on these components (Shestha et al., 2000). The best of the three charge controllers required the addition of heat sinks and improved regulation of reference voltage, but otherwise performed well. The best of three lamps required reduction of the radio frequency interference, but otherwise performed well. The modified designs from the project were transferred by the Nepalese manufacturer and CRE to engineers and technicians from Laos and Bangladesh during a three day training program including theory, assembly and testing procedures. This is an effective example of technology transfer for PV between developing countries. The Danish government, through the Energy Sector Assistance Programme (ESAP) later supported the establishment of training and certification of installers, interim quality standards for BOS and modules and a national solar energy test station (SETS) in Nepal (Katic, 2002; Nieuwenhout et al., 2004). 14 manufacturers who qualified were able to participate in a national subsidy scheme, ranging from 30-50% of system cost. Minimum standards were agreed with the manufacturers, and an interim technical standard was established. The SETS facility was given a mandate to provide useful feedback from test results and assist manufacturers in improving quality through quality product development and testing training programs. SETS is also conducting field studies and providing feedback to manufacturers. SETS intends to gain PV-GAP testing accreditation.
4.5.8. Sri Lanka In the late 1980s, Solar Power and Light Company (SPLC) manufactured PV modules, 12V lamps and charge controller domestically, importing the PV cells and components, and fabricating them into finished panels. Other solar PV firms, NGOs and government programs involved with the Sri Lankan ASTAE World Bank Program purchased the modules at wholesale prices (Cabraal et al., 1996). High import duties on imported components eroded the
109 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
manufacturing cost advantages over imported modules, and manufacturing was abandoned (Huacuz & Gunarante, 2003).
4.5.9. Zambia The competitive tender process used in the Zambian ESCO project made it difficult for local equipment supply companies to compete, due to large warranties and the need to extend credit for long periods (Ellegard et al., 2004). The use of imported system components in Zambia made it difficult to keep ESCO fees down in conditions of high inflation. In addition, none of the imported batteries provided were according to specifications (Ellegard et al., 2004). In one ESCO, half of the batteries failed in less than a year, where it was suspected that batteries supplied by Siemens Zambia did not meet specifications. Siemens prepayment devices also failed in 50% cases. Siemens said the technology was proven at the outset of the project, but they had not been field tested.
4.5.10. Zimbabwe In Zimbabwe, since local manufacture was seen as vital for the sustainability of the PV industry, the Zimbabwae GEF Solar project attempted to promote local manufacture and build supply and service capacity. Under the project, standards were developed in the 1990s for installation and specification of PV systems and components. Manufacturing companies were able to take advantage of training via the project (Mulugetta et al., 2000). A PV laboratory was established for quality testing of imported components used in the GEF project. Two staff members were seconded to the Danish Technical Institute to train in solar laboratory technologies. Project participants needed to comply with standards in order to be eligible for compulsory registration for the project with the newly formed Solar Energy Industries Association of Zimbabwe (SEIAZ). They were then eligible for training programmes in manufacturing, installation and maintenance. However, existing smaller companies often were unable to qualify for the project and were therefore unable to benefit from the subsidies available, while unsustainable larger companies only operated as long as the subsidies existed. 60 companies were registered with SEIAZ in 1997, reducing to 30 in 1998 and by 2000, only 15 companies renewed their registration. Very little success was achieved in stimulating the design and manufacture sector. No new module company was formed due to lack of funds for equipment (Mulugetta et al., 2000). Importation of components became much more expensive in Zimbabwe as a result of the decline in the value of the Zimbabwe dollar since 1997. Materials imported for the local manufacture of components were also affected (Mulugetta et al., 2000). The intervention also suffered from opportunistic enterprises cashing in on market subsidies provided in parallel with the support, most of which stopped producing at the end of the project, when there was no
110 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
longer a financial incentive, thus leaving no-one responsible for maintenance and fulfilling warranties on systems in some cases (Mulugetta et al., 2000). A project was later carried out with the support of JICA, Japan to address some of the problems of the GEF project. Further training in testing technologies was received during the JICA Zimbabwe project from Japanese experts (Tani, 2003a). Selected charge controller and lamp manufacturers were supported to improve their designs and quality control through the project. One of the existing domestically manufactured charge controllers was initially found to have excessive internal power consumption. The design of the charge controller was improved to reduce power consumption. Further problems were encountered with the three companies that were eventually accepted as suppliers, including: High voltage set point too high (12.5V instead of 13.5), Dead insects, loose nuts and plastic shavings left inside the controller box, No covered contacts on relays, Long leads on transistors, Circuits not able to carry rated current without excessive heating and voltage drop. One of the companies did not have adequate DC power supply and equipment to perform adjustment of charge controllers and hence used an outside contractor, who soldered the components to the printed circuit board and delivered incomplete circuits (Tani, 2003a). In the end the manufacturers could not achieve the required quality in the project timeframe and the controllers were eventually replaced with a Japanese manufactured model (IEA, 2003a). Despite serious setbacks in stimulating local manufacture through these projects, in 2003, Zimbabwe had about 50 small PV companies engaged in import, sales, installation and manufacture of PV system components. Most of these firms had less than 20 employees and annual sales below US$10 000 (Tani, 2003a).
111 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.6. Barriers and Advantages of Small Scale Manufacture
In this section, the barriers and advantages of small scale PV manufacture in developing countries are identified. While successful deployment has occurred, the use of PV for rural electrification has also resulted in many failures (Cabraal et al., 1996; Martinot et al., 2000a; Wilkins, 2002). The success of local manufacture and commercialisation of PV system components and the satisfaction of social and business goals is strongly impacted by the way that the technology is delivered, including the financing, installation, after-sales service and training, and the linkages between actors in the value chain. A number of studies have summarised lessons learnt or key barriers to success of PV projects (Barnes & Floor, 1996; Cabraal et al., 1996; IEA, 2003a; Lorenzo, 1997; Lorenzo, 2000; Martinot et al., 2000a; Martinot et al., 2000b; Nieuwenhout et al., 2004; Nieuwenhout et al., 2000; van Campen et al., 2000; Wilkins, 2002). Many case studies of PV diffusion and use have also been published. In order to provide a context for the operation of small scale PV manufacturing enterprises, Appendix 2 reviews the literature and extracts the main barriers to the successful commercialisation of PV in developing countries, which fall within the following categories: High upfront cost, Low quality of the hardware, Poor technical capability in system design, installation, maintenance and use of the technology, Unsuccessful organisational structures in projects and programmes, Inadequate market institutions resulting in poor information flows and contract enforcement.
Local manufacture has the potential to mitigate problems in relation to some of these issues, whereas other issues represent particular barriers for small scale manufacturers. The implications of these characteristics of PV markets for small-scale manufacturers are now discussed.
4.6.1. High Upfront Costs Although PV is often the cheapest option for rural electrification over the life cycle of the system, the high upfront cost of the hardware is one of the major barriers to the diffusion of the technology in these markets. The cases reviewed in section 4.5 indicate that local manufacture is cheaper in many cases such as in the Bolivian altiplano project (Aguilera &
112 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
Lorenzo, 1996), Brazilian manufactured lamps (Zilles et al., 2000) and Kenyan batteries and lamps (Acker & Kammen, 1996). Local manufacturers in Kenya and China benefited from avoidance of transport costs, sales tax and import duties (Acker & Kammen, 1996; Hankins, 2000; Stone et al., 1998). However, local manufacturers usually need to use imported components, and in some cases become the sole importer of electronics in the country, such as in Bolivia (Aguilera & Lorenzo, 1996). They are therefore vulnerable to changes in the value of the local currency, which was problematic in Zimbabwe (Mulugetta et al., 2000) and import duties, which was problematic in Sri Lanka (Huacuz & Gunarante, 2003).
4.6.2. Quality of the Hardware Achieving sufficient quality has been a problem in many countries, and the issue of quality constitutes a particular barrier for local manufacturers, since it is often assumed that the quality of locally manufactured components will be lower (Ling et al., 2002; Stone et al., 1998), which is one of the main barriers to the support of local manufacture. Indeed, product has proved to be difficult to achieve in practice, with the documented quality of locally made BOS components mixed. The quality of local components was found to be initially inadequate in Indonesia (Fitriana et al., 1998) and in the case of Brazilian-made charge controllers (Zilles et al., 2000). Quality control was identified as particularly problematic for Zimbabwean manufacturers (Tani, 2003a). However, locally made products in Bolivia performed equally to those made in Spain (Aguilera & Lorenzo, 1996) and better reliability was recorded for locally manufactured lamps than imported ones in Brazil (Zilles et al., 2000). Locally made components in Kiribati also recorded low failure rates (Gillett & Wilkins, 1999). There are also numerous examples of poor quality imported components in rural electrification projects (Nieuwenhout et al., 2004; Wilkins, 2002). For instance, in Zambia, imported prepayment devices and batteries were found to be poor quality and 28% of imported lamps were found to have malfunctioned after a few years (Gustavsson & Ellegard, 2004). The quality of both locally made and imported modules has also been a problem (King et al., 2000). In 57% of the imported modules installed in Tonga, the EVA was found to have melted (Tukunga, 2002). Cracks in glass, water intrusion, dust and algal growth, as well as damaged junction boxes due to polarity reversal have also been found in modules in Thai battery charging stations (Green, 2004). Black market amorphous modules from South Africa and China have been found across Zimbabwe and Kenya to be poor quality (Duke et al., 2002; Jacobson & Kammen, 2005; Tani, 2003a). Locally made modules in Nicaragua were observed during field work to have poor fill factors, corrosion of solder joints and contacts (Bruce & Watt, 2006). In an effort to confirm whether locally manufactured components are in fact of lower quality on average than imported ones, 20 projects which had some analysis of failure modes
113 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
were reviewed, and where it could be confirmed that the components were either locally manufactured or imported, the failure rates were recorded (Table 4-3).
Table 4-3: Failure Modes in 20 PV Projects (number of project surveys that mention significant failure) Cause of Module Battery Charge Lights Other Total Locally Imported Problem Contr. H’dware Manuf. Misuse 2 3 1 0 0 6 3 2 Theft/Vandalism 3 1 0 0 0 4 0 0 Poor System Design 0 0 0 0 0 0 0 0 Poor Maintenance 0 1 0 0 2 3 0 0 Poor Quality 2 2 5 3 0 12 5 5
The projects reviewed and the references are listed below:
Project Name Region Reference Implementation Purchase Sao Paulo Brazil SHS (Zilles et al., 2000) 1 South America subsidised Thailand Battery Charging (PWD) (Green, 2004; Sriuthaisiriwong & Southeast Asia 2 with no fee or technician Kumar, 2001) Battery Charging Thailand Battery Charging (PWD) (Green, 2004; Sriuthaisiriwong & Southeast Asia 3 with fee and technician Kumar, 2001) Battery Charging 4 Thailand Battery Charging (DEDP) Southeast Asia (Green, 2004) Battery Charging 5 South Africa schools (RDP) Southern Africa (Klunne et al., 2001) Donated 6 South Africa schools (EU) Southern Africa (Klunne et al., 2001) Donated 7 South Africa clincs (IDT) Southern Africa (Klunne et al., 2001) Donated (Ellegard et al., 2004; Gustavsson Nyimba, Zambia ESCO Southern Africa 8 & Ellegard, 2004) Fee for Service (Tukunga, 2002; Tukunga et al., Tonga Lighting Pacific Islands 9 2002) Donated 10 Senegal Battery Charging West Africa (Seck, 2002) Fee for Service 11 Thailand Water Pumping Southeast Asia (Kaunmuang et al., 2001) Donated (Mulugetta et al., 2000; Tani, Zimbabwe GEF Project Southern Africa 12 2003a) Credit Purchase Indian India (West Bengal) – NREL (Stone et al., 1998) 13 Subcontinent Credit Purchase 14 Western China – NREL East Asia (Stone et al., 1998) Purchase 15 Rural Mexico Central America (Huacuz & Agredano, 1998) 16 Kiribati PV Pacific Islands (Tani, 2003b) Fee for Service 17 JICA Zimbabwae Southern Africa (Tani, 2003a) Fee for Service Indian Sub India Urjagram (Sharma, 2000) 18 Continent Donated Indian Sub India Assam (Sharma, 2000) 19 Continent 20 Brazil Pernabucco Pilot Projects South America Pernambucco Electricity Utility
In many of these cases, it is not clear what percentage of the components failed, or what constitutes ‘failed’ in each survey. The surveys were also conducted when the system components were a variety of ages. It is therefore not possible to draw any concrete conclusions about the quality of locally manufactured and imported components from this survey of the literature, but imported and locally made components were found to be of poor quality in an equal number of projects. There is no resounding message that locally manufactured components fail significantly more frequently than imported ones. However, the perception that small manufacturers are unable to produce products of sufficient quality remains likely to be a barrier for small enterprises.
114 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
Although small enterprises can potentially produce good quality products, achieving quality in product design and production is a problem for many local manufactures. Small local companies making fluorescent lamp inverters do not always have enough highly educated technicians available. Even young graduates in electronics prefer to work for large organisations and usually do not want to work in rural areas. As a result of lack of skilled personnel and isolation from other enterprises and organisations, many local manufacturers use old circuit diagrams, often based on poor designs (Vervaart & Nieuwenhout, 2000). Quality control is a particular problem, since its importance may be overlooked, it requires testing equipment and it is labour intensive. Once the initial cost barrier is overcome, however, quality control can make manufacturers more cost effective. In China, manufacturers who accessed assistance through the World Bank / GEF China Renewable Energy Development Program improved their product quality with no additional cost, some even decreasing costs (de Villers, 2005).
4.6.3. Local Manufacture and Local Technical Capabilities Inadequate local capabilities in system design, installation, maintenance and use have led to the premature failure of many PV systems in rural areas of developing countries (Cabraal et al., 1996; IEA, 2003b; Lorenzo, 2000; Martinot et al., 2000a; Wilkins, 2002). In particular, mismatched or inappropriate components and misuse or failure of the charge controller requires frequent replacement of batteries, leading to higher life cycle costs of systems. Good installation and maintenance depends on user education, the availability and capabilities of technicians and the availability of spare parts. The review of PV diffusion in Appendix 2 indicates that projects & programmes have tend to have inadequate links to local communities and therefore inadequate local availability of spare parts, skilled personnel and funding. The training of local technicians in projects is advocated, but a full time technician job relies on a minimum number of systems in a small geographical area. Rural technicians, once trained, may also find better employment in urban areas. After-sales service quality can be better in fee-for-service modes, since the utility has the incentive to ensure the energy service is available in order to get paid, rather than providing the hardware. Many systems are sold in cash-sales markets without any maintenance agreement or user training. The retailer often has very little capability in installation and maintenance, since it is common for PV to be one of the many products they sell. Since small enterprises are at risk of going out of business, users may be left without any after-sales service. Local manufacture may have a role to play in improving the local capabilities for installation, maintenance and use. Fitriana et al (1998) advocate the use of locally produced PV system components in Indonesia where possible, since spare parts for locally manufactured
115 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
components are more easily available within the country and components can be repaired within the country. The local manufacture of BOS components has resulted in better local availability of spare parts in Brazil, Kiribati & Tuvalu (Gillett & Wilkins, 1999; Zilles et al., 2000), and has resulted in products which are better suited to local conditions (Acker & Kammen, 1996). In the case of Kiribati and Tuvalu, the locally manufactured products were designed to be more easily repaired in the field (Gillett & Wilkins, 1999). In Kiribati and Tuvalu (Gillett & Wilkins, 1999) and in Bolivia (Aguilera & Lorenzo, 1996), technical capabilities for maintenance and installation were enhanced by local manufacturers, who interacted closely with technicians.
4.6.4. Local Manufacture and Project-Based Diffusion Projects and programmes constitute around a third of the total market for PV in developing countries (Nieuwenhout et al., 2001). They often have a demonstration objective, or are designed to improve rural electrification in a geographically or socially targeted manner. Projects can more easily guarantee quality of PV systems through centralised procurement and testing of components, but often result in more expensive and imported hardware, money spent on the wages of foreign consultants and administrators, and a limited number of designs approved by the project developers, which may not suit the preferences of the users. Small enterprises face high barriers to accessing projects and programmes, due to prohibitively strict standards, complex or costly administrative procedures, fiscal procedures, the complexity of tendering processes, or contracts that are too large. Small firms may not have the capability to achieve high technical standards, or the financial or administrative capabilities to comply with procedures. For example, In an ESCO pilot project in Zambia, local suppliers could not participate due to the long guarantees and credit periods required (Gustavsson & Ellegard, 2004). In Zimbabwe, small manufacturers were unable to meet the quality requirements (Mulugetta et al., 2000). Favouritism for known brands or influential suppliers may also be a barrier to the selection of local enterprises for projects or programmes. Large-scale schemes may exclude the many small, financially weak, firms that have local knowledge and commitment and often end up delivering the services on behalf of larger suppliers. IEA’s (2003b) report on capacity building recommends that regulatory frameworks need to consider barriers to new entrants and unfair competition. Projects inherently have a limited life. There has been little success in establishing self- sustaining markets after projects, even where the market is not saturated (de Villers, 2005). Uncertainties in PV markets resulting from projects coming and going impede investment in labour and equipment. Enterprises need to recruit temporary staff for projects and may find it difficult to develop a skilled trained workforce without a permanent team. Many small
116 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
enterprises are unsustainable because they are not able to cope with vulnerability and will often need to have other business activities to sustain them when PV demand is low. While projects may promote the use of good quality hardware, systems installed through projects often fail in the medium or long term after the support for installation and maintenance is removed. The instability of markets, subsidies and technical support has detrimental effects on the reputation of the technology. Customers may also wait for subsidies to appear before making a purchase. In some cases, these subsidies may not materialise and the cash market is damaged.
4.6.5. Local Manufacture and Diffusion via Cash- Markets While projects are unstable and usually have poor linkages with local communities, they often provide financing and the training to build capabilities locally. The provision of finance to the poor users of PV systems and to small enterprises involved in retailing, servicing or manufacturing systems is likely to be inadequate without the intervention of projects, because of the high risks and transaction costs of providing credit to poor and dispersed customers with little ability to make repayments. In cash markets, small manufacturers generally rely on their own resources to provide products at an appropriate quality and cost and may therefore be unfeasible. Small scale manufacturers and retailers also face market barriers related to preferences for brand names or imported goods. Unfair competition in the form of dumping by larger suppliers, both national and international has been reported (de Villers et al., 2004).
4.6.6. Non-Technical Capabilities of Small Scale Manufacturers Poor financial management of the dissemination of locally manufactured PV components has caused problems in Bolivia (Huacuz & Gunarante, 2003) and Tuvalu (Gillett & Wilkins, 1999). Attempts at small scale module manufacture in Nepal and Zimbabwe have failed due to inadequate infrastructure (power supply) and insufficient financial capital respectively. Complementary capabilities in business skills and financial management are clearly required by small scale manufacturers, but there is very little information available in the literature in relation to these requirements.
4.6.7. Supporting Small Scale PV Manufacture Programmes have attempted to improve the quality of local manufacture through the use of standards, the establishment of local testing facilities and support for compliance with standards in Nepal, China, Zimbabwe and Indonesia (Katic, 2002; Li, 2003; Mulugetta et al.,
117 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
2000; Nieuwenhout et al., 2004; Shestha et al., 2000). Certification and improved components improved the confidence of Chinese consumers in PV systems and increased the willingness to pay for higher prices, better quality products (Li, 2003). While the introduction of standards provides incentives for manufacturers to improve, in many cases, such as in Zimbabwe, small manufacturers have not been able to qualify for these programmes, resulting in their collapse and encouraging unsustainable larger companies to enter the market (Mulugetta et al., 2000). The absence of a sufficiently large market in Indonesia prevented manufacturers from realising their improved designs (Nieuwenhout et al., 2004), while project timeframes in Zimbabwe were too short for manufacturers to improve their designs sufficiently (Tani, 2003a). In Indonesia, products which passed testing were sometimes not robust and did not perform well over the long term in the field (Nieuwenhout et al., 2004). In order to nurture local manufacturing, the introduction of standards should be combined with capability building, financial assistance and design assistance and should involve consultation with local manufacturers. It should also be recognised that time and iterations are needed to improve the design and production capabilities of small scale manufacturers. South-south technology transfer has proved effective in transferring capabilities from Nepal to Laos and Bangladesh (Shestha et al., 2000) and this potential could be further explored. It has been suggested that local standards and testing laboratories may make compliance more affordable than international standards and allow standards to be increased incrementally via the introduction of interim standards (Nieuwenhout et al., 2004). In addition, sales tax and import tariffs on parts and materials required by manufacturers should be avoided. In general, attention should also be given to the impacts of PV project design on small enterprises. Long term electrification programmes and smart subsidies that can be removed or targeted are preferable to short term projects.
118 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
4.7. Conclusion
This chapter has identified roles for small scale PV manufacture in poverty reduction and in improving the use of the technology locally. Small manufacturing enterprises can contribute to local employment and allow people to diversify their income away from agriculture. The small scale manufacture of BOS in developing countries also has the potential to reduce leakages from the local economy in the form of payments for energy services, save foreign currency and possibly develop into an export industry. Small enterprises may use local materials, and locally manufactured PV products may be better adapted to serve the local market than imported ones. Although there is only a small amount of literature documenting experiences with small scale manufacture, there is evidence to suggest that the local manufacture of PV system components can potentially address some of the barriers to sustainable PV rural electrification via improved local technical knowledge and skills, availability of spare parts, information flows and potentially lower cost manufacture and maintenance (Acker & Kammen, 1996; Gillett & Wilkins, 1999; Green, 2004; Mulugetta et al., 2000). It should be noted that local maintenance capabilities have been improved only where there have been good linkages between technicians and manufacturers. Designs that are more easily repairable have also facilitated better local maintenance. However, small enterprises in the least developed countries operate with a scarcity of resources and capabilities and the assembly of balance of systems components is likely to be at the limit of their capabilities. The quality of locally made BOS has been found to be inadequate in many cases. The evidence available suggests that small manufacturers tend to have inadequate technical capabilities, particularly in quality control and product design and inadequate complementary business capabilities. They are also likely to be particularly affected by the small and dispersed markets for PV, market fluctuations, and lack of access to markets, especially projects and programmes which may have challenging quality, warranty, credit and tendering requirements. Support for local manufacture has been primarily related to the introduction of standards and the support of manufacturers in compliance with them. Although some manufacturers have been able to improve their products and expand their markets through these programmes, small manufacturers have been disadvantaged by these programmes in many cases and issues related to market size, project timeframes and lack of complementary capabilities have prevented them from benefiting from these interventions. Because of the paucity of information available on the small scale manufacture of PV components in developing countries, an incomplete picture of the factors which impact
119 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
capability building has been established through this review. There are almost certainly a number of other factors that impact small PV manufacturers, such as low levels of financial and physical capital, isolation from value chains and poor infrastructure. There is clearly a need for a better understanding of the determinants of capability building in small scale PV manufacturers in developing countries. The following chapter develops an analytical framework with reference to what is known about capability building, including literature specific to small enterprises. The framework will be used to study capability building in three case studies, including two small scale manufacturers, and to therefore suggest appropriate capability building strategies, identify the characteristics of enabling environments and suggest suitable interventions to assist manufacturers.
120 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
References
Acker, R.H. and Kammen, D.M. (1996), The quiet (energy) revolution : Analysing the dissemination of photovoltaic power systems in Kenya, Energy Policy, 24 (1), p 81. Aftab, K. and Rahim, E. (1987), The Emergence of a Small-Scale Engineering Sector: The Case of Tubewell Production in the Pakistan Punjab, Journal of Development Studies, 23 (1), pp 60-76. Aftab, K. and Rahim, E. (1989), 'Barriers' to the Growth of Informal Sector Firms: A Case Study, Journal of Development Studies, 25 (4), p 490. Aguilera, J. and Lorenzo, E. (1996), Rural photovoltaic electrification programme on the Bolivian high plateau, Progress in Photovoltaics: Research and Applications, 4 (1), pp 77 - 84. Albu, M. (2001), International Strategy for ITDG’s work with Manufacturing Enterprise: Improving Livelihood Opportunities from Small-scale Manufacturing and Processing Enterprises, Intermediate Technology Development Group. Barnes, D.F. and Floor, W.M. (1996), Rural Energy in Developing Countries: A Challenge for Economic Development, Annual Review of Energy and the Environment, 21 (497-530). Barnett, A. (2001), Looking at household energy provision in a new way: The Sustainable Livelihoods approach, Boiling Point (46), pp 30-32. Barnett, A. (2005), Sustainable Livelihoods, Energy and Poverty, Access (EASE Newsletter), 2, pp 3-5. Bruce, A. and Watt, M. (2006), Small-Scale Manufacture of PV System Components in Developing Countries: Case Studies in India and Nicaragua, Australian & New Zealand Solar Energy Society Conference 2006:Clean Energy, Can Do!, Canberra, Australia, 13-15 September 2006. Cabraal, A. (2000), Building on Experience: Assuring Quality in the World Bank/GEF-Assisted China Renewable Energy Development Project, 16th European Photovoltaic Solar Energy Conference, Glasgow, U.K. Cabraal, A. (2004), Improving Quality and Building Capacity for Private Provision of Rural Energy Services: The China Renewable Energy Development Project, The World Bank. Cabraal, A., Cosgrove-Davies, M. and Schaeffer, L. (1998), Accelerating sustainable photovoltaic market development, Progress in Photovoltaics: Research and Applications, 6 (5), pp 297-306. Cabraal, A., Davies, M.C. and Schaeffer., L. (1996), Best Practices for Photovoltaic Household Rural Electrification Programs - Lessons from Experiences in Selected Countries, ASTAE World Bank. Cecelski, E. (2003), Enabling Equitable Access to Rural Electrification: Current Thinking on Energy, Poverty, and Gender, EnPoGen, The World Bank, Washington D.C. Corkish, R., Bruce, A., Bocking, J., Linney, M. and Sproul, A.B. (2004), Small Photovoltaic Systems in Rural Nicaragua, ANZSES Solar 2004, Perth, Western Australia. Czanderna, A.W. and Pern, F.J. (1996), Encapsulation of PV modules using ethylene vinyl acetate copolymer as a pottant: A critical review, Solar Energy Materials and Solar Cells, 43 (2), p 101. Davis, J.R. (2004), The Rural Non-Farm Economy, Livelihoods and their Diversification: Issues and Options, Natural Resources Institute, Chatham, U.K. de Villers, T. (2005), Strategies to increase the confidence in PV and expected impact on the PV Market, Tackling the Quality in Solar Rural Electrification, TaQSolRE. de Villers, T., Watchueng, S., Shanker, A. and Rambaud-Méasson, D. (2004), Successful PV SHS project in developing countries? barriers and way foreword, 19th European Photovoltaic Solar Energy Conference, Paris, France, 7-11 June 2004. Dorward, A. (2001), Pro-Poor Livelihoods: Addressing the Market/ Private Sector Gap, Sustainable Livelihoods Seminar on ‘Private Sector and Enterprise Development’, Manchester, U.K., 19th November 2001.
121 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
Dorward, A., Poole, N., Morrison, J., Kydd, J. and Urey, I. (2003), Markets, Institutions and Technology: Missing Links in Livelihoods Analysis, Development Policy Review 21 (3), pp 319-332. Duke, R.D., Jacobson, A. and Kammen, D.M. (2002), Photovoltaic module quality in the Kenyan solar home systems market, Energy Policy, 30 (6), p 477. Dunlop, J.P. (1997), Batteries and Charge Control in Stand-Alone Photovoltaic Systems: Fundamentals and Application, Sandia National Laboratories, Albuquerque, NM, USA. Dunnett, S. (2001), Entrepreneurs by choice? – or by necessity?, Boiling Point (47), pp 1-2. EDRC (2003), A review of international literature of ESCOs and fee-for-service approaches to rural electrification (solar home systems). Energy and Development Research Centre, University of Cape Town. Ellegard, A., Arvidson, A., Nordstrom, M., Kalumiana, O.S. and Mwanza, C. (2004), Rural people pay for solar: experiences from the Zambia PV-ESCO project, Renewable Energy, 29 (8), p 1251. Fernando, S. (1992), Chapter 3: Technology Transfer for Small and Cottage Industries: The Sri Lankan Experience, in Islam, R. (ed), "Transfer, Adoption and Diffusion of Technology for Small and Cottage Industries", Asian Regional Team for Employment Promotion, World Employment Programme, ILO, New Delhi, India, pp 74-108. Fitriana, I., Kantosa, E., Sudradjat, A., Kuhmann, J., Preiser, K. and Schweizer-Ries, P. (1998), On the Way from Sukatani to the 50 MW Programme - A Socio-Technical Analysis of Solar Home Systems in Indonesia, IEEE 2nd World Conference on Photovoltaic Solar Energy Conversion, Vienna, Austria, 1998. Foley, G. (1990), Electricity for Rural People, Panos, London. Gillett, B. and Wilkins, G. (1999), Evaluation of the PREP Component : PV Systems for Rural Electrification in Kiribati & Tuvalu, Final report for the European Commission DGVIII Development, AEA Technology Environment (ETSU). Green, D. (2004), Thailand's solar white elephants: an analysis of 15 yr of solar battery charging programmes in northern Thailand, Energy Policy, 32 (6), p 747. Gustavsson, M. and Ellegard, A. (2004), The impact of solar home systems on rural livelihoods. Experiences from the Nyimba Energy Service Company in Zambia, Renewable Energy, 29 (7), p 1059. Hankins, M. (2000), A case study on private provision of photovoltaic systems in Kenya, Energy and Development Report: Energy Services for the World's Poor, World Bank Energy Sector Management Assistance Programme (ESMAP), U.S.A. Hirshman, W.P., Hering, G. and Schmela, M. (2007), Gigawatts - the measure of things to come: Market survey on global solar cell and module production in 2006, Photon International (March 2007), pp 136-166. Honsberg, C. and Bowden, S. (1999), Photovoltaics: Devices, Systems and Applications CD ROM – Volume 1: Devices, UNSW Centre for Photovoltaics, Sydney, Australia. Huacuz, J.M. and Agredano, J. (1998), Beyond the grid: photovoltaic electrification in rural Mexico, Progress in Photovoltaics: Research and Applications, 6 (5), pp 379 - 395. Huacuz, J.M. and Gunarante, L. (2003), Photovoltaics and Development, in Luque, A. & Hegedus, S.S. (eds), "Handbook of Photovoltaic Science and Engineering", John Wiley & Sons. IDS (2003), Energy, Poverty and Gender: a Review of the Evidence and Case Studies in Rural China, Report for the World Bank EnPoGen Project, Institute of Development Studies. IEA (2003a), 16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries, Deployment of Photovoltaic Technologies: Co-operation with Developing Countries, IEA PVPS Task 9. IEA (2003b), PV for Rural Electrification in Developing Countries – A Guide to Capacity Building Requirements, Deployment of Photovoltaic Technologies: Co-operation with Developing Countries, IEA PVPS Task 9. Islam, R. (1992), Chapter 1: Transfer, Dissemination and Adoption of Technology for Small and Cottage Industries: An Overview, in Islam, R. (ed), "Transfer, Dissemination and
122 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
Adoption of Technology for Small and Cottage Industries", Asian Regional Team for Employment Promotion, World Employment Programme, ILO, New Delhi, India. Jacobson, A. and Kammen, D.M. (2005), The Value of Vigilance: Evaluating Product Quality in the Kenyan Solar Photovoltaics Industry, Report delivered to solar industry stakeholders in East Africa, July 2, 2005. James, J. (1989), Improving traditional rural technologies, Macmillan, Basingstoke. Katic, I. (2002), Quality Assurance of Solar Home Systems in Nepal, 18th European Photovoltaic Solar Energy Conference, Rome, 7 - 11 October 2002. Kaunmuang, P., Kirtikara, K., Songprakorb, R., Thepa, S. and Suwannakum, T. (2001), Assessment of photovoltaic pumping systems in Thailand - one decade experience, Solar Energy Materials and Solar Cells, 67 (1-4), p 529. Khosla, A. (1994), Effective Design and Diffusion of Rural Technologies, in Bhalla, A.S. & Reddy, A.K.N. (eds), "The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme", Intermediate Technology Pub., London, p 44. King, D.E., Pern, F.J., Pitts, J.R., Bingham, C.E. and Czanderna, A.W. (1997), Optical Changes in Cerium-Containing Glass as a Result of Accelerated Exposure Testing, 26th IEEE Photovoltaic Specialists Conference, Anaheim, California, U.S.A., 29 September - 3 October 1997. King, D.L., Quintana, M.A., Kratochvil, J.A., Ellibee, D.E. and Hansen, B.R. (2000), Photovoltaic module performance and durability following long-term field exposure, Progress in Photovoltaics: Research and Applications, 8 (2), pp 241-256. Klunne, W.E., Westra, C.A., Cox, R.L.M. and Gys, L.J.C. (2001), Improving the Avialabilty of PV-Systems at South African Schools and Clincs, 17th European Photovoltaic Solar Energy Conference, Munich, Germany, October 22-26, 2001. Krishnaswamy, K.N. and Reddy, A.K.N. (1994), The Commercialization of Improved Technologies in Rural Areas, in Bhalla, A.S. & Reddy, A.K.N. (eds), "The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme", Intermediate Technology Pub., London, p 23. Kumar, S., S.C.Bhattacharya and Leon, M.A. (2000), A Status Survey on PV Systems and Accessories in Asia, World Renewable Energy Congress, Brighton, U.K., July 2000. Li, Z. (2003), Global Environment Facility (GEF) / World Bank Assisted China Renewable Energy Development Project PV Component, in IEA (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA PVPS Task 9 Deployment of Photovoltaic Technologies: Co-operation with Developing Countries. Li, Z. (2004), Made In China, Renewable Energy World, 7 (1), pp 70-80. Ling, S., Twidell, J. and Boardman, B. (2002), Household photovoltaic market in Xining, Qinghai province, China: the role of local PV business, Solar Energy, 73 (4), p 227. Lorenzo, E. (1997), Photovoltaic Rural Electrification, Progress in Photovoltaics: Research and Applications, 5 (1), pp 3-27. Lorenzo, E. (2000), In the field - Realities of some PV rural electrification projects, Renewable Energy World (Sept-Oct 2000). Martinot, E., Cabraal, A. and Mathur, S. (2000a), World Bank/GEF Solar Home Systems Projects: Experiences and Lessons Learned 1993-2000, World Bank/GEF. Martinot, E., Ramankutty, R. and Rittner, F. (2000b), The GEF Solar PV Portfolio: Emerging Experience and Lessons, Monitoring and Evaluation Working Paper 2, Global Environment Facility. Masum, M. (1992), Chapter 2: The Bangladesh Experience, in Islam, R. (ed), "Transfer, Dissemination and Adoption of Technology for Small and Cottage Industries", Asian Regional Team for Employment Promotion, World Employment Programme, ILO, New Delhi, India. Medellin-Erdmann, R. (1992), AT and New World Order, Appropriate Technology in Post- Modern Times, Frankfurt, Germany.
123 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
Moner-Girona, M., Ghanadan, R., Jacobson, A. and Kammen, D.M. (2006), Decreasing PV costs in Africa: Opportunities for Rural Electrification using Solar PV in Sub-Saharan Africa, Refocus, 7 (1), pp 40-45. Mulugetta, Y., Nhete, T. and Jackson, T. (2000), Photovoltaics in Zimbabwe: lessons from the GEF Solar project, Energy Policy, 28 (14), p 1069. Nieuwenhout, F., de Villers, T., Mate, N. and Aguilera, M.E. (2004), Reliability of PV stand- alone systems for rural electrification, Part 1: Literature Findings, Tackling the Quality in Solar Rural Electrification, TaQSolRE. Nieuwenhout, F.D.J., van Dijk, A., Lasschuit, P.E., van Roekel, G., van Dijk, V.A.P., Hirsch, D., Arriaza, H., Hankins, M., Sharma, B.D. and Wade, H. (2001), Experience with solar home systems in developing countries: a review, Progress in Photovoltaics: Research and Applications, 9 (6), pp 455-474. Nieuwenhout, F.D.J., van Dijk, A., van Dijk, V.A.P., Hirsch, D., Lasschuit, P.E., van Roekel, G., Arriaza, H., Hankins, M., Sharma, B.D. and Wade, H. (2000), Monitoring and Evaluation of Solar Home Systems - Experiences with applications of solar PV for households in developing countries, Netherlands Energy Research Foundation ECN and Department of Science, Technology and Society of Utrecht University. Overseas Development Institute (2002), Non-Farm Income in Rural Areas, Keysheets for Sustainable Livelihoods (Last Updated: October 2002), Accessed from: http://www.keysheets.org/, on: August 2005. Panditrao, Y.A. (1994), Experiences of the Khadi and Village Industries Comission in Technology Transfer, in Bhalla, A.S. & Reddy, A.K.N. (eds), "The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme", Intermediate Technology Pub., London, p 113. Pern, F.J. and Glick, S.H. (2000), Photothermal stability of encapsulated Si solar cells and encapsulation materials upon accelerated exposures, Solar Energy Materials and Solar Cells, 61 (2), p 153. Ramani, K.V. and Heijndermans, E. (2003), Energy, Poverty and Gender: A Synthesis, EnPoGen Project ASTAE, The World Bank. Sandia (1999), Volume 3 - Module Durability, Quarterly Highlights of Sandia’s Photovoltaics Program, Sandia National Laboratories. Schumacher, E.F. (1973), Small is beautiful: a study of economics as if people mattered, Blond and Briggs, London,. Scoones, I. (1998), Sustainable Rural Livelihoods: A Framework for Analysis, IDS Working Paper 72, Institute for Development Studies, Brighton, U.K. Seck, L. (2002), Charging up: Senegal, UNDP Special Unit for SSC pp 89-95. Sharma, B.D. (2000), Annex 4. Experiences in India, in Nieuwenhout, F.D.J. (ed), "Monitoring and Evaluation of Solar Home Systems - Experiences with applications of solar PV for households in developing countries", Netherlands Energy Research Foundation ECN and Department of Science, Technology and Society of Utrecht University. Shestha, J.N., Sharma, D. and Adhikari, R.B. (2000), Application of Photovoltaic Technology in Nepal: An Assessment, World Renewable Energy Congress, Brighton, U.K., July 2000. Sriuthaisiriwong, Y. and Kumar, S. (2001), Rural electrification using photovoltaic battery charging stations: a performance study in northern Thailand, Progress in Photovoltaics: Research and Applications, 9 (3), pp 223 - 234. Stone, J.L., Tsuo, Y.S., Ullal, H.S., Wallace, W.L., Sastry, E.V.R. and Baoshan, L. (1998), PV Electrification in India and China: The NREL's Experience in International Cooperation, Progress in Photovoltaics: Research and Applications, 6 (5), pp 341-356. Tani, T. (2003a), JICA PV Project - Case of Zimbabwae, in IEA PVPS Task 9 (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA. Tani, T. (2003b), PV Rural Electrification in Kiribati, in IEA PVPS Task 9 (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA.
124 Chapter 4. The PV Industry and Small Scale Manufacture in Developing Countries
ter Horst, E. and Zhang, C. (2005), Impacts of Technology Improvement and Quality Assurance in the WB/GEF China Renewable Energy Development Project on PV industry and market development in China, 15th PVSEC. Todaro, M.P. (1982), Economics for a developing world : an introduction to principles, problems and policies for development, 2nd ed, Longman, Harlow, Essex. Tukunga, T. (2002), The Sustainability of Stand-Alone Photovoltaic Lighting Systems in Tonga, ME (Electrical) Thesis, School of Electrical Engineering and Telecommunications, The University of New South Wales, Sydney. Tukunga, T., Healy, S. and Outhred, H. (2002), Experience with PV Lighting Systems in Tonga, Solar 2002 - Australian and New Zealand Solar Energy Society, 2002. UN (2005), Investing in Development: A Practical Plan to Achieve the Millennium Development Goals., UN Millennium Project report to the UN Secretary-General, New York. UNDP (1999), Human Development Report: Globalization with a Human Face, United Nations Development Programme. UNDP (2001), Human Development Report: Making new technologies work for human development, United Nations Development Programme. UNEP (2003), Open for Business Entrepreneurs, Clean Energy and Sustainable Development, United Nations Environment Programme. Usher, E. and Ross, M.P. (1998), Recommended practices for charge controllers, IEA PVPS Task III. van Campen, B., Guidi, D. and Best, G. (2000), Solar photovoltaics for sustainable agriculture and rural development, Environment and Natural Resources Working Paper No 2, FAO, Rome. Varadi, P.F., Domínguez, R. and Schaeffer, L. (2003), Quality Management in Photovoltaics:Quality Control Training Manual for Manufacturers, Quality Program for Photovoltaics (QuaP-PV), ASTAE PV GAP. Vervaart, M.R. and Nieuwenhout, F.D.J. (2000), Solar Home Systems: Manual for the Design and Modification of Solar Home System Components Quality Program for Photovoltaics (QuaP-PV), ECN. Wade, H. (2003), Summary of PV Rural Electrification Experiences in the Pacific Islands, in IEA PVPS Task 9 (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA. Wenham, S., Green, M., Watt, M. and Corkish, R. (2006), Applied Photovoltaics - Second Edition, UNSW Centre for Photovoltaic Engineering, Sydney, Australia. Wilkins, G. (2002), Technology Transfer for Renewable Energy: Overcoming Barriers in Developing Countries, Earthscan Publications, London. Wilkinson, R. (2002), The impact of energy infrastructure projects on poverty, Boiling Point (48), pp 16-19. Wohlgemuth, J.H. (2003), Long Term Photovoltaic Module Reliability, NCPV and Solar Program Review Meeting, Denver, CO, U.S.A. Zilles, R., Lorenzo, E. and Serpa, P. (2000), From candles to PV electricity: a four-year experience at Iguape-Cananéia, Brazil, Progress in Photovoltaics: Research and Applications, 8 (4), pp 421-434.
125 126 CChhaapptteerr 55.. AA FFrraammeewwoorrkk ffoorr tthhee AAnnaallyyssiiss ooff CCaappaabbiilliittyy BBuuiillddiinngg iinn DDeevveellooppiinngg CCoouunnttrriieess
In the preceding three chapters, roles, barriers and advantages for both PV cell manufacturers and small scale PV BOS manufacturers in developing countries have been established. Despite the potential benefits of both small-scale and modern sector PV manufacture in developing countries, there has been very little published about, and there is therefore an inadequate understanding of, the critical factors influencing capability building in PV manufacturing enterprises. In this chapter, a framework is therefore developed to facilitate the analysis of capability building in several case studies, with a view to exploring its wider application to assisting decision making related to supporting PV manufacture in developing countries. Section 5.1 of this chapter defines knowledge and technology and discusses the characteristics of knowledge that impact learning and technology transfer. A typology of capabilities is presented in section 5.2, which will facilitate the qualitative identification of technological capabilities in enterprises. In section 5.3, the technological learning literature is introduced. A variety of learning processes by which enterprises build different types of capabilities are identified in section 5.4. The capability building literature, which is specifically concerned with learning in enterprises in developing countries, is reviewed in section 5.5, and the literature specific to capability building in small scale enterprises is reviewed in section 5.6, providing some background on the importance of different types of capabilities and capability building strategies, and the barriers to each. The technological systems approach, which identifies the components and interactions within technology systems that impact technological learning, is described in section 5.7. On the basis of this discussion, a framework for the analysis of capability building is proposed in section 5.8, which links the system-level factors to the capabilities and types of learning previously identified. In the interests of clarity, the framework is presented here as though it was developed linearly. In reality, its development was iterative and influenced by the process of data collection and analysis in the case studies as well as the paths that were followed in reviewing the PV, learning and technological systems literature. The framework emerged progressively and in no particular order, with different parts developing at different rates at different times, while the choices that have been made in selecting and combining parts of the literature have been influenced along the way by the experience gained in the case studies.
127 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.1. Knowledge and Technology
In this section, knowledge, innovation and technology are defined for the purposes of this study. Some aspects of technological knowledge and knowledge creation that influence technology acquisition and use are discussed.
5.1.1. Knowledge Knowledge is described by Nonaka (1994) as ‘justified true belief’. Although the terms ‘information’ and ‘knowledge’ are often used interchangeably, information is defined as data that needs to be internalised and interpreted in the context of what is already known by an individual before it becomes knowledge (ibid). Strictly speaking, skills refer to abilities acquired through training, often in relation to physical dexterity or otherwise in executing a task, whereas knowledge refers to understanding resulting from reasoning. The term ‘knowledge’, however, is widely used in innovation and learning literature to refer to skills, knowledge and expertise (Kamp, 2002). It is in this broad sense that it will be used in this thesis. In order to distinguish the types of knowledge that are significant in the use of technology, technological knowledge has been categorised as ‘know-how’, ‘know-why’ and ‘know-what’. Garud (1997) describes know-why as the understanding of the principles that explain why something works the way it does, know-how as the knowledge about how to do something in an efficient way, and know-what as the knowledge of the uses of a technology. Although know-how enables one to get things done, other types of knowledge are required to solve problems and intentionally acquire or generate new knowledge.
5.1.2. Tacit and Explicit Knowledge Explicit knowledge is knowledge which can be codified: written down or otherwise recorded and can be transmitted to another. Once codified, it can be accessed through books, technical specifications, designs and material embodied in machines. Tacit knowledge, in contrast is difficult to communicate or codify, and can more easily be expressed in actions, especially in a specific context (Kim, 1997). In Polanyi’s influential (1967) treatise, tacit knowledge was distinguished from explicit knowledge by its subconscious nature. Polanyi demonstrated the existence of peripheral awareness of things that cannot be explicitly identified by using the example of the ability to identify a particular human face without being able to explain why. There are a number of important implications arising from the tacitness of knowledge. First, the codified knowledge held by an enterprise, such as design blueprints, process instruction and standard operating procedures, requires accompanying tacit knowledge before it
128 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
is useful (Kim, 1997; Marcotte & Niosi, 2000; Nelson & Winter, 1982). Second, tacit knowledge is viewed as being difficult to share and hence belonging to individuals (Marcotte & Niosi, 2000; Nonaka, 1994). The knowledge residing in individuals must therefore be diffused throughout an organisation, and knowledge may be lost when individuals leave the organisation, or transferred when they move organisations. Third, the existence of a tacit component to knowledge implies difficulty in technology transfer. The learning process requires face-to-face communication of those aspects that cannot be codified. Although there may be difficulties in interpreting codified information (e.g. language difficulties or differences in codification practices), cultural and communication difficulties are magnified in person-to-person learning (Cowan et al., 2000). The idea that knowledge contains tacit components that are complementary to its explicit components, and therefore that the spread of knowledge does not take place solely by transmission of information, gained authority with studies in the 1970s on the construction of lasers in laboratories, where someone with experience of laser construction was present in every case where a laser was successfully built (Cowan et al., 2000).
5.1.3. Knowledge Creation As a result of the tacit nature of knowledge, and the need for face-to-face interaction for its effective communication, knowledge is seen to be created and reside in social situations. Nonaka (1994) introduces the idea of a spiral of organisational knowledge creation. He identifies the processes of transformation of knowledge from codified to tacit, through internalisation of codified knowledge to tacit knowledge; and tacit to codified, through externalisation of tacit knowledge to codified knowledge. These processes involve reinterpretation and expansion of knowledge. Socialisation processes can convey tacit knowledge from one individual to another individual (via training, for example), and combination processes integrate different pieces of codified knowledge. These processes allow people to extend and reframe their own tacit knowledge and lead to an increase of knowledge at an organisational level that accelerates as more people become involved in knowledge conversions. Although tacit knowledge has come to mean knowledge that cannot be codified, Cowan et al. (2000) point out that this distinction between tacit and codified knowledge is an oversimplification, and that tacit knowledge may be articulated if the incentives were high enough. Nelson and Winter (1982) refer to Polanyi’s example of a swimmer remaining buoyant by retaining air in his lungs, but not being conscious of doing so. The principle of buoyancy could be recognized and articulated to another, but in this case remains tacit. Nelson and Winter note that the tacitness is not inherent to the knowledge and that incentives to articulate it are imperative:
129 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
“…costs matter. Whether a particular bit of knowledge is in principle articulable or necessarily tacit is not the relevant question in most behavioral situations. Rather, the question is whether the costs … are sufficiently high so that the knowledge in fact remains tacit.” (Nelson & Winter, 1982, p 80)
The nature of knowledge, and the accumulation of knowledge, individually and collectively, has important implications for the way enterprises acquire technological capabilities.
5.1.4. Innovation Schumpeter (1961 {1911}) described invention as the development of a new technical idea, innovation as the successful introduction of a new or improved product to the market, and diffusion as the process by which improvements become widely used. Innovations are more broadly defined by the innovations systems writer Edquist (1997, p 1) as ‘new creations of economic significance’, which may be of various kinds, including technological, organizational or institutional. Although the term innovation is often associated with enterprises at the technological frontier of advanced nations, it is used in the literature about latecomers to refer also to improvements within enterprises that may not stretch the technological frontier, including the introduction of new processes or products to an enterprise. Within this thesis, I will use the term ‘innovation’ to refer to new technical, organisational or institutional creations and ‘learning’ to refer to the acquisition of technology more broadly.
5.1.5. Technology Technology has been called “the skills, knowledge and procedures for making, using and doing useful things” (Stewart, 1978, p 1). Technology is widely recognised as having the following four major interrelated parts that must be brought together in order for technology to be an effective asset (Kim & Nelson, 2000; Sharif, 1992): Equipment-embodied (machines and equipment), Human-embodied (knowledge and skills), Organisation-embodied (processes, systems, procedures, relationships) Document-embodied (codified knowledge, design, specification) While equipment and document embodied technology may be purchased, human- embodied technology is more difficult to transfer because of the tacit nature of much of this knowledge. Organisation-embodied technology, such as routines and relationships are also tacit in nature and may depend on each other and the local context, therefore being difficult to transfer.
130 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.2. Technological Capabilities
In this section, a general typology of technological capabilities is developed. The term ‘technological capabilities’ has arisen from a body of research which has attempted to identify the processes by which enterprises in developing countries have become competitive in international markets. There has been particular interest in the east-Asian Tigers (South Korea, Taiwan, Hong Kong and Singapore), as they have maintained high growth rates and rapid industrialization between the early 1960s and 1990s. Technological capabilities authors have focused not on simple technology transfer, which tends to be hardware-focused, but have been primarily concerned with human-embodied knowledge-related aspects of technology, and the ability to use that knowledge, which has an organisational dimension. Kim (1997, p 4) defines technological capability as “the ability to make effective use of technological knowledge in efforts to assimilate, use, adapt, and change existing technologies. It also enables one to create new technologies, and develop new products and processes”. Since technology has been defined here inclusively of the human-embodied knowledge and skills as well as organisational institutions, documents and equipment which comprise the technology; technological capabilities can be understood as abilities which result from the possession and application of an appropriate combination of the four aspects of technology. An enterprise may possess technology that does not translate into capability if the complementary aspects of technology are not available. For example, an enterprise may have ample technological knowledge, but may not have the equipment to effectively employ that knowledge. Within this study, therefore, the term technological capabilities is defined as the abilities to acquire, use, adapt and build on technology, which arise through the possession of the appropriate combination of human-embodied knowledge and skills, documented knowledge, organisational arrangements and physical equipment that embody a technology.
5.2.1. Types of Technological Capabilities Kim (1997) identified three elements of technological capabilities (Table 5-1): Production capabilities, which include the management, monitoring and optimization of production, and the repair and maintenance of equipment; Investment capabilities, which include training, feasibility studies and activities related to the establishment or expansion of production facilities including engineering design and procurement.
131 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Innovation capabilities, which include basic research, external search and commercialization.
Table 5-1: Linsu Kim’s Elements of Technological Capabilities Production Capability Production management to oversee operation of established facilities
Production engineering to provide information required to optimize operation of established facilities, including raw material control, production scheduling, quality control, troubleshooting and adaptations of processes and products to changing circumstances
Repair and maintenance of physical capital according to regular schedule and as needed Investment Capability Manpower training to impart skills and abilities of all kinds
Investment feasibility studies to identify possible projects and ascertain prospects for viability under alternative design concepts
Project execution to establish or expand facilities, including project management, project engineering (detailed studies, basic engineering, and detailed engineering), procurement, embodied in physical capital, and start-up Innovation Capability Basic research to obtain knowledge for its own sake
Applied research to obtain knowledge with specific commercial implications
Development to translate technical and scientific knowledge into concrete new products, processes and services Source: (Kim, 1997)
A number of other authors including Lall (1992), Katz (1984), Figueiredo (2002) and Ariffin (2002) have proposed alternative classifications of capabilities, according to the focus of their research. Katz sub-divided production capabilities into process engineering, product engineering and industrial engineering. Lall separated out the functions by investment, production and linkage capabilities. Linkage capabilities allow enterprises to liaise with suppliers, consultants, subcontractors, educational and technical institutions and therefore to acquire technology through these interactions. The inclusion of the linkage category reflects the importance placed on relationships with entities outside of the enterprise. Teece (2000) identified organisational functions that facilitate technological change and sorted them into three major categories: routinization, coordination/integration, and reconfiguration. The term ‘continuous improvement’ is used to refer to an organisational system in which learning of the workforce translates into a steady stream of improvements and assists the development of new products, as well as adjusting to minor innovations (Wilson et al., 1995). The ability to carry out the relevant organisational functions will therefore be referred to in this thesis as ‘improvement capabilities’. Teece’s first category of functions, routinization, involve the standardization of procedures, such as production monitoring and inventory control, which facilitate information collection, decision making, coordination and change. His second category of functions,
132 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
coordination and integration, ensure that the learning of individuals leads to growth in the total knowledge held by the firm (Nonaka, 1994). Examples of successful internal coordination and integration include long term training, the rotation of workers and the formation of teams such as quality control circles that facilitate interaction between different parts of the organisation and management structure (Fujimoto, 2001; Teece, 2000; Wilson et al., 1995). Externally, coordination and integration functions also facilitate absorption of learning and inform technology collaborations and relations with suppliers and markets. For example, quality deployment and just in time manufacturing strategies have been used by advanced producers to coordinate interactions with production priorities and to increase the integration of learning through closer links (Wilson et al., 1995). Teece’s third category of functions, reconfiguration of the enterprise’s assets and processes is crucial in being able to stay competitive in rapidly changing environments. The capacity to reconfigure cost effectively involves constant surveillance of technologies and markets, and the ability to see things differently and act accordingly.
5.2.2. A Typology of Capabilities Because this thesis is focused on technological learning, it is useful to separate capabilities that facilitate learning from those relating to production or innovation. Four types of capabilities are therefore distinguished. The first type is production capabilities, relating to the operation of facilities. The second type is innovation capabilities, relating to research activities, both basic and applied. The third type is investment and linkage capabilities, which enable relationships with other actors. The fourth type is improvement capabilities, which enable enterprises to learn from production, coordinate and integrate learning activities and reconfigure to take advantage of what is learnt.
133 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.3. Technological Learning
Having established a typology for capabilities, this section considers how capabilities are acquired. In common usage, learning refers to the process by which individuals acquire new skills and knowledge. In the context of technological learning, the term is also applied to organisations or innovation systems, and refers to the acquisition of technology, which may be embodied in equipment, documents and organisations, as well as people. The term ‘learning’ has been most commonly used in the economics literature about technology acquisition that is not specific to latecomers. The term ‘capability building’ has been used specifically to describe the learning processes by which enterprises in latecomer countries acquire technological capabilities. Since technological capabilities are the result of the possession of appropriate knowledge, equipment, documents and organisational routines and structures, capability building occurs through the acquisition of technology. The terms ‘learning’ and ‘capability building’ (the acquisition of technological capability) are therefore used interchangeably in this thesis. This section explores different conceptions of learning in economic theory.
5.3.1. Improved Performance via Learning by Doing There is a tradition of usage of the term ‘learning’ in economics literature to refer to improved performance, which is assumed to be achieved via ‘learning by doing’. In this document, however, the term ‘learning’ is attributed a broader set of meanings, referring to the acquisition of technology via a variety of internal and external processes which will be explored in the following sections. The concept of learning by doing as the acquisition of know-how through practice (in manufacturing) was introduced by Arrow (1962). Through learning by doing, production skills are accumulated over time as problems and bottlenecks are identified and solved, increasing the efficiency of production operations (Rosenberg, 1982). Learning by doing is a by product of production, and is purported to occur at all times, without conscious effort or specific resources (Bell, 1984). Learning by doing has been empirically modelled by learning curves, previously described in chapter 2. Much of the work on learning curves has been applied to whole industries, including the photovoltaics industry (Masini & Frankl, 2003; Schaeffer et al., 2004; van der Zwaan & Rabl, 2003, 2004). Although the unit costs of producing various products has been shown to follow a learning curve pattern within an enterprise, large differences between organisations have been found, even when enterprises have been producing the same product with virtually identical equipment. Dutton & Thomas (1984) surveyed 108 studies on learning curves, and found that there had been a wide variation in learning rates between enterprises.
134 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
They concluded that the learning curve, which aggregates the results of all the factors that contribute to cost reduction, includes many effects which are not related to automatic learning by doing. Consequently, they attempted to identify the factors that do influence learning rates. Four types of learning were suggested, which may be either exogenous or endogenous to the enterprise. Some may arise automatically, whereas others must be induced, as described in Table 5-2. The learning rate was found to be influenced by technical change (endogenous and exogenous to the enterprise), local industry and enterprise characteristics and scale effects, in addition to labour learning (the traditional conception of learning by doing). Dutton & Thomas assert that the learning rate as predicted by past performance can not be seen as an independent variable, but rather as dependent upon an enterprise’s behaviour.
Table 5-2: Learning Types Identified by Dutton & Thomas Autonomous Learning Induced Learning Exogenous General growth in scientific and Learning of capital good suppliers Origins technical knowledge that flows induced by the users’ experience with freely into the enterprise. the equipment.
Continually improving productivity Investments in improved capital goods garnered when an enterprise in order to hasten the rate of progress. periodically replaces its equipment. Copying and adapting the technical innovation of a competitor. Endogenous Direct labour learning due to the Increased tooling. Origins “practice-makes-perfect” principle or wage incentive plans. Manufacturing process changes.
Routine production planning. Model or product design changes to effect efficiencies in production. Source: (Dutton & Thomas, 1984)
Argote & Epple (1990) also provide evidence to suggest that organisational forgetting (the obsolescence of knowledge or the loss of knowledge through disuse, or other loss), employee turnover; as well as economies of scale and knowledge spillovers (due to improvements in technical knowledge in the wider environment) may be factors influencing the variation in organisational learning rates. These studies suggest that learning curves as a function of output are simplistic; and that there is room for a better understanding of the factors that influence learning rates, including acknowledgement of the role of technological change and enterprises’ interactions with the external environment.
5.3.2. Traditional Views of Technology and Technological Change Two divergent views of technology and the drivers of technological change; the ‘demand-pull’ view and the ‘supply-push’ view have coexisted in economic theory. In demand- pull theory, favoured by neo-classical economists, and used in trade theory, enterprises freely
135 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
and costlessly select optimal technologies from the technology shelf in response to markets. In this paradigm, production is assumed to depend on inputs of labour and capital according to production functions. Technology is seen as an exogenous variable. When innovation or technical change occurs, a new production function describes the new situation (Lall, 1992). In the supply-push view, technology does not arise costlessly, but must be generated through intentional effort. The direction in which technology develops is determined by these efforts.
Figure 5-1: A Production Function with output Y, inputs K (capital) and L (labour)
Figure has been removed due to copyright restrictions.
Source: (Fonseca, 2007) Both of these views fail to adequately explain technological change. While market forces influence the direction of technology, as described by the demand-pull theory, suitable technology may not be available. Even if it is available, the tacit component of technological knowledge is not shared easily, so enterprises cannot easily access the technology to move about the production function in response to factor prices. Enterprises instead operate at a point, and their technical progress is localized to that point (Lall, 1992). While it is true that technical change may be induced by the R&D investments of enterprises and other organisations, as in the supply-push view, the direction of these efforts is influenced by markets, and enterprises do not act, or learn, in isolation, but through their interactions with other actors.
5.3.3. Technological Change via Evolutionary Processes Contrary to traditional views, evolutionary economists see technological change as arising through complex, path-dependent processes, similar to natural selection. In the theory of natural selection, mutations are the source of change, whereas in evolutionary economics, innovations fulfil this role. Circumstances make some mutations (innovations) more favourable, and hence constitute a natural (market) selection processes. Most innovations are incremental,
136 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
rather than radical (Malerba, 1992) and the process of technological change is an open ended process, which never reaches a state of equilibrium (Edquist, 1997). What new knowledge is able to be acquired by an enterprise depends on what has already been accumulated, because understanding often relies on other pieces of knowledge. Organisational learning and the effective use of equipment and document embodied technology is also path-dependent and an enterprise’s previous investments, its assets, and its institutions such as relationships and routines constrain its future behaviour. Enterprises therefore accumulate knowledge and equipment, and develop organisational routines and structures in different ways and with different resultant capabilities from each other. Dosi (1988, pp 1155- 1156) describes the “permanent existence of asymmetries among enterprises, in terms of their process technologies and quality of output”. He points out that economies of scale and the vintage of plant explains part of the inequality, but that also important is “different innovative capabilities, that is, different degrees of technology accumulation and different efficiencies in the innovative search process”. The evolutionary approach views knowledge as primarily tacit and an important source of competitive advantage. Technology cannot be costlessly acquired, nor can it be easily generated without building upon an existing base of technology.
137 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.4. Types of Learning
A variety of different types of learning have been identified by various authors. Learning by doing, Internal learning by searching (R&D), and Learning through interactions external to the enterprise, such as: o Purchase or transfer of technology, personnel or training services, o Learning by interacting in value chains, and o Learning through cooperation with other organisations.
In this section, these learning mechanisms, the types of knowledge likely to be acquired through each type, and the requirements for each to occur are identified.
5.4.1. Learning by Doing Learning by doing is the increased understanding and improved execution of tasks resulting from experience. It can improve production and enable search efforts to be focused towards feasible alternatives (Garud, 1997). As previously discussed, learning by doing is traditionally viewed as being automatic, requiring no explicit effort. The benefits of automatic learning by doing, however, are limited, since only ‘know-how’ required for routine tasks is learned automatically through production, whereas insight, understanding why things work the way they do (‘know-why’) is required for modification or improvement of tasks, and is considered to be acquired only through active and often difficult interpretation of information (Argote & Epple, 1990; Marcotte & Niosi, 2000). Teece (2000) pointed out that much of the learning by doing in production is not automatic, but depends on routine production systems monitoring (Garud, 1997). UNIDO published a report (Wilson et al., 1995) specifically focused on the importance of routines, referred to as continuous learning techniques, in the improvement of manufacturing capabilities in developing countries, which helps to explain why some firms have able to learn more effectively than others. Shop floor routines that have been central to the competitive advantages enjoyed by Japanese car manufacturers include (Fujimoto, 2001; Teece, 2000; Wilson et al., 1995): Monitoring and quality control procedures that identify better ways of doing things, including foolproof prevention of defects and real time feedback by production workers; Total productive maintenance, where operators take responsibility for on the spot identification of and prevention of problems; and
138 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Continuous revision of standard operating procedures by managers.
Since learning by doing arises through experience in production, the production of many units is the facilitating factor. Learning by doing therefore relies on a number of factors, as illustrated in Figure 5-2: Availability of the resources required for production, which are obtained mainly through markets, but governments may also alter the allocation of resources (indicated by a dotted line); Investment opportunities, including markets and a sound business environment; Incentives to invest in production, which arise mainly through markets, but may be altered by governments (indicated by a dotted line); Production capabilities, which enable enterprises to produce products and services at an appropriate quality and cost; and Production monitoring routines for reviewing and interpreting the experience of production.
Figure 5-2: Capabilities and Learning by Doing
production resources investment opportunities
PRODUCTION CAPABILITIES incentives & resources ROUTINES Learning by Doing
DOING
5.4.2. Learning by Searching Beyond achieving efficiency in relation to one production technique, there is a need to move from one set of capabilities to another in relation to a particular technology, and from one type of technology to another in the learning process (Bell, 1984). Nelson & Winter (1982) and Dosi (1988) have interpreted R&D activities as a search process, whereby enterprises look for new technological options, test them and generate technical knowledge in particular directions. Learning by searching is active acquisition of new technology. This technology may be generated internally, through change processes or formal
139 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
R&D processes, or through journals and conferences, which are especially useful in providing information about the technological frontier and the direction of technological change. External contracts with technology suppliers or collaborative research arrangements often complement learning by searching. Spillovers may result from advances in science and technology external to the enterprise. Spillover learning may occur through active or passive absorption of knowledge resulting from the technological progress in the industry, in other industries, or in science more broadly (Malerba, 1992). Bell (1984) distinguishes learning by changing from learning by doing. Learning by changing occurs when the design and implementation of an improvement increases understanding of the technology, knowledge of the principles behind the technology and confidence in using the technology. In Bell’s (1984, p 192) words: ‘One may learn a little from using improved methods, but a great deal from defining and implementing them.’ In this thesis, learning by changing is considered to be a subset of learning by searching, since it depends on active search for new technology. While learning by doing mainly contributes to know-how, ‘know-why’ is likely to emerge from learning through change processes and internal and external search processes. Know-why enables the prediction of the effects of change, and therefore enables enterprises to further modify technologies (Garud, 1997). Know-why also increases the ability to absorb more knowledge (Lall, 1992) and strongly influences the appropriateness of the direction of search since prior knowledge dictates the problems that are addressed, the means used to solve them, and the solutions that are found (Garud, 1997). Learning by searching requires (Figure 5-3): Resources for search, including funding, knowledge and facilities, which may be sourced from other actors, but the allocation of these resources may be altered by governments (indicated by a dotted line); Incentive to search, usually the profit motive, but incentives may also be altered by governments (indicated by a dotted line); Direction for search, through feedback from internal monitoring and also external stimuli (indicated by a dotted line); Innovative capabilities, which enable enterprises to create new technology including new processes, equipment and quality control routines; and Improvement capabilities, which enable enterprises to coordinate research with production objectives, integrate learning and reconfigure assets to take advantage of R&D.
140 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Figure 5-3: Capabilities and Learning by Searching
innovation resources investment opportunities
COORDINATION & INTEGRATION
incentives, direction INNOVATIVE resouces for search CAPABILITIES & direction for search
Learning by Searching
R&D
new production RECONFIGURATION technique
5.4.3. Learning by Interacting The idea of learning by interacting is central to evolutionary theories and to the work of innovation systems authors, whose work will be further explored in section 5.7. Interactive learning occurs when enterprises interact with other organisations, whether through market interactions with suppliers or customers; through horizontal cooperation with other organisations such as research institutes, industry associations or competitors. Interactive learning gives enterprises access to the technology possessed by other organisations, and the sharing of technology provides opportunities for new knowledge creation, as existing knowledge is reinterpreted and combined in different ways.
Value Chain Interactions Interactions with suppliers are particularly important sources of learning, as enterprises may acquire technology embodied in equipment or materials or in the form of knowledge about its use, sometimes through training or documentation. Interacting in value chains often also gives enterprises access to the knowledge developed in other industries, for example, through equipment and materials purchases or through learning by interacting with customers in other industries. Interacting also increases the knowledge of the supplier when new knowledge about equipment or other inputs not only improves their use within the customer enterprise, but also feeds back to the supplier to improve design (Malerba, 1992). Learning through interactions with customers is also called learning by using and may involve learning from the users of final products as well as the users of equipment mentioned previously. Rosenberg (1982) introduced the concept of learning by using in a chapter of his book ‘Inside the Black Box’. Rosenberg describes two types of knowledge gained by learning
141 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
by using: ‘embodied’ knowledge that relates to the product and contributes to a better understanding of the design characteristics and thus performance of the product; and ‘disembodied’ knowledge that does not relate to the physical product, but rather to the development of new ways of using the product in order to increase its productivity. Both the embodied and disembodied knowledge may lead to new physical designs or new ways of using the product. Learning by using generates ‘know-what’, since customers often use products in a different way to which it was designed (Garud, 1997; Rosenberg, 1982), or provide information about what would be useful in the product.
Technology Transfers and Licensing Technology may be acquired externally through equity-based arrangements, or through licensing. Equity-based technology transfers include foreign direct investment and joint ventures, where a more advanced partner firm owns some or all of the equity in the recipient firm and freely transfers technology. Licensing arrangements often bundle equipment, document-embodied technology and know-how, often including commissioning and training. The most advanced technology is usually not for sale, it is usually mature technology that is available through these types of arrangements.
Horizontal Co operations Learning through horizontal cooperation with competitor enterprises or other organisations, such as research institutes, gives each organisation some access to the technology of the other. Organisations may interact horizontally through explicit technology collaborations, which usually also involve some search-based learning, informal interactions, through the exchange of personnel or through explicit effort in formalized training. The closer, more open and more frequent the contact, the more learning will take place. Positive feedback loops may result as organisations build upon what has been learnt through interactions and bring new knowledge to future interactions. Horizontal relations (inter-organisational cooperation) have more potential to produce radical innovations than vertical (value-chain) relations, which have been found to be more likely to lead to incremental innovations (Piek, 1998).
In conclusion, external learning depends on the nature of connectivity with other actors. Learning by interacting and the acquisition of resources and opportunities for production and innovation require (Figure 5-4): The presence of actors (enterprises, suppliers, users, research institutes, industry or advocacy groups); Fruitful interactions between actors, such that resources and knowledge can be created and exchanged;
142 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Investment and linkage capabilities, which include the abilities to search for and negotiate over technology, schedule investment and procure materials and equipment; and Coordination and integration capabilities to guide technology choice and deploy knowledge throughout the firm.
Figure 5-4: Capabilities and Learning by Interacting
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
5.4.4. A Learning Framework Bringing together the three types of learning identified in the preceding sections, Figure 5-5 provides a comprehensive view of learning processes within enterprises, linking specific types of capabilities with each process. Because each of the types of learning relies on the resources, opportunities and incentive available external to the enterprise, the environment within which an enterprise operates is critical in determining its success in capability building. In addition, learning is usually not automatic, and requires explicit effort and investment in capability acquisition (Dosi, 1988; Lall, 1992; Teece, 2000). Bell and Pavitt (1995) point out that this effort is primarily a management and organizational problem. A variety of capability building strategies may be adopted by enterprises, most of which involve more than one type of learning. Technology purchases, for example, often involve the acquisition of the technology from a supplier, internal searching in order to adapt the product and process and learning by doing in order to streamline production. In order to understand more about the value of different strategies and the critical factors that impact capability building in developing countries, the following two sections provide a review of the literature on capability building both modern sector and small-scale enterprises in developing countries.
143 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Figure 5-5: A Learning Framework
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
incentives & incentives, resources PRODUCTION direction INNOVATIVE resources CAPABILITIES for search CAPABILITIES and direction for search ROUTINES Learning by Doing Learning by Searching new production technique DOING R&D Improvement Capabilities
RECONFIGURATION
Later in this chapter, the technological systems literature will be introduced and used to extend the learning framework to include the system of factors external to enterprises that influence their capability building.
144 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.5. Capability Building in Modern Sector Latecomers
While the previous discussion about learning has been based on literature that is primarily concerned with innovation at the technological frontier, this section is based on the work of the capability building authors and is specifically focused on the acquisition of technology by latecomers. Capability building authors have been particularly interested in the strategies employed by enterprises in those newly industrialising countries that have successfully moved from building basic capabilities, often by acquiring foreign technology, to building increasingly more independent innovative technological capabilities. Kim (1997) observed that, contrary to the model of innovation whereby technological development progresses from product innovation to process innovation; enterprises in developing countries generally begin by acquiring mature technology, focusing more on learning by doing to improve production processes and reduce manufacturing costs. Katz (1984) found that in Latin American latecomers, product engineering capabilities were the first to develop, followed by process engineering skills. The production planning and industrial organisation capabilities were found to be the last to develop, often after at least a decade of operation. Mature technology can be acquired through technology transfers and licensing, such as: Equity based technology transfer arrangements, Forms of subcontracting, Technology purchases, and also often involves Hiring and training.
As latecomers become more capable, they gradually assimilate technology closer to the technology frontier and carry out more of their own searching, until they are finally in a position to bring their own innovative products to market (Figure 5-6). Capability building strategies which involve less dependence and more learning by searching include: Imitation, and Research collaborations (which also rely on learning by interacting).
UNIDO’s review of capability building for catching up (UNIDO, 2005) emphasised the importance of interactive learning. Interactive capability building has commonly occurred through: Interactions with suppliers, and Clusters and special industrial districts.
145 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Figure 5-6: Technological Trajectories at the Frontier and in Catching Up Countries
Product Process Innovation Innovation
Technology Frontier
Rate of Innovation Time
Fluid Transition Specific (emergence) (consolidation) (maturity)
Technology Transfer
Generation Improvement
Catching Assimilation Improvement Up Countries Acquisition Assimilation
Acquisition Technological Capability Time
Source: Adapted from (Kim, 1997, p 89)
The following sections describe some of the capability building strategies that have enabled manufacturers in developing countries to catch up, concluding with a comparison of different strategies, advantages and barriers to their adoption and the degree to which they afford the recipient opportunities to develop technological independence.
5.5.1. Equity-based Technology Transfer Equity-based technology transfer includes foreign direct investment (FDI) and joint ventures with foreign partners, usually large multinationals, whose contribution to total manufactures in developing countries form a large proportion of the total (Lall, 2000b), and who are the biggest exporters of technology (Ingham, 1993). A multinational firm will locate its production facilities in a developing country if it is profitable and low-risk to do so. The incentives for foreign investment by multinationals are access to natural resources, local or regional markets and local labour as well as any investment incentives. The basic requirements include adequate infrastructure, protection of proprietary rights, low government interference and stable government policies (Sharif, 1992). Inadequate
146 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
intellectual property and contract enforcement has been found to discourage foreign investment (ibid). US and Japanese electronics firms set up manufacturing operations via foregn direct investment in South Korea, Hong Kong, Taiwan and Singapore from the 1950s in order to access low cost labour and to gain entry into local markets (Hobday, 1995). These multinationals acted as examples to local firms. They helped promote the entry of local firms through subcontracting arrangements, and trained local people. The largest South Korean electronics producer, Samsung, was formed as a joint venture with Sanyo in 1969. 106 employees were sent to Sanyo in Japan for training (Hobday, 1995). As the capabilities of firms in the Asian newly industrialising countries grew, and the cost of labour also increased, multinationals began transferring more complex parts of their production to these countries. Later in the learning cycle, Korean firms have acquired knowledge through their own international mergers and acquisitions (Kim, 1997). Samsung and Hyundai acquired firms that possessed the technology and skilled engineers and equipment they needed to approach the frontier when the technological leaders were not willing to sell the technology (Hobday, 1995). The involvement of foreign partners may be detrimental to accumulation of technologies, since recipient firms are not forced to develop their own capabilities (Lall, 1990). Singapore has continued to encourage foreign investment, believing that the local entrepreneurial activity was insufficient. Most of the capabilities acquired have been through multinationals. In comparison with Korea and Taiwan, where local firms have comprised a growing share of electronics manufacture, Singapore’s indigenous capabilities, particularly their design capabilities, are relatively weak (Hobday, 1995).
5.5.2. Subcontracting Original Equipment Manufacture (OEM) is a form of subcontracting whereby a product is made according to the specification of the buyer, who then markets the product under its own brand. Under OEM arrangements, buyers may help with the selection of capital equipment, the training of engineers and managers, and technical and organizational assistance (Hobday, 2000). Own Design and Manufacture (ODM) occurs when the local firm carries out some or all of the product design, but usually according to specifications from the buyer. Firms may progress to Own Brand Manufacture (OBM), where they design the product and distribute it under its own brand. Hobday (2000) has studied the transition of latecomer firms in Asia from Original Equipment Manufacture to Own Design and Manufacture to Own Brand Manufacture, as described in Table 5-3.
147 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Table 5-3: Transition of Asian Latecomer firms in Subcontracting Arrangements
Technological Transition Market Transition
1960s/1970s Learns assembly process for Foreign TNC/buyer designs, brands and OEM standard, simple goods distributes.
1980s Local firm designs and learns product TNC buys, brand and distributes TNC gains PPVA ODM innovation skills. (post-production value-added).
1990s Local firm designs and conducts Local firm organises distribution, uses own brand OBM R&D for new products. name, and captures PPVA. Source: (Hobday, 2000, p 135)
Hobday found that Original Equipment Manufacture (often linked to licensing) allowed firms to export large volumes of goods, but usually at the low-end market segments and often with low value-added. The buyer could switch to another supplier at any time if they were not satisfied. For the catching-up firm, dependency on the buyer for technology, capital and brand resulted. This dependency constrained the development of the capital goods sector and indigenous R&D in the Taiwan, Korea, Singapore and Hong Kong (Hobday, 1995). The opportunity to gain scale economies, low risk entry, and access to markets and technology, however, allowed these firms to gain valuable experience via learning by doing and acquire technology and marketing skills by interacting with the buyers. Managers, engineers and technicians directly received training from the buyers, who wanted to upgrade the capabilities of the supplier firms, building local human capital (Hobday, 1995). The foreign buyers often established local offices and visited the supplier’s plant, advising the supplier on design, quality and cost; assisted in the establishment of operations; and advised on the purchase of capital equipment and materials (Hobday, 1995). Exporting to buyers who provide expertise has been a critical source of technology in the electronics industry in the east-Asian tigers (Hobday, 1995).
5.5.3. Purchase of Technology Technology purchases, such as technology licensing, often provide an entry point into the manufacture of products, but technologies at the frontier are generally not for sale. The nature of the technology being sourced and the willingness of the supplier to sell the technology will limit the capacity of firms to acquire technology by direct purchases (Lall, 1992). When the technology being sourced is further from the technology frontier, firms may not compete directly with their supplier, reducing the difficulty of acquiring technology (Marcelle, 2002). Developing countries have weak bargaining power in technology purchases, leading to high cost of technology purchases (Fransman, 1984). Foreign currency scarcity can also limit the amount of technology that developing countries will be able to purchase.
148 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Licensing requires more technical capacity than joint-ventures, where the senior partner trains the latecomer to manufacture (Hobday, 1995). Successful acquisition of technology by licensing usually requires transfer of tacit knowledge, so a face-to-face component to such transfers is important. In the case of Korean television manufacture, for which the design was still too complex in 1965, packaged technology was acquired and foreign engineers employed in production systems. Local engineers were also intensively trained and took over within a year (Kim, 1997). (Bell & Pavitt, 1995) demonstrate that effective systems for importing technology also involve building local capabilities and developing local technology, and see a synergistic balance between technology imports and local technological development. Local technological capability also allows recipient countries to have more influence over the terms and conditions of the transfer (Marcelle, 2002). Training and learning components and an explicit effort to acquire design, engineering and project management skills was found to lead to successful capability building (Hobday, 1995). Kim’s (1997) study of Hyundai in Korea traces the firm’s technological learning from assembly of foreign models from 1967-1983, manufacturing an indigenous model with foreign technology licenses from 1976-1995; and completely indigenous models from 1994 onwards. Hyundai followed an independent strategy in developing technological capabilities, learning first from manufacturing foreign designs, and then sourcing technology by employing foreign engineers and buying it from diverse foreign suppliers in the US, Japan and Europe. Large investments were made in technology search, training and R&D, raising indigenous knowledge and facilitating the absorption of the purchased technology.
5.5.4. Hiring and Training The hiring of experts is a means to obtain the technology embodied in people. (Sharif, 1992) identifies salary, work environment and living conditions as the primary determinants of labour migration. Quality staff can be imported temporarily to developing countries, but the transfer of skills depends on the firm’s improvement capabilities (Sharif, 1992). Many east Asian engineers were trained abroad in foreign firms, universities, and R&D institutes (Hobday, 1995). In particular, Korean firms have employed foreign experts in order to access the global R&D frontier (Kim, 1997). The flow of trained Taiwanese returning to Taiwan rose from 250 in 1985, to 750 in 1989 to over 1000 in 1991 (Hobday, 1995). Bell (1984) has also observed that sponsoring or otherwise encouraging overseas training and work experience for technical and managerial staff has assisted in the acquisition of problem-solving skills and provided linkages to informal international knowledge networks.
149 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.5.5. Imitation Imitation is not only about following the same path to industrialisation. It is often crucial to learning to industrialise (Juma & Clark, 2002). Imitation is often an entry point into R&D. Learning by searching, including internal R&D and learning from advances in science and technology, tend to begin later in the learning cycle, when firms approach the leader. These are particularly important sources of technology for firms in industries that rely on innovation for competitive advantage. Intellectual property laws, such as patents, trade secrets and trademarks protect firms from imitation. The need to apply for patents, however, exposes the knowledge (Teece, 2000). Product technology is more vulnerable to reverse engineering than process technology. Imitation, however, need not be illegal. Products or processes that do not have intellectual property protection may be copied legally. Teece (2000) believes that resistance to imitation determines the sustainability of competitive advantage. The more tacit a firm’s knowledge, the more difficult it will be to imitate. In the case of simple products, reverse engineering may not require much R&D investment. This kind of imitation will not result in a high level of learning. It does, however involve learning to find market opportunities and to find the knowledge or technology to take advantage of these opportunities; as well as learning to interact with suppliers and customers (Kim & Nelson, 2000). Some forms of imitation are legal because they surpass existing patents. These, called ‘creative imitation’ (Kim & Nelson, 2000), include creative adaptation, technological leapfrogging or adaptation to another industry. These types of imitation generally require investments in R&D and an element of innovation. The ability to imitate will be influenced by existing skills and knowledge. Organisational routines and complementary assets may be central to the use of a technology and may be difficult to imitate. As followers approach the leader, they need to invest in basic research to keep up. As they catch up, however, the R&D capabilities developed through imitation are a basis for developing innovative R&D capabilities. More complex imitation may also involve interactions with research institutions. In the Korean electronics industry, reverse engineering in simple products (with assistance from a German engineer) allowed the tacit knowledge for reverse engineering in more complex products to be developed (Kim, 1997). When the Korean electronics manufacturers entered the microwave market, they reverse engineered the product in order to force the foreign license holders to grant them a license to produce the patented product (Kim, 1997).
150 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.5.6. Research Collaborations and International R&D Small firms do not generally have the resources to invest heavily in internal R&D facilities. Research collaborations can enable firms to overcome their dependency on licensing of technology and subcontracting arrangements, and can help firms move from R&D aimed at improving manufacturing technology, to basic research aimed at developing new products and processes. In the Korean electronics industry, firms invested aggressively in R&D and placed fully funded R&D laboratories within universities that are jointly occupied by university and corporate personnel (Kim, 1997). Hobday (1995) observed that the establishment of overseas R&D centres or other outposts of technological learning has assisted technological capability building in the Asian tigers. Where global leaders in technology, such as manufacturers of colour TV and LCD displays have been unwilling to share the emerging technology, Korean firms formed alliances with local research institutions to enter the export market (Kim, 1997). These Korean manufacturers have also invested heavily in R&D, set up joint ventures with foreign experts and established international research facilities (Hobday, 1995; Kim, 1997).
5.5.7. Early Entry and Technological Niches It is often assumed that technological catch up involves imitation and following the same path of development, with success being dependent on speed. In fact many countries have become technological leaders by going in a different direction (Juma & Clark, 2002). Established firms have technology investments, assets and organisational routines that constrain their future behaviour. Learning will be based upon existing knowledge, and the technological options that are seen by management will be influenced by their experience (Teece, 2000). This gives newcomer firms an advantage in flexibility, since they are not constrained by existing investments and path dependence. Taiwanese firms, such as ACER achieved early entry by producing state-of-the-art PC components in the 1970s. In new fields such as biotechnology, some developing countries are aiming to move to the forefront and develop products for local markets (Juma & Clark, 2002). Early entry to a technology, (leapfrogging), however, is the exception to the rule. In most cases, firms have begun by manufacturing another’s design, progressively building their own innovative capabilities. Technological niches are also seen as a way for a new technological system to find a market in cases where the incumbent technologies have become entrenched.
151 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.5.8. Supplier Interactions Interactions within value chains are important sources of technology. It has been found that smaller firms benefit greatly from close interactions with larger firms, particularly in more developed areas (Piek, 1998). (UNIDO, 2005) have drawn attention to the political, economic and social factors that influence the connectivity between actors and therefore opportunities for learning by interacting. Equipment and service suppliers have been found to be particularly important sources of technology. Marcelle (2002) observed that African telecommunications firms acquired technology through purchases; product demonstrations, including field trials; training programmes, including long-term visits to the supplier’s site; and through technical support and documentation. She noted that although the contracts primarily involved the sale of equipment and codified knowledge; these relationships often included the transfer of tacit knowledge. However, the knowledge acquired was primarily ‘know-how’ rather than ‘know-why’.
5.5.9. Clusters and Special Industrial Districts Networks between firms, customers and suppliers in close proximity (clusters) have intensified interactions which enhance learning benefits, and are considered to be of substantial importance. Firms benefit from knowledge spillovers, the transfer of skilled personnel, and new ideas can circulate between firms that are geographically and socially close (Dutrénit, 2004). When firms develop backward linkages with local suppliers, new support industries may emerge and compete amongst themselves (Hobday, 1995). Clusters are considered to have attributes as follows (Albu, 1997): Specialisation, forward and backward linkages and exchanges of goods, information, services and personnel within the cluster; Cooperation when there is a common advantage despite competition between members of the cluster; A shared cultural identity, facilitating trust and behavioural conventions; The formation of self-help organisations within the cluster; Financial support and services provided by government. Special industrial districts, also known as development zones or science parks are set aside in many developing countries to encourage foreign investment and interaction between high-tech firms. Firms in special industrial districts often benefit from exemption from taxes and import duties. Large firms usually dominate production in these areas, which often contain high levels of competence and access to resources. There is often a strong cooperative culture (Piek, 1998).
152 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.5.10. Barriers and Advantages to Capability Building Strategies The capability building literature has provided a significant amount of empirical evidence to suggest that firms accumulate capabilities, in particular innovative capabilities, in a path dependent manner, and has identified the technology acquisition strategies that have been effective. Equity-based arrangements, such as foreign direct investment (FDI), licensing of technology, and subcontracting arrangements that include technology transfer generally involve international organisations and are important sources of catching up for industrialising countries. However, technological dependence is generally highest in the case of equity based technology transfers. Innovative capabilities, which are critical in high-technology industries, have been found to develop further where technology transfer arrangements have involved more internal innovative effort and less reliance on external sources of technology. Subcontracting therefore generally affords an enterprise greater independence in technological functions and greater opportunities for learning. Training staff and hiring experts have been found to be important in facilitating the absorption of technology from external sources and have allowed firms to access international knowledge. Research collaborations are powerful sources of innovative capability building, but are generally appropriate to firms that already have a base of innovative capabilities and are often approaching the technological frontier. Firms often build their own innovative capabilities by imitation, and at more advanced stages in their development may exploit technological niches by producing a product before it reaches maturity. The purchase of capital equipment often involves training components which transfer tacit knowledge along with the technology embedded in the equipment. Clusters of domestic firms have also provided synergies for technological learning and resulted in local knowledge spillovers. Firms benefit from the transfer of skilled personnel, and new ideas can circulate between firms that are geographically and socially close.
5.5.11. The Role of Learning and Technological Systems Literature in the Analysis of Capability Building While most capability building studies have focused on firm-level strategies for managing learning activities in order to accumulate knowledge, some authors (Kim, 1997; Lall, 1992, 2000a, 2003; UNIDO, 2005) have stressed that technological learning relies on an effective national technological system and have attempted to identify the role of policy in the catching up process and national competitiveness. Lall (2000a) identified three factors which impact the development of national technological capability, and therefore the capability accumulation process at the firm level:
153 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
The presence of incentives for firms to invest in learning, The ability of factor markets to provide the resources required for production and innovation, and The appropriate institutions to support the provision of incentives and resources.
The ways in which enterprises and other organisations interact within developing countries, their linkages with international actors and the ways in which institutional arrangements support or constrain them are clearly important factors influencing capability building in latecomers. The preceding review has provided a general picture of the value of different capability building strategies for enterprises in developing countries, but the capability building literature does not fulfil the requirements of this thesis because it does not (a) explain the relationship between capabilities and learning, or (b) explain the factors external to enterprises that facilitate capability building through each strategy. The framework based on the learning literature previously developed in this chapter describes the relationship between different types of learning and the capabilities which are built through them, and upon which they also depend. The technological systems literature, which provides a systems-based view of technological learning is introduced in section 5.7, and will be used to expand the learning framework developed in section 5.4 by linking types of capabilities and external factors to types of learning.
154 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.6. Capability Building in Small Scale Enterprises
The preceding discussion has been based on literature which has been primarily concerned with modern sector manufacturing enterprises. In this section, the literature on capability building in small-scale enterprises is reviewed in order to discover what is known about the types of capabilities required by small scale enterprises, the constraints they face and the capability building strategies and support typically available to them. This background will inform the analysis of small scale PV manufacture and help to illustrate the applicability of the framework being developed in this chapter to small scale enterprises. Despite their important employment and economic linkage benefits for local economies small scale industries are characterised by low productivity of labour. It is contested (Fernando, 1992; Islam, 1992), but also generally accepted that modern sector industries also use capital more efficiently (Kathuria, 1992). There is a strong correlation between the level of sophistication of technology used and labour productivity in small enterprises (Islam, 1992). In the Intermediate Technology Development Group (ITDG)’s work with small manufacturing enterprises, lack of technological capabilities has been found consistently to be one of the most important barriers to small producers finding and participating in new markets (Dawson & Jeans, 1997). Despite this, very little has been written about capability building in small enterprises. The focus of the published literature on technology and small enterprises has historically been on appropriate technology choice and technology transfer. There is a consensus that there has been too much focus on the technology itself, rather than the capabilities required to use it. ITDG has recently changed its name to ‘Practical Action’, accompanied by slogans such as “technology is only half the answer” and “practical answers to poverty” in order to reflect a more holistic view. The categories of technological capabilities previously identified can be usefully applied to small enterprises (Table 5-4), since, similarly to modern sector enterprises; they require abilities in production, innovation, investment and linkage, as well as organisational abilities to facilitate improvement.
155 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Table 5-4: Capabilities for Small Scale Manufacturing Enterprise by Function Investment and Linkage Production Innovation Improvement Technology choice Assimilation of Development of new Internal training Feasibility and market product design production techniques (often via studies Assimilation of Adaptation of products apprenticeship) Recruitment & training process for local markets or Organisation of Acquiring premises, technology production capabilities knowledge sharing equipment and materials Quality control Ability to Business management Inventory control reconfigure skills Repair and Linkage with technology maintenance of suppliers and markets, equipment including after sales service Accessing infrastructure and services
Since small scale enterprises generally lack the resources for technology development, they often acquire technology through their interactions with other enterprises or people, such as by purchasing equipment or through other informal modes of technology transfer. However, because they have limited resources, capabilities and poor connectivity with other actors, opportunities to interact are much more limited for small scale enterprises than for enterprises in the modern sector. NGOs and public organisations may therefore provide important avenues for small scale enterprises to build capabilities and acquire the resources they need. Appropriate technology NGOs and public technology institutes may provide technology development services. Small enterprise support such as education and training, finance, and business development services may also be provided by NGOs, private organisations or the public sector. These capability building mechanisms are now discussed.
5.6.1. Learning by Searching In general, small enterprises use established technology which is often traditional, and manufacture products involving simple technology and tools, such as carpentry, metal working, leather working, making construction materials or agricultural processing. Small enterprises rarely have the capabilities for independent technology development. It is often assumed that R&D capabilities need to be transferred to small enterprises from western societies, despite the fact that western methods have led to many innovations never emerging from developing country research institutions (Gass et al., 1997). However, the perception that small firms do not improve and innovate is called into question by some studies. While the majority of small firms (61%) in a 1990 survey of 186 Indian firms in the small-scale sector did not obtain any technology externally (Desai & Taneja, 1990), most of the firms developed technologically, with 25% dropping a product, 18% substantially altering a product and 50% making their own equipment. However, enterprises that use more complex technologies are less likely to be able to engage in innovation and will be more reliant on external sources of technology.
156 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.6.2. Subcontracting Joint ventures and foreign investment arise primarily in the modern sector, as most small firms lack the technical (in particular quality control) and linkage capabilities required (James, 1989). There are, however, sometimes opportunities for more accomplished small scale enterprises to participate in globalised manufacturing value chains, mainly through subcontracting arrangements. For instance, subcontracting of footwear production and tobacco growing and processing to the small and cottage industries sector has occurred in Sri Lanka (Fernando, 1992). Similarly to subcontracting arrangements in the modern sector, a buyer may provide a small enterprise with technology and market access. Under subcontracting arrangements or in clusters, where a larger entity is responsible for marketing, greater technological upgrading can be achieved.
5.6.3. Purchase of Equipment In the absence of commercial technology transfer such as foreign investment and joint ventures in the small scale sector, import of equipment is the most common and important mode of technology transfer from industrialised countries to small scale enterprises (Islam, 1992). However, the production techniques used in industrialised countries are often unsuitable for use in small scale production, and the materials required and the capabilities to operate and maintain equipment may not be available. Equipment developed in similar developing countries may therefore be more appropriate for the factors of production and is often cheaper (Masum, 1992).
5.6.4. Small Scale Technology Development Technology appropriate to small scale production may be developed either by upgrading traditional techniques, by descaling and adapting modern techniques, or by inventing new appropriate techniques as illustrated in Table 5-5.
Table 5-5: Alternative frames of reference in the improvement of traditional rural techniques Method Associated Frame of Reference Retention of traditional techniques Upgrading ‘Bottom-up’ Significant Descaling ‘Embodied modern technology’ None New Technique Existing scientific knowledge None Source: (James, 1989, p 9)
Descaling and upgrading are both examples of what Bhalla and Reddy (1994) describe as technology blending, where some characteristics of traditional and modern technologies are combined. Technology blending may better enable the poor to make the transition to improved technology because some elements of the traditional knowledge, skills and institutions are relevant to the new technology. Replacement investment involves investment in the development of an entirely new technology where it is not technically, socially or organisationally feasible to adapt existing traditional and modern technologies. 157 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
The development or adaptation of technology to suit local factors of production requires engineering skills which small enterprises are unlikely to possess. Government organisations, programmes and NGOs have therefore been formed to provide technology and product development services for small enterprises. While the purchase of technology or commissioning of technology development is not usually feasible for small scale enterprises, it can be economically feasible for local technology institutes or NGOs to import, develop and adapt technology which can then be made available to small enterprises. Technology transfer in small scale industries mainly takes place through these agencies and programmes.
Technology R&D Institutes in Developing Countries Government technology institutes have found it difficult to establish contact with small scale producers, due to the highly centralised structure of these organisations in most developing countries (Islam, 1992). They often also lack the capabilities to develop technology that is appropriate and disseminate it effectively (Islam, 1992). They often have good engineers, but know little about the needs of small manufacturers and rural communities. In particular, it has been assumed that small firms were not efficient because they did not have appropriate equipment. The need for skills, knowledge and organisational capacity have often been overlooked, as have the potential advantages of networks and clusters (Allal, 1999; Bhalla & Reddy, 1994).
Appropriate Technology NGOs A number of appropriate technology organisations; notably the Intermediate Technology Group (ITDG) and Volunteers for Technical Assistance (VITA); have developed an array of small-scale, simple technologies and transferred these to small producers in developing countries. NGOs have often been found to have a stronger motivation and commitment, and closer links to the community than government-provided services. The good local representation of NGOs enables them to better identify needs and provide technologies to small entrepreneurs (Bennell, 1999; Bhalla & Reddy, 1994; Islam, 1992; Kumar, 1994). They also often have the capacity to provide a package of services, including credit, technology and business services. The impact of NGOs, however, may be limited, since they operate at the fringes and are unable to make a large scale, sustained contribution to skills development (Bennell, 1999).
5.6.5. Support for Small Enterprises
Education and Training Training programmes provided by private organisations, NGOs or governments may complement learning on the job. However, few training schemes for the informal sector have been successful (Bennell, 1999). These programmes have been found to be too generalised and
158 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
too supply driven, failing to respond to the needs of the enterprises. The trainers have also often been inadequately qualified and there has often been no follow up of trainees (Allal, 1999). In addition, poor people are time constrained and generally cannot afford the time for training (Masum, 1992). New training approaches attempt to overcome some of these weaknesses. For example, an NGO in Maharashtra, India implemented a Ghandian concept where a school system was used to allow students to engage in profitable production while learning (Kumar, 1994), and ensured the training was targeted to the priorities of trainees. Technologies, including non- traditional ones, were developed into trades.
Finance Financing may involve support to enterprises for investments in capital or expansion. It may also involve the provision of credit, which has been found to be an effective mechanism in promoting sales to the poor customers of small scale enterprises who are often unable to afford the products on offer (Mishra, 1994). The leasing of equipment, joint ventures and special types of subcontracting arrangements are innovative types of financial services that have been offered to medium and small enterprises (Allal, 1999). The integration of business and financial services is recommended, since lack of financial capital is just one of the constraints faced by small producers.
Business Development Services There is a new push toward the provision of non-financial services to small enterprises, called Business Development Services (BDS). Business development services have a broad focus, variously providing small firms with assistance in aspects of running a business, such as management, organisation, sales, and employment, income and quality issues. These services are designed to assist small and medium enterprises to increase productivity, reduce costs or access markets (Allal, 1999). BDS commonly include the provision of information, through consultancies or exhibitions and trade fairs. Legal services may be provided in the registration of businesses or in arranging contracts with other (usually larger) firms. Training, testing and certification services may also be provided. There is not usually expertise in the relevant technologies, and hence the focus is not on upgrading of products, processes and organisation to increase technological competence. ITDG’s experience indicates that the credibility and ability to reach more enterprises of such services is enhanced by linking to micro-credit programmes (Albu, 2001). BDS may be provided by public or private sector organisations, and may be free or provided for a fee, often subsidised. NGOs providing BDS are few and, where they exist, they are most often involved in BDS for vulnerable groups. The services provided by NGOs and the public sector in the past have not been of sufficient quality (Albu, 2001) or relevance (Allal,
159 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
1999) and have not been valued because they have been given for free. A new BDS paradigm involves private delivery of BDS with full cost-recovery. When paid for, it is supposed that clients will ensure that the services address their needs, rather than being supply-driven (particularly in relation to training and information services, which may be accepted if for free or if there is a financial incentive to participate). According to ITDG, however: “there is little evidence that private actors can provide the services needed by IGAs and micro-enterprises at costs they can afford. In many cases, non-profit community-based organisations provide far more realistic channels for extending BDS to the poorest people… Furthermore, some BDS services (e.g. market information services, basic technical training) have characteristics hat justify and may necessitate some form of public provision.”
(Albu, 2001, p 10) 5.6.6. South-South Technology Transfer India has emerged as a significant exporter of technology to other developing countries including African and South Asian countries. Indian firms and institutions have offered services such as feasibility studies and investment planning, turkey projects, setting up industrial estates and training schemes in a number of countries under government economic cooperation programmes and other international programmes. Indian technology has the advantages of being low-cost and simple, with cottage industry technology usually being free of intellectual property constraints and often no fees for technical know-how (Kathuria, 1992). Fernando (1992) documents south-south TT routes used in Sri Lanka, including Sri Lankans working abroad, foreigners setting up cottage industries and a people-to-people technology exchange mechanism developed by IRED (international NGO) whereby village level users visit poor communities in another country, learning from people with the same socio-economic background and identifying suitable technologies for their community. Technologies from other developing countries tend to be more appropriate as they use more local raw materials and capital goods therefore generating more linkages domestically. The cost of technology from developing countries is also lower, usually containing less proprietary technology and therefore requiring less foreign currency (Islam, 1992). Islam (1992), Kathuria (1992) and Pavlic (1983) have noted the following challenges to south-south technology transfer: Lack of information about technology, Lack of institutions with appropriate capabilities in supply and buyer countries, Lack of currency, and Lack of incentives.
160 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.6.7. Barriers to Small Scale Capability Building Four important characteristics of small enterprises in developing countries that influence their technological capabilities will now be discussed: Reliance on local capabilities, Reliance on traditional technology and lack of innovation, Lack of investment and linkage capabilities, Importance of social capital in determination of access and influence.
Reliance on Local Capabilities The capabilities and constraints of small scale enterprises are those of their owners and employees, who are generally poor or very poor, and live in rural areas (Carney, 1999). They employ workers mainly from within the family, and are often located within the family home (Masum, 1992). The rural poor often lack the material, human and social assets required to use technology and to take advantage of small enterprise livelihood opportunities. Aspects of human capital such as health, formal education and training, as well as the integration of prior experiential learning gained through learning by doing, directly impact technological capabilities. The ability of workers to use technology depends on their level of education and their on the job training. Skill drain, when the educated emigrate to urban centres to take advantage of their education in higher skilled and paid work poses a challenge for rural enterprises (Bhalla & Reddy, 1994). Economic and social constraints also affect the capabilities of people and small enterprises to use skills and knowledge productively.
Reliance on Traditional Technology and Lack of Innovation Workers in micro and small enterprises usually possess traditional skills which have been acquired informally (Masum, 1992). Small entrepreneurs have been found to be risk adverse, since they cannot afford to take risks with the assets they depend on for their basic needs; and therefore tend to resist new ideas and technologies (Islam, 1992). While small enterprises do innovate, the use of modern technology such as that required for the manufacture of PV system components is likely to pose a greater challenge than that used by most small enterprises, because it is further from traditional knowledge.
Lack of Investment and Linkage Capabilities Investment capabilities for small scale manufacturing enterprises include those required to choose and acquire technology, to establish the feasibility of the venture, to employ and train staff, and those for business planning and management (Bennell, 1999). Knowledge about business management, technology and markets has been found to be low among small entrepreneurs (Fernando, 1992; Masum, 1992), and to be the primary causes of their failure.
161 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
These low levels of knowledge have been attributed to lack of access to information and high transaction costs due to inadequate linkage capabilities and institutional arrangements (Dorward et al., 2003).
Social Capital and Linkage Capabilities in Small Enterprises Social capital includes all the social bonds (relationships and memberships of groups), agreements and norms (accepted ways of behaving) that can result in cooperation and collective action and influence access to wider institutions and services (Overseas Development Institute, 1999). Social capital can be seen as the primary determinant of the linkage capabilities of small enterprises, since it is central to determining their access to resources, their influence, and their ability to organise into groups. The sustainability of rural development projects has been closely linked to social capital formation: “A variety of studies of rural development have shown that when people are well organized in groups, and their knowledge is sought, incorporated and built upon during planning and implementation, then they are more likely to sustain activities after project completion.” (Pretty & Ward, 2001, p 210)
Such studies substantiate the value of working with community and decentralised and participatory approaches.
5.6.8. Lessons Learnt in Small Enterprise Technology Development and Support Interventions to assist small scale enterprises with technology have been found to be more successful if they: Are participatory, Are demand driven, Have appropriate timescales, Are at an appropriate scale, Are integrated and link micro and macro organisations and institutions, Involve collective action.
These elements ensure that small enterprises can successfully link with the organisations implementing the intervention and with other actors in the supply chain.
Participatory A high degree of local involvement has been found to be critical to the long term effectiveness of technology programs. Local involvement in technology development enables producers and users to adapt and improve the technology so that it is locally appropriate. It is recommended that external institutions should transfer only basic technology, allowing potential
162 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
producers to develop the technology themselves. Douthwaite (2002) in researching agricultural harvest equipment in the Philippines and Vietnam, that the equipment was more likely to be used and of use if people could understand it and adapt it to their needs. Technologies that were adapted were the most widely adopted. He believes that agricultural extension should be about “helping farmers to understand and innovate”, rather than promoting technologies to achieve the widest adoption. However, a review by Stephen Biggs of around 40 participatory technology development projects in agriculture found that very few of them had strengthening the research capabilities of poor farmers as a primary goal (Gass et al., 1997). Project teams tend to focus on more measurable outcomes such as rates of adoption. The capabilities for management, accounting for, maintenance and proper use of technology must be developed locally, otherwise programs have frequently failed when the government/NGO workers are withdrawn (James, 1989). When temporary staff and structures that are part of a technology project control the learning process too much, the learning accrues to the wrong people and local groups remain inexperienced and uninformed as they were to begin with (James, 1989). In a study of Indian watershed management programmes, government funded and executed programmes were found to be the least able to institutionalise participation, government-funded and NGO-executed programmes were average, and NGO-funded and NGO- executed programmes performed well (Vania & Taneja, 2005).
Demand Driven Approaches ITDG has found that technology institutions and NGOs (ITDG included) have been too supply-focused and tended to advocate solutions that are supply-driven, particularly biased towards the skills and knowledge of the individuals and institutions offering the service. Other constraints are often not addressed. Technology development must meet the needs of the producers and users in order of priority if is to succeed in warranting sustained effort and investment on their part (Krishnaswamy & Reddy, 1994). A commercial orientation may lead to a focus on real needs, through incentive structures based on market principles (Romijn, 2001). Field demonstrations, the provision of credit and training can help overcome risk aversion and low adoption rates. Small scale enterprises may compete with medium and large enterprises that use more sophisticated technology to produce goods more cheaply and of higher quality (Panditrao, 1994). Larger enterprises also employ superior marketing and business strategies (Fernando, 1992), and benefit from better access to capital, information, training and management experience (de Wilde, 1991). Even when the production technology itself is scale neutral or can be produced more economically at a small scale, small enterprises will find it difficult to
163 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
compete. Basic needs are often more cheaply satisfied by mass production, and small industries are often better aimed at niche markets.
Timescales Studies have revealed that it took on average ten years for agricultural innovations to move through basic research to the point that farmers began to spend their own money on improving the innovation and adopt it widely (Douthwaite, 2002). The amount of time allowed for technology to be introduced through appropriate technology projects has been shown to strongly influence the success of the interventions (James, 1989). Rogers (2003) points out that most innovations in any environment diffuse slowly. Foreign donors, however, want to see results and provide funding over a few years’ work, with the need for success to be demonstrated for further funding. Programmes thus benefit from showing results early, and place often insufficient emphasis on long term sustainability. Appropriate timescales are also important in relation to market development. Small enterprises usually lack sophisticated business and marketing experience, so they generally produce products to supply local markets. The users of small scale technology are generally poor, vulnerable and risk-adverse. They are therefore only able to make incremental changes to their practices (James, 1989) and require time to absorb unfamiliar technology (Fernando, 1992). The price of small scale products is critical and the upgrading of quality may exclude customers. Technological and market upgrading of technologies for the poor should therefore be incremental, since there is a limited capacity to penetrate unfamiliar markets (Fernando, 1992). The market may also be limited in size and dispersed. An initial burst of production can potentially saturate a small market (Mishra, 1994).
Scale Large scale projects are usually administered centrally, and have a hierarchical structure which excludes people from participation and influence. The implementation of large scale programmes tend to focus on what is feasible politically and bureaucratically feasible, rather than what is best for community development (James, 1989). There is often an emphasis on targets and implementation with the fewest resources possible, rather than on using the resources that are available in communities. Existing political and bureaucratic structures are often used in programme implementation, which may be effective and conserve resources, but may also obstruct optimal institutional design. The institutional arrangements in large scale projects are likely to be rigid and therefore not able to suit local preferences and requirements. Poor people are also more comfortable dealing with their peers and unlikely to interact easily with formal organisations. The poor therefore may not understand their rights and may not be able to effect change (DFID, undated). New organisational structures where decision making is handed to the community, particularly the disempowered, may be better able to assist the poor.
164 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Small scale projects can employ small scale organisational structures, and are therefore more likely to fit local needs and capabilities. The people involved in small scale projects are also usually more closely linked to the communities and therefore more engaged and motivated.
Integration and Micro-Meso Linkages There is often a multitude of different schemes, such as (technology, training, credit) and bureaucracies (health, education, BDS, agricultural) involved in small enterprise development, but these are not usually coordinated (Saxena, 2003). Capabilities and development needs are often interrelated and interventions should be harmonized. Skills development should be mobilised around specific community development options (Bennell, 1999). While small scale organisations are often best placed to deliver appropriate services to enterprises and communities, some resources are best developed centrally, to avoid replication, while some services are the responsibility of the government. Bhalla & Reddy (1994) suggest that private-public partnerships are required, since each of the private sector (innovation, production, marketing skills), government (R&D resources, incentives) and NGOs (community links & knowledge) have a part to play. Bridging institutions can link micro and macro organisations and provide better access and influence for the poor (James, 1989). If technology is developed centrally, effective institutions are required to ensure feedback from users to product/process developers, manufacturers, installers and maintenance technicians.
Collective Action Collective action such as through producer groups or clusters of similar enterprises, especially those at close proximity to each other, can an play an important role in increasing the influence of small scale enterprises over the institutional arrangements that affect them. Well organized associations may combine to organize the upgrading of infrastructure such as water, power or transport. Producer associations can help to achieve a more favourable regulatory environment in marketplaces. These groups can also increase the access of their members to knowledge, materials and markets where they have shared needs, giving enterprises the ability to access economies of scale in innovation, production and diffusion. Bulk purchasing is common in small manufacturing enterprises, but joint tendering or marketing of products has been found by ITDG to be less common (Albu, 2001). Since the benefits of collective action are well established, the current research into collective action is primarily concerned with identifying the conditions under which individuals will have sufficient incentive to contribute to collective action that will benefit a group. (Romijn, 2001, p 59) points out that “intense competitive rivalry, lack of trust, poorly functioning factor markets and underdeveloped private-sector services” often preclude small
165 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
firms from inter-firm cooperation. The experience of ITDG supporting small scale manufacturing enterprises indicates that “collective action (for example in purchasing materials or marketing goods) by isolated small-scale producers, is slow and costly to facilitate, especially when competition is intense.” (Albu, 2001, p 4) There is a danger of such organizations being controlled by powerful interests. Harper (1996) suggests that group enterprises may be slower to make decisions and changes and potentially benefit some more than others.
5.6.9. Concluding Remarks This review has provided some background on the typical constraints and capability building strategies of small scale enterprises. In particular, it has highlighted the importance of good linkages between enterprises and other actors. PV is a modern technology, and while the process technology is fairly simple, the product technology is more complex than is usually used by small enterprises. It may therefore be considered inappropriate for small scale enterprises. Small enterprises in the PV industry are likely to be more dependent on external sources of technology and support. Linkages are therefore particularly important.
166 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.7. Technological Systems
PV enterprises of different sizes and in different locations have unequal technological capabilities, which impact their relative ability to manufacture and sell their products profitably, and to build upon these capabilities further. There has also been variation in the success of different countries in nurturing PV enterprises and fostering a successful PV industry. Developing countries, in particular, have had little success in supporting PV manufacturing. While firm-level learning processes and associated capabilities have been identified in the previous sections, a framework that explains capability building must also identify the external influences on learning. In this section, the implications of the evolutionary theory of learning on system-wide technological change processes are discussed, providing a basis for the introduction of the ‘technological systems’ approach. The technological systems literature, which brings a systems-based perspective to the analysis of technology development, is then introduced. The body of literature is focused on explaining the emergence of innovations (new creations), rather than learning (including technology acquisition) more generally. However, because it deals with technological change, it provides a useful classification of the components and interactions within technology systems that participate in and impact technological learning within enterprises, within industries and within nations. The approach acknowledges the importance of interactive and evolutionary learning processes and the influence of institutions on technological trajectories. The learning literature previously reviewed and the technological systems approach, reviewed here, will then be integrated in section 5.8, to propose a framework which identifies the ways in which elements of both technological systems and firm capabilities influence the learning processes of enterprises.
5.7.1. Evolutionary Learning and Technological Trajectories The evolutionary and path-dependent nature of technological learning within enterprises, and the resulting asymmetries in their technological capabilities has been described in section 5.3.3. Path dependence also exists in markets and institutions external to the enterprise, since greater adoption of a product will lead to supporting infrastructure, institutions and networks for its manufacture and use as well as scale economies and learning by using effects. As a technology is adopted, there are increasing returns to adoption by new users. Early leads can be advantageous, sometimes resulting in lock-in of technologies, due to high switching costs; well-know examples of which include the QWERTY keyboard and the DOS operating system.
167 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Dosi (1982) introduced the idea of technological paradigms that describe the evolution of technological change. Similarly to scientific paradigms, technological paradigms embody an outlook on the part of individuals, enterprises, organisations and institutions. A technological paradigm is a branch among the evolution of technological development. The paradigm includes perceptions of problems that need to be solved and the range of technologies that can provide solutions. It therefore constrains and enables technological development. In the evolutionary view, technological change and learning arises through systems of interdependent actors within a technological paradigm. The emergence of new technological paradigms brings about radical technological change, or a new technological ‘trajectory’, which is “the activity of technological process along the economic and technological trade-offs defined by the paradigm” (Dosi, 1988). Technological paradigms emerge through a combination of factors, including economic factors such as cost-saving and labour saving, institutional effects and the priorities and expertise of the enterprises involved. The rate and direction of technical change is decided via competition between innovation systems for various technologies1, both fully developed, and emerging ones. How the capabilities of actors change, and how institutions and networks evolve, will shape the growth path of a new system and its ability to compete with other systems. The evolutionary view draws attention to the path dependence of technological learning and innovation; and to the interdependence of actors and institutions. In the following sections, a systems-based evolutionary view of technological change and it’s implications for enterprise capability building is further explored.
5.7.2. Innovation Systems The innovation systems approach emerged in the late 1980s, and was brought to prominence with the publication of Lundvall’s book ‘National Systems of Innovation: towards a theory of innovation and interactive learning’ (Lundvall, 1992). Lundvall and other authors who studied national systems of innovation were interested in demonstrating the need for an alternative approach to neoclassical economics that explained innovation and learning. The idea of the isolated profit-maximising enterprise was felt to be inappropriate (Edquist, 1997). The approach has developed around theories on: Learning by interacting: The idea that innovations come about within systems comprised of many actors, and that learning emerges from the interactions between actors.
1 An innovation system is a system of actors, institutions, and their interactions, within which innovations are created. The following section further defines and explores the innovation systems concept. 168 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Evolutionary theory: The path-dependent and non-equilibrium nature of technological change, and the role of diffusion as well as invention and adoption of technology in the emergence of technological paradigms. Institutional Economics: The idea that institutions strongly influence the operation of markets, the distribution of resources, interactions between actors and hence technological paradigms. Organisational change and institutional change are considered to be important, as well as technical change.
5.7.3. Technological Systems Innovation systems authors have concentrated on national-level analysis of factors influencing the innovative performance of enterprises. The ‘technological systems’ approach was developed with a focus on sectoral, rather than national, innovation systems, since countries perform differently in different sectors. A technological system is defined by Carlsson & Stankiewicz (1991, p 111) as a “…network of agents interacting in a specific economic/industrial area under a particular institutional infrastructure… and involved in the generation, diffusion, and utilization of technology.” The national unit for studying innovation systems may be too broad, when it is considered that countries perform differently in different sectors; but in some cases too narrow, when globalised industries are considered. However, from a policy point of view, a national focus may make sense, since policies and institutions are often nationally based. The study of regional and national technological systems for a particular technology has gained importance as this is the context within which enterprises operate (Dutrénit, 2004). Since this thesis is focused on PV technology in particular, the term ‘technological system’ will be used henceforth.
5.7.4. Components of Technological Systems and their Roles Technological systems are commonly viewed as being comprised of three parts: actors, the networks that link them, and the institutions that influence them. The roles and requirements of each of these components in the creation of new knowledge and the emergence of technological paradigms are now identified.
5.7.4.1. Actors (Organisations) Actors in technological systems are organisations that influence or participate in innovation; including those in the value chain, such as suppliers, customers and competitors; but also other organisations such as universities, research institutes, industry organisations, banks and government ministries. Organisations, including enterprises and research institutes, are the primary agents for technological change. In addition to the learning of existing actors, one of the
169 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
main ways in which technological systems emerge and grow is via the entry of new organisations into the system. These actors contribute to learning by interacting, and bring new knowledge and resources from other industries.
The Role of Actors Actors individually and through their interactions carry out the work of knowledge creation. Enterprises are often the most important actors in technological systems since, although their mandate is to generate profits, innovation is often the primary source of competitive advantage (Edquist & Johnson, 1997). Actors also influence laws, rules and professional or societal attitudes and norms in the institutional environment. They can influence expectations and confidence in technology, disseminate information and lobby on behalf of a technology for favourable institutional arrangements (Unruh, 2000). Marketing efforts may also influence the formation of markets, since innovations rarely find ready-made markets. Private non-market actors such as advocacy and user and industry associations may be formed specifically for this purpose. Public organisations often directly shape institutions by formulating and implementing technology policy and regulating aspects of product design, production or other processes that occur within enterprises, as well as the relationships between enterprises. Particularly influential actors in technological systems have been named ‘prime movers’ (Jacobsson et al., 2002), who are technically, financially or politically powerful. These actors can strongly influence the success of the technology: they ‘raise awareness, undertake investments, provide legitimacy and diffuse the new technology’. Capital goods suppliers are often actors of primary importance in technological systems. The actions of enterprises, and their learning processes, are shaped by the nature of linkages (networks) and through the influence of institutional requirements or economic incentives. In summary, the role of actors is to: Create knowledge, Influence institutions, Influence perceptions of technology.
5.7.4.2. Networks Networks are made up of relationships between actors. They are important channels for the transfer of technology, since it is within networks that learning by interacting, using and external searching occurs. Network interactions usually involve exchanges of resources, such as financial capital, physical resources, personnel or knowledge. These interactions may be based on markets, within supplier or customer relationships, for example; or may be non-market related, such as technical collaborations or membership of industry associations.
170 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
The Role of Networks Håkansson (1989) distinguishes three components of networks: actors; activities, in which resources are combined, developed, exchanged or created; and resources, including physical, financial and human capital. Market and non-market networks and interactions in technological systems facilitate learning and the exchange of resources. External learning processes, such as learning by interacting, learning by searching, learning by using and learning through spillovers are primarily a function of an enterprise’s engagement with networks of suppliers, competitors, research institutes and users. The processes of knowledge creation and learning within organisations are complemented by processes involving external interactions, which may lead to the recombination and reframing of existing knowledge, extending it and giving it new applications. Internal learning processes also rely on the enterprise’s access to the resources required for production (doing) and for innovative activities (searching). This includes financial capital, technology, inputs to production, human capital and infrastructure. The access that enterprises have to these resources is determined by the nature of their participation in networks. By facilitating access to resources and access to relationships within which to exchange resources, networks provide opportunities for the entry of actors. The nature of interactions in networks influences opportunities to make profits or threats to existing profits, and therefore alters incentives for enterprises to innovate. Incentives affect the type and quantity of investments organisations make in technology. Without threats to existing profits, or opportunities to make further profits, enterprises will be unlikely to invest in improving or expanding. The nature of competitive incentives for profits is influenced by the size of markets, the number of actors in the industry, its level of development and rivalry between actors. While individual actors, particularly prime movers, can influence institutions and legitimise technology, when enterprises combine in networks, they increase their reach and influence. Networks are important drivers of technological trajectories, since actors and their interactions influence perceptions about future technical possibilities and market opportunities. Through the activities and learning of networks of actors, technological paradigms emerge and grow. Increasing adoption leads to what Sanden (2005) calls institutional virtuous circles. The entry of more actors, the formation of market and non-market networks, increasing adoption, and the development of favourable institutions legitimise the technology, act to make the technology more entrenched and work for its survival. The legitimisation of technology will in turn impact markets and willingness to invest in technology from both public and private perspectives. The development of the technological paradigm also provides direction for search.
171 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
The roles of market and non-market interactions in networks are (Figure 5-7): Exchange of resources, Provision of investment opportunities and incentives (mainly through the market), Knowledge creation and exchange through interactions.
Networks of actors also: Influence policy and institutions, Influence perceptions of the technological paradigm.
Figure 5-7: The Functions of Networks in Technological Systems
TECHNOLOGICAL TRAJECTORY
influence perceptions of the technological paradigm
Networks
OTHER ACTORS influence policy and institutions
MARKET & NON-MARKET INTERACTIONS INSTITUTIONS
knowledge creation & exchange resources for production and innovation investment opportunities incentives
ENTERPRISE
5.7.4.3. Institutions Authors vary in their use of the term ‘institution’. Some authors, include “both regimes and organisations” under the institutional umbrella (Carlsson & Stankiewicz, 1991, p 45). Other authors, such as institutional economists2, separate organisations, such as enterprises, universities and government agencies, from institutions such as laws, rules and norms, which govern relations between individuals and/or groups. Since organisations, or actors, such as
2 Institutional Economics is a multidisciplinary area of investigation that attempts to ‘explain the determinants of institutions [and to]…evaluate their economic efficiency and distributional implications’ (Nabli & Nugent, 1989, p 1335). Although economists often advocate institutional change, institutions are generally not incorporated into neoclassical analytical frameworks. Institutional economists draw attention to the importance of institutions in determining economic growth and in particular market development (Dorward et al., 2003).
172 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
enterprises, universities and government agencies, directly participate in technical change, it is useful for the purposes of this study to distinguish them from institutions.
The Role of Institutions Actors are strongly influenced by institutions. They are constrained by the legal system, standards and norms of behaviour. The formation of networks, exchange of resources and learning by interacting depend on the cost of carrying out interactions, whether they are market or non-market transactions. New institutional economists consider that transaction costs and risks are increased by incomplete or asymmetrical distribution of information and by uncertainty (Nabli & Nugent, 1989). Conflicts and cooperation within technological system interactions are mediated by institutions, including institutions that manage potentially conflict-causing events such as the creation of new enterprises, the dissolution of old enterprises, the transfer of personnel, technology or other goods (Edquist & Johnson, 1997). The management of these conflicts reduces transaction costs, enables actors to participate in networks and increases their connectivity. Rules, accepted practices and stable relationships between parties provide information that reduces uncertainties in the conduct of business, particularly in relation to innovation activities. The enforcement of property rights through patents and the intellectual property rights laws, for example, reduces uncertainties about the rights to use technology and the extent to which an enterprise can be sure of claiming the returns on their learning investments. The enforcement of standards also provides information to actors and reduces uncertainties. Standards may also provide guidance in relation to the choice of a specific design approach. Institutions affect the opportunities that actors have to access networks, and the incentives for actors to engage in innovation (Jacobsson et al., 2002), by influencing the operation of markets and the allocation of resources. They may therefore strongly influence the rate and direction in which technology develops. The role of policies that determine these institutional arrangements is further discussed in the following section. In summary, the roles of institutions are illustrated in Figure 5-8: Influence market operation, Facilitate connectivity and information flows, Alter resource allocation, Provide incentives to invest and innovate.
173 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Figure 5-8: The Functions of Institutions in Technological Systems
Networks
OTHER ACTORS influence policy and institutions
MARKET & NON-MARKET INTERACTIONS influence operation of markets influence connectivity INSTITUTIONS
knowledge creation & exchange resources for production and innovation investment opportunities provide incentives to invest and improve incentives alter allocation of resources
ENTERPRISE
5.7.5. Networks and Institutions in Developing Countries Developing countries often have cumbersome laws and administrative procedures (Tybout, 2000) that increase the cost of carrying out transactions and reduce connectivity between actors in technological systems. Low transaction costs make firms more competitive and create a more attractive environment for investment in technology. For example, the IPPC Working Group III special report, ‘Methodological and Technological Issues in Technology Transfer’ suggests that changes in legislation on intellectual property rights, for example, have facilitated technology transfers to developing countries (IPCC, 2001). The innovation climate within developing countries is generally poor, since there are not good linkages between universities, research institutions and industry while public technology policy planning is usually short term (Sharif, 1992). For latecomer firms, international linkages to the sources of frontier innovation are perhaps even more important than domestic interactions. However, firms in developing countries are usually disconnected from the main international sources of R&D and technology (Hobday, 1995). They are also disconnected from advanced markets and user-producer relationships that are important sources of interactive learning and incentives to innovate (Lundvall, 1992). They therefore face two sources of innovative disadvantage that constrain their access to knowledge inputs.
174 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.7.6. Small Scale Enterprises and Technological Systems Similarly to the modern sector, technological systems for small scale technology include technology users, competitors, and suppliers of equipment and materials. Small enterprises also rely on supporting organisations and institutions such as financial services, product development services, infrastructure, effective technology diffusion channels and accessible input markets. However, markets and networks often fail to provide adequate incentives for technology development and improvement, the linkages between actors in small scale systems are usually poor and small enterprises have little access to and influence over the actors and institutions in technological systems.
Market Failures The livelihoods of most people (poor and otherwise) depend heavily on their interactions with markets. Markets, however, often perform imperfectly and particularly fail to serve the interests of the poor. The markets operating in rural areas are likely to be very imperfect, due to inadequate communication and legal structures in relation to contract enforcement and property rights. Actors, particularly those with little power or financial and social capital, face high transaction costs in accessing information and property rights enforcement, and this in turn constrains their access to markets (Dorward et al., 2005). In summary, the markets in which small enterprises and the poor participate are affected by the following circumstances (Albu & Scott, 2001; Dorward et al., 2003; Nabli & Nugent, 1989): High transaction costs, Monopoly distortions (due to limited choices of buyers or suppliers), Problems with information flows, Market access failures, Lack of clarity over property rights.
At the root of these problems are inadequate infrastructure, low connectivity, small, dispersed markets and minimal ability to influence institutions. Institutions play an important role in determining the nature of markets. In addition to the macro-level determinants of market operation such as fiscal, trade and investment policies and legal infrastructure relating to commerce and finance, more municipal-scale factors that influence the interactions of small scale enterprises with markets include (Albu & Scott, 2001): Local institutions that are often dominated by the power of local officials; Other players in the market such as large firms, service providers and influential buyers or sellers (especially where there are oligopolistic conditions) will have much more
175 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
power to affect the outcome of transactions than the poor. Fees may be payable to access the market & different sellers may pay different amounts.
As a result, markets do not deliver technology that fulfils the needs of the poor at an appropriate level of quality and price. The technology developed for industrialised nations is not necessarily suited to the needs of or factors of production available in developing countries, particularly in relation to the needs of rural populations (Bhalla & Reddy, 1994). The poor, who often largely rely on subsistence agriculture, do not have the purchasing power to send demand signals to trigger the development of technology that fulfils their needs (Bhalla & Reddy, 1994). The private sector therefore does not invest in technology development for small scale rural enterprises.
Non-Market Service Provision Non-market interactions are particularly important in the small scale technological system, since the market often fails to operate effectively. The delivery of goods and services to the poor may be free of charge or subsidised via welfare support or NGOs (Krishnaswamy & Reddy, 1994). For instance, services such as health, education, water supply and energy supply may be provided by government bodies or other programmes regulated by government policy. These programmes influence prices and product standards and have the potential to encourage forward and backward consumption and production linkages between markets. However, inadequate information flows and lack of clarity over rights and obligations, access failures, and high transaction costs can also affect non-market interactions (Kumar, 1994). The institutional structures in programmes that support the use of and the maintenance of technology are critical in determining its successful diffusion (Krishnaswamy & Reddy, 1994).
Social Capital and Networks for Small Scale Enterprises One of the major barriers for small scale enterprises is their inability to access and influence other organisations and institutions. Organisations and institutions often do not reach or work for the poor. They are usually developed through processes where the poor have not played a role and their interests are therefore not accounted for. Market institutions also generally suit the powerful and the modern sector. Elite groups tend to hijack the benefits of projects and programmes. (North, 1995) explains that as strong groups perceive changes in relative prices, technologies and transaction costs they modify institutions to suit their own interests. The relative power of different groups of people and their perception of possible effects of alternative paths of institutional and technological development are crucial in determining the direction in which change will occur. Social capital affects the access enterprises have to institutions and organisations and therefore markets and technology, as well as the distribution of earnings between buyers and
176 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
sellers and also within households and enterprises (Overseas Development Institute, 1999). Support for collective action is therefore advocated (Nabli & Nugent, 1989; Romijn, 2001). When people organise in groups they are more likely to be able to effect change, but their ability to form groups is also dependent on their stock of social capital.
5.7.7. The Role of Policies in Influencing Learning Technological change is considered to have characteristics of a ‘public good’3, since complementarities and positive externalities; such as inter-enterprise and inter-industry spillovers and benefits obtained through learning by using and interacting give technological capacity additional value that may not be accounted for by the market. Because of its public good nature, individual profit seeking behaviour is not efficient and technological change is therefore subject to market failures. Private enterprises will tend to under-invest in technological effort, as they may only be able to capture a portion of the total returns on their investment due to spillovers. Markets do not reward them sufficiently for the high outlays and risks involved in technological effort (Romijn, 2001). Technological inertia also arises because new technological systems are not well developed to supply the requirements of new technologies; and technological ‘lock-in’ arises due to previous investments and vested interests in technology infrastructure and institutions for existing technologies, which may not suit the incumbent technologies. This phenomenon has been observed in the case of fossil fuel technologies (Shum & Watanabe, 2006; Unruh, 2000), for which a suitable institutional structure has been built over 100 years. Institutional change (the emergence of new institutions and the disappearance of old ones) is usually slower than technical change, and technical change is classically regarded as the primary driver of institutional change (Edquist & Johnson, 1997). Governments and non-market institutions therefore have a role to play in creating institutional arrangements that encourage enterprises to invest in technological development. Selective support for technology is commonly advocated since countries have limited resources and must choose where they will develop their competitive advantage (Lall, 1992; Sharif, 1992). Environmentally beneficial technologies, such as PV, suffer from market failures in relation to the avoidance of negative environmental externalities in addition to positive technological externalities. In the face of these externalities, most commonly, deployment policies are used to promote learning-by-doing by supporting market growth while R&D investments directly support learning-by-searching (IEA, 2000; Jacobsson et al., 2002; Watanabe et al., 2000).
3 Public goods are goods which are not depleted by use, so after consumption by one individual, are available for another’s consumption. In the case of pure public goods, it is not possible to exclude people from using the good. The social benefits of public goods generally outweigh their private costs. 177 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.7.7.1. Public Investment in Knowledge Creation Governments and industry or interest groups may support the creation of new knowledge by directly funding R&D and appropriate education within private or public organisations. NGOs may also fulfil this role in the small scale sector. Public organisations may engage in collaborative research with, or otherwise share knowledge with enterprises. Governments may also use rules for R&D funding to encourage collaborations between research institutes and industry, and to encourage the involvement of particular types of enterprises in the development of a technology. Governments encourage private investment in learning by providing R&D funding that reduces the cost of the R&D investment required of enterprises and hence make the risks and spillovers more attractive. In the successful Korean electronics industry, the government invested in R&D in the early stages, setting up institutes for basic research and prototype construction (Kim, 1997). Hobday (1995) believes that the importance of these direct interventions, such as technology institutes and programs are unclear. However, supportive government policies towards a technology evidenced by such interventions may provide a favourable environment for enterprises to pursue R&D in the technological area.
5.7.7.2. Government Resource Allocation Governments may alter incentives to innovate via the allocation of resources and in the allocation of property rights. They may distribute resources via tax arrangements, or offer cheap land or services, such as electricity or water, to increase the profitability and the expansion capabilities of enterprises in a particular industry.
5.7.7.3. The Role of Government Intervention in Markets Governments may intervene in the market by providing subsidies to customers, artificially setting prices for products, imposing tariffs on competitors, or through procurement. Governments may also protect certain industries from international competition, ensuring markets for domestic enterprises. Market interventions generally aim to expand production; thereby facilitating scale economies, learning by doing and the entry of more players into the industry value-chain, which in turn bring with them knowledge and resources, providing more opportunities for learning by interacting. Governments may also intervene in markets, either through subsides or through direct procurement, to provide critical services to the poor at an affordable price. These interventions provide similar learning opportunities. Government intervention may also provide guidance with respect to the growth potential of a new technology, which may be closely linked to the legitimacy of it. These interventions are often selective, and in many cases are designed to select the modern sector, and particularly the export sector, and especially large enterprises (Tybout, 2000).
178 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Protectionism is criticised for causing distortions in investment, resulting in the non- optimal allocation of resources. Decision makers are not necessarily in possession of complete information and may make inappropriate decisions. Neoclassical approaches to technology policy include strategies such as getting prices right, encouraging free international trade in capital and technology and reduced government intervention, including reduced protection and privatisation (Lall, 1992). There is a consensus that structural adjustment reform policies have, however, overemphasised the issues of pricing, and provided insufficient support for technological development and diffusion in the face of market failures (Lall, 1992). The need for policy interventions in order to improve incentives is accepted in cases where the market does not allocate sufficient resources to learning. Lall (1992) believes that when the market failure is external to the enterprise, such as deficient national human capital, infrastructure or institutions, protection will have no impact in improving competitiveness in the long term. It is when the protection gives the enterprise time to overcome its own lack of investment in capability building that intervention may restore efficient resource allocation. Policies that provide the necessary resources are therefore needed in parallel with protectionism.
179 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.8. A Capability Building Framework
The functions of and relationships between actors, networks, institutions and technological trajectories in technological systems have been established. Figure 5-9 brings these elements together and illustrates how the enterprise level learning framework fits into the technological system.
Figure 5-9: A Framework for the Analysis of Capability Building
TECHNOLOGICAL TRAJECTORY
influence perceptions of the provide direction for technological paradigm search
Technological System
Networks
OTHER ACTORS influence policy and institutions
MARKET & NON-MARKET INTERACTIONS influence operation of markets influence connectivity INSTITUTIONS
knowledge creation & exchange resources for production and innovation investment opportunities provide incentives to invest and improve incentives alter allocation of resources
ENTERPRISE
Enterprise
180 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
The framework describes: The roles of networks of actors in providing the resources, opportunities and incentives for production and innovation to enterprises and in facilitating knowledge creation and exchange between actors. The ways the institutions can influence the operation of and connectivity in networks and alter the allocation of resources and incentives. The ways that networks and actors can in turn influence institutions and the technological trajectory.
At the enterprise level, the framework identifies three types of observable learning processes: learning by doing, searching and interacting; links each to the capabilities upon which they depend, and to the types of capabilities that are most commonly built through them in a path dependent manner. Improvement capabilities that allow enterprises to coordinate their activities are also identified. The ability of enterprises to interact with other actors viewed as being dependent on their linkage and investment capabilities. Through these interactions, they can access the resources and opportunities for production and innovation. Dotted lines indicate incentives, resources and direction for innovation being provided by institutions and the technological trajectory.
181 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
5.9. Conclusion
The studies on learning in the PV industry reviewed in chapter 3 provide insight into the character and complexity of the technologies for PV cell manufacture, module assembly and balance of systems manufacture, and the rate and types of technological change that are likely to occur in the cell manufacturing industry. These studies do not, however, identify the technological capabilities enterprises require, how they can build these capabilities, or the factors in technological systems that impact learning. An analytical framework has been proposed which identifies the factors that influence capability building. The framework adapts concepts and combines existing frameworks that are primarily drawn from two bodies of learning relating to technological learning: the learning literature and the innovation systems literature. The process of adaptation has been informed by insights from: The technological capabilities literature, which highlights the technology transfer processes whereby latecomers catch up technologically and eventually gain technological independence; The literature on technology in small scale enterprises, which includes the hardware- focussed appropriate technology literature, and also that related to small and rural non- farm enterprises, which is more focused on institutions, linkages and complementary capabilities; and Observations of capability building from the case study field work.
Through consideration of this literature and these observations, the author has been able to select and combine elements such that the framework can usefully explain the factors that influence capability building in both small scale and modern sector enterprises, with particular attention paid to the development of manufacturing capabilities and especially those required for relatively complex technologies, such that the framework is suited to the assessment of PV manufacturing enterprises. The framework captures the following key elements of capability building processes and systems: Learning processes; Types of capabilities that can be built through various learning processes; The capabilities upon which learning processes depend; The factors in the technological system that enable capability building; and The ways that actors can influence institutional arrangements.
182 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
The framework will be applied in subsequent chapters to identify barriers to capability building, suitable capability building strategies and critical factors in institutional environments for PV manufacturing enterprises in developing countries through the investigation of the cases of Suntech Power, Grupo Fénix and the Barefoot College. The analysis will serve to assess the value of the framework in guiding decision making in relation to PV manufacture in developing countries, and will enable a preliminary identification of typical barriers to PV manufacture in developing countries and useful capability strategies and interventions to support PV manufacturers.
183 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
References
Albu, M. (1997), Technological Learning and Innovation in Industrial Clusters in the South, MSc in Science and Technology Policy Thesis, Science Policy Research Unit, University of Sussex, Brighton, U.K. Albu, M. (2001), International Strategy for ITDG’s work with Manufacturing Enterprise: Improving Livelihood Opportunities from Small-scale Manufacturing and Processing Enterprises, Intermediate Technology Development Group. Albu, M. and Scott, A. (2001), Understanding livelihoods that involve micro-enterprise: markets and technological capabilities in the SL framework, Intermediate Technology Development Group. Allal, M. (1999), Working Paper 2: International best practice in micro and small enterprise development, ILO/UNDP Micro and Small Enterprise Development and Poverty Alleviation Project in Thailand, Finnegan, G. (ed), International Labour Organization / United Nations Development Project. Argote, L. and Epple, D. (1990), Learning Curves in Manufacturing, Science, 247 (4945), p 920. Ariffin, N. and Figueiredo, P.N. (2002), Internationalisation of Innovative Capabilities: Evidence from the Electronics Industry in Malaysia and Brazil, work paper derived from PhD research program, SPRU, University of Sussex, Sussex, U.K. Arrow, K. (1962), The economic implications of learning by doing, The Review of Economic Studies, 29 (3), pp 155-173. Bell, M. (1984), 'Learning' and the Accumulation of Industrial Technological Capacity in Developing Countries, in Fransman, M. & King, K. (eds), "Technological capability in the Third World", St. Martin's Press, New York, pp 187-209. Bell, M. and Pavitt, K. (1995), The Development of Technological Capabilities, in Haque, I.U. (ed), "Trade, Technology and International Competitiveness", The World Bank, Washington D.C., USA, pp 69-101. Bennell, P. (1999), Learning to change: Skills development among the economically vulnerable and socially excluded in developing countries, Employment and Training Papers: 43, International Labour Organization, Geneva. Bhalla, A.S. and Reddy, A.K.N. (1994), The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme, Intermediate Technology Pub., London. Carlsson, B. and Stankiewicz, R. (1991), On the nature, function and composition of technological systems, Journal of Evolutionary Economics, 1 (2). Carney, D. (1999), Approaches to Sustainable Livelihoods for the Rural Poor, Policy Briefing no 2, Overseas Development Institute, U.K. Cowan, R., David, P.A. and Foray, D. (2000), The explicit economics of knowledge codification and tacitness, Industrial and Corporate Change, 9 (2), p 211. Dawson, J. and Jeans, A. (1997), Looking Beyond Credit: Business development services and the promotion of innovation among small producers, ITDG Publications, London, U.K. de Wilde, T. (1991), Small Business Development: Changing paradigms and contests, in Stevens, R.W. (ed), "Appropriate technology : a focus for the nineties", Intermediate Technology Development Group of North America, New York, p 98. Desai, A.V. and Taneja, N. (1990), The Role of Small and Medium-Scale Industries in the Industrial Development of India', in ADB, The Role of Small and Medium-Scale Manufacturing Industries in Industrial Development: The Experience of Selected Asian Countries, Asian Development Bank, Manila, Phillipines. DFID (undated), DFID Sustainable Livelihoods Guidance Sheets, Accessed from: http://www.livelihoods.org/info/info_guidancesheets.html#7, on: July 2005.
184 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Dorward, A., Kydd, J., Morrison, J. and Poulton, C. (2005), Institutions, markets and economic co-ordination: Linking development policy to theory and praxis, Development and Change, 36 (1), pp 1-25. Dorward, A., Poole, N., Morrison, J., Kydd, J. and Urey, I. (2003), Markets, Institutions and Technology: Missing Links in Livelihoods Analysis, Development Policy Review 21 (3), pp 319-332. Dosi, G. (1982), Technological paradigms and technological trajectories : A suggested interpretation of the determinants and directions of technical change, Research Policy, 11 (3), pp 147-162. Dosi, G. (1988), Sources, procedures and microeconomic effects of innovation, Journal of Economic Literature, 26 (3), pp 1120-1171. Douthwaite, B. (2002), Enabling Innovation: A Practical Guide to Understanding and Fostering Technological Change, Zed Books in Association with CAMBIA, London and New York. Dutrénit, G. (2004), Building Technological Capabilities in Latecomer Firms: A Review Essay, Science, Technology & Society, 9 (2). Dutton, J.M. and Thomas, A. (1984), Treating progress functions as a managerial opportunity, Academy of Management The Academy of Management Review (pre-1986), 9 (000002), p 235. Edquist, C. (1997), Systems of Innvovation Approaches - Their emergence and characteristics, in Edquist, C. (ed), "Systems of innovation : technologies, institutions, and organizations", Pinter, London ; Washington, pp 1-40. Edquist, C. and Johnson, B. (1997), Institutions and Organisations in Systems of Innovation, in Edquist, C. (ed), "Systems of innovation : technologies, institutions, and organizations", Pinter, London ; Washington, pp 41-63. Fernando, S. (1992), Chapter 3: Technology Transfer for Small and Cottage Industries: The Sri Lankan Experience, in Islam, R. (ed), "Transfer, Adoption and Diffusion of Technology for Small and Cottage Industries", Asian Regional Team for Employment Promotion, World Employment Programme, ILO, New Delhi, India, pp 74-108. Figueiredo, P.N. (2002), Does technological learning pay off? Inter-firm differences in technological capability-accumulation paths and operational performance improvement, Research Policy, 31 (1), pp 73-94. Fonseca, G. (2007), The Production Function, The History of Economic Thought Website, Department of Economics of the New School for Social Research, Accessed from: http://cepa.newschool.edu/het/essays/product/prodfunc.htm, on: March 2007. Fransman, M. (1984), Technological Capability in the Third World: An Overview and Introduction, in Fransman, M. & King, K. (eds), "Technological capability in the Third World", St. Martin's Press, New York, pp x, 404 p. Fujimoto, T. (2001), Automobiles: Strategy-based Lean Production System, CIRJE Discussion Paper, Tokyo University, Tokyo, Japan. Garud, R. (1997), On the Distinction Between Know-How, Know-What and Know-Why, Advances in Strategic Management, 14, pp 81-101. Gass, G., Biggs, S. and Kelly, A. (1997), Stakeholders, science and decision making for poverty-focused rural mechanization research and development, World Development, 25 (1), p 115. Håkansson, H. (1989), Corporate technological behaviour : co-operation and networks, Routledge, London ; New York. Harper, M. (1996), Empowerment Through Enterprise, Intermediate Technology Publications Ltd., London, U.K. Hobday, M. (1995), East Asian latecomer firms: Learning the technology of electronics, World Development, 23 (7), p 1171. Hobday, M. (2000), Chapter 5: East versus Southeast Asian Innovation Systems: Comparing OEM- and TNC-led Growth in Electronics, in Kim, L. & Nelson, R.R. (eds), "Technology, Learning and Innovation: Experiences of Newly Industrializing Economies", Cambridge University Press, Cambridge, U.K.
185 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
IEA (2000), Experience Curves for Energy Technology Policy, OECD/IEA, France. Ingham, B. (1993), The meaning of development: Interactions between "new" and "old" ideas, World Development, 21 (11), p 1803. IPCC (2001), Managing Technological Change, An explanatory summary of the IPCC Working Group III Special Report: Methodological and Technological Issues in Technology Transfer, United Nations Environment Programme Division of Technology, Industry and Economics. Islam, R. (1992), Chapter 1: Transfer, Dissemination and Adoption of Technology for Small and Cottage Industries: An Overview, in Islam, R. (ed), "Transfer, Dissemination and Adoption of Technology for Small and Cottage Industries", Asian Regional Team for Employment Promotion, World Employment Programme, ILO, New Delhi, India. Jacobsson, S., Andersson, B.A. and Bångens, L. (2002), Transforming the energy system - the evolution of the German technological system for solar cells, SPRU Electronic Working Paper Series Paper No 84, Science and Technology Research, University of Sussex, Brighton, U.K. James, J. (1989), Improving traditional rural technologies, Macmillan, Basingstoke. Juma, C. and Clark, N. (2002), Technological Catch-Up: Opportunities and Challenges for Developing Countries, SUPRA Occasional Paper, Research Centre for the Social Sciences, University of Edinburgh, Edinburgh. Kamp, L.M. (2002), Learning in wind turbine development: A comparison between the Netherlands and Denmark, PhD Thesis, Universiteit Utrecht. Kathuria, S. (1992), Chapter 4: Transfer of Small-Scale Industrial Technology from India to Africa and other Less Developed Regions - Experience and Potential, in Islam, R. (ed), "Transfer, Adoption and Diffusion of Technology for Small and Cottage Industries", Asian Regional Team for Employment Promotion, World Employment Programme, ILO, New Delhi, India, pp 109-163. Katz, J.M. (1984), Technological Innovation, Industrial Organisation and Comparative Advatages of Latin American Metalworking Industries, in Fransman, M. & King, K. (eds), "Technological Capability in the Third World", St. Martin's Press, New York, pp 113-136. Kim, L. (1997), Imitation to innovation: the dynamics of Korea's technological learning, Harvard Business School Press, Boston, USA. Kim, L. and Nelson, R.R. (2000), Introduction, in Kim, L. & Nelson, R.R. (eds), "Technology, Learning and Innovation: Experiences of Newly Industrializing Economies", Cambridge University Press, Cambridge, U.K. Krishnaswamy, K.N. and Reddy, A.K.N. (1994), The Commercialization of Improved Technologies in Rural Areas, in Bhalla, A.S. & Reddy, A.K.N. (eds), "The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme", Intermediate Technology Pub., London, p 23. Kumar, D. (1994), The Role of Non-Governmental Organisations in the Diffusion of Rural Technologies, in Bhalla, A.S. & Reddy, A.K.N. (eds), "The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme", Intermediate Technology Pub., London, p 133. Lall, S. (1990), Building industrial competitiveness in developing countries, Development Centre of the Organisation for Economic Co-operation and Development; OECD Publications and Information Centre, distributor, Paris, France; Washington, D.C. Lall, S. (1992), Technological capabilities and industrialization, World Development, 20 (2), p 165. Lall, S. (2000a), Technological Change and Industrialization in the Asian Newly Industrializing Economies: Achievements and Challanges, in Kim, L. & Nelson, R.R. (eds), "Technology, Learning and Innovation: Experiences of Newly Industrializing Economies", Cambridge University Press, Cambridge, U.K., pp 13-68.
186 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Lall, S. (2000b), The Technological Structure and Performance of Developing Country Manufactured Exports, 1985-1998, QEH Working Paper Series – QEHWPS44, Queen Elizabeth House, Oxford University, Oxford, U.K. Lall, S. (2003), Investment And Technology Policies For Competitiveness: Review of successful country experiences, Technology for Development Series, United Nations Conference on Trade and Development, United Nations, New York and Geneva. Lundvall, B. (1992), National systems of innovation : towards a theory of innovation and interactive learning, Pinter Publishers; St. Martin's Press, London, New York. Malerba, F. (1992), Learning by Firms and Incremental Technical Change, The Economic Journal, 102 (413), pp 845-859. Marcelle, G.M. (2002), Reconsidering Conventional Wisdom on Technology Transfer, International Conference on Science, Technology and Innovation Emerging International Policy Issues, Harvard University, Cambridge MA, September 2002. Marcotte, C. and Niosi, J. (2000), Technology Transfer to China: The Issues of Knowledge and Learning, Journal of Technology Transfer, 25 (1), pp 43-57. Masini, A. and Frankl, P. (2003), Forecasting the diffusion of photovoltaic systems in southern Europe: A learning curve approach, Technological Forecasting and Social Change, 70 (1), p 39. Masum, M. (1992), Chapter 2: The Bangladesh Experience, in Islam, R. (ed), "Transfer, Dissemination and Adoption of Technology for Small and Cottage Industries", Asian Regional Team for Employment Promotion, World Employment Programme, ILO, New Delhi, India. Mishra, S.C. (1994), The Small-Scale Farm Machinery Industry in Uttar Pradesh, in Bhalla, A.S. & Reddy, A.K.N. (eds), "The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme", Intermediate Technology Pub., London, p 201. Nabli, M.K. and Nugent, J.B. (1989), The New Institutional Economics and its applicability to development, World Development, 17 (9), p 1333. Nelson, R.R. and Winter, S. (1982), An Evolutionary Theory of Economic Change, Harvard University Press, Cambridge, U.S.A. Nonaka, I. (1994), A Dynamic Theory of Organizational Knowledge Creation, Organization Science: A Journal of the Institute of Management Sciences, 5 (1), p 14. North, D.C. (1995), The New Institutional Economics and Third World Development, in Harriss, J., Hunter, J. & Lewis, C. (eds), "The New Institutional Economics and Third World Development", Routledge, London, U.K. Overseas Development Institute (1999), Social Capital, Keysheets for Sustainable Livelihoods (Last Updated: May 1999), Accessed from: http://www.keysheets.org/, on: August 2005. Panditrao, Y.A. (1994), Experiences of the Khadi and Village Industries Comission in Technology Transfer, in Bhalla, A.S. & Reddy, A.K.N. (eds), "The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme", Intermediate Technology Pub., London, p 113. Pavlic, B. (1983), The Challenges of south-south cooperation, Westview Press, Boulder, Colorado, USA. Piek, H. (1998), Technology Development in Rural Industries - A study of China's collectives, Intermediate Technology Publications Ltd., London, U.K. Polanyi, M. (1967), The Tacit Dimension, Doubleday, New York, U.S.A. Pretty, J. and Ward, H. (2001), Social Capital and the Environment, World Development, 29 (2), p 209. Rogers, E.M. (2003), Diffusion of innovations, 5th ed, Free Press, New York., NY. Romijn, H. (2001), Technology Support for Small-scale Industry in Developing Countries: A Review of Concepts and Project Practices, Oxford Development Studies, 29 (1). Rosenberg, N. (1982), Inside the black box : technology and economics, Cambridge University Press, Cambridge [Cambridgeshire] ; New York.
187 Chapter 5. A Framework for the Analysis of Capability Building in Developing Countries
Sanden, B.A. (2005), The economic and institutional rationale of PV subsidies, Solar Energy, 78 (2), pp 137-146. Saxena, N.C. (2003), The Rural Non-Farm Economy in India: Some Policy Issues, Rural Non- Farm Economy and Livelihood Enhancement, DFID-World Bank Collaborative Research Project, Natural Resources Institute. Schaeffer, G.J., Alsema, E., Seebregts, A., Beurskens, L., de Moor, H., van Sark, W., Durstewitz, M., Perrin, M., Boulanger, P., Laukamp, H. and Zuccaro, C. (2004), Learning from the Sun Analysis of the use of experience curves for energy policy purposes: The case of photovoltaic power, Final report of the Photex project: Report ECN-C-04-035, ECN Renewable Energy in the Built Environment. Schumpeter, J. (1961 {1911}), Theory of Economic Development (English Translation), Oxford University Press., New York. Sharif, N. (1992), Technological dimensions of international cooperation and sustainable development, Technological Forecasting and Social Change, 42 (4), p 367. Shum, K.L. and Watanabe, C. (2006), Photovoltaic deployment strategy in Japan and the USA-- an institutional appraisal, Energy Policy, In Press, Corrected Proof. Stewart, F. (1978), Technology and underdevelopment, 2d ed, Macmillan, London. Teece, D.J. (2000), Firm Capabilities and Economic Development: Implications for newly industrializing economies, in Kim, L. & Nelson, R.R. (eds), "Technology, Learning, and Innovation", Cambridge University Press, Cambridge, UK. Tybout, J.R. (2000), Manufacturing Firms in Developing Countries: How Well Do They Do, and Why?, Journal of Economic Literature, 38 (1), pp 11-44. UNIDO (2005), Capability building for catching-up: Historical, empirical and policy dimensions, Industrial Development Report 2005, United Nations Industrial Development Organisation, Vienna, Austria. Unruh, G.C. (2000), Understanding carbon lock-in, Energy Policy, 28 (12), pp 817-830. van der Zwaan, B. and Rabl, A. (2003), Prospects for PV: a learning curve analysis, Solar Energy, 74 (1), p 19. van der Zwaan, B. and Rabl, A. (2004), The learning potential of photovoltaics: implications for energy policy, Energy Policy, 32 (13), p 1545. Vania, F. and Taneja, B. (2005), People, Policy, Participation: Making Watershed Management work in India, International Institute for Environment and Development (IIED), London. U.K. Watanabe, C., Wakabayashi, K. and Miyazawa, T. (2000), Industrial dynamism and the creation of a "virtuous cycle" between R&D, market growth and price reduction: The case of photovoltaic power generation (PV) development in Japan, Technovation, 20 (6), p 299. Wilson, S.R., Ballance, R. and Pogany, J. (1995), Beyond Quality: An agenda for improving manufacturing capabilities in developing countries, Edward Elgar Publishing, Cambridge, U.K.
188 CChhaapptteerr 66.. AA VViirrttuuaall PPrroodduuccttiioonn LLiinnee ffoorr tthhee MMaannuuffaaccttuurree ooff SSccrreeeenn--PPrriinntteedd SSoollaarr CCeellllss
This chapter describes the development, use and effectiveness of the Virtual Production Line, a software simulation tool for the training of production line engineers that was co- developed by Stuart Wenham and the author at the University of New South Wales between 2002 and 2007. A copy of the Virtual Production Line (VPL) software is contained on a CD in Appendix 1. Section 6.1 discusses the need for training tools in the PV industry and describes the design and development of the VPL to respond to this need. Section 6.2 details the manufacturing processes modelled by the VPL, describing the parameters that can be controlled by the user and the impact they have on the solar cells being produced. In section 6.3, the quality control tests available in the VPL which enable the user to obtain feedback on the effects of varying the processing parameters are described. Educational features of the VPL are described in section 6.4, including help files, assignments, and the ability to export the data from processes batches for analysis. In section 6.5, the effectiveness of the VPL as a training tool and its role in capability building is considered through the application of the framework developed in chapter 5. The operation of solar cells and the equation which governs their current voltage (IV) characteristics are described in Appendix 3. The reader may refer to this appendix and the help files contained in the VPL CD in Appendix 1 for a more detailed explanation of the details of solar cell operation and manufacture than is contained in this chapter.
189 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
6.1. The Value and Development of the Virtual Production Line
As the photovoltaics manufacturing industry has grown rapidly, considerable job creation has occurred. The European Photovoltaics Industry Association (EPIA) and Greenpeace (2007) reported that 10 jobs are created in manufacturing per MWp of PV production. They also reported that in 2006, there were 48,000 PV manufacturing jobs worldwide, and a total of 73,000 jobs in the PV industry. The ‘moderate scenario’ of industry growth in their report predicted that over 400,000 jobs will be created in PV manufacturing by 2030 and a total of almost 3 million jobs in the industry as a whole. There is clearly a large and growing need for skilled manufacturing personnel in the PV industry. There is an inadequate understanding of quality control tests (Haase, 2005). The dissemination of PV manufacturing knowledge through international conferences is, limited, since they are mainly dedicated to basic research and applications. The first international conference dedicated to mass production of PV was only held in 2005 (Koch et al., 2005). The training of manufacturing engineers and technicians is challenging for educational institutions, because it is difficult to gain access to commercial facilities due to commercial sensitivities and it is expensive and impractical to duplicate high volume production lines in educational institutions. Even laboratory or pilot scale production facilities are too expensive to equip and run and are too delicate for educational purposes. Without any hands-on experience, manufacturing concepts such as process optimisation, parameter tolerances, yield, throughput, in-line quality control, fault diagnosis and reliability are therefore difficult to teach (Wenham et al., 2002). Technology transfer of complex technologies require extensive training and may take a long time, depending on the R&D capability, manufacturing experience and level of prior related knowledge of the recipient. Traditional technology transfer methods such as instructor presentation of lectures, demonstrations, visiting production facilities and on the job training require large investments in travel, human resources and time, which may not be affordable for firms in developing countries. A commercial solar cell production line is usually not available for training the engineers since the line must be kept available for producing solar cells, and the processing parameters therefore cannot be altered from the settings currently being used for cell production. Given the concentration of the industry in a few countries and the difficulty of training engineers, the shortage of skilled personnel is likely to be a significant barrier to manufacturers in developing countries. The Virtual Production Line (VPL) is a software package, developed by the author and Stewart Wenham as a teaching and training tool for university students and
190 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
production line engineers and technicians, which has the potential to mitigate some of the problems of training production line engineers. The VPL is based on varying a set of variables which describe the characteristics of a solar cell as the cell progresses through the steps in the manufacturing sequence. In each process, the VPL performs algorithms which determine the changes in the physical properties of the solar cell on the basis of the processing parameters used during its manufacture. A simulation relies on the ability to construct a simplified model of sufficient accuracy. Often a simulation is based on modelling of the physical parameters and known fundamental relationships between them. In this case, these interactions in the production environment are too complex. It has been found to be more practical to base the simulation is based on real world data describing the interactions between controllable production line parameters and the resulting device parameters. A large amount of data from three different commercial production lines has been collected and analysed and the algorithms have been developed for each process to reflect the impact on the properties of the solar cell of variations in the processing parameters and relationships between processing parameters for each process, as well as the interdependencies of parameters between different processes. The algorithms are contained in Appendix 4. Extensive testing of the VPL and comparison with data from real production lines has demonstrated that it accurately simulates the impact and interactions of processing parameters over a reasonable range of processing parameters. In order to account for the things which can’t be controlled, such as variations in the properties of materials, the operation of equipment, statistical variability is built into the system. In order to analyse the performance of the virtual solar cells at various stages in their manufacture, an extensive array of quality control tests which are used in commercial production lines have been included. These tests both enable the user to observe the effects of varying processing parameters and train the user in the use of and importance of these tests in the operation of production lines. While some of these tests could be implemented simply by returning the stored values which describe the characteristics of the device, others produce graphs which rely on simulation of the operation of the device. In order to achieve this, the VPL runs a modified version of a two-dimensional device modelling package called PC1D (Basore, 1990). The VPL supplies PC1D with the device parameters, which are then used to generate IV curves, graphs of reflection and QE spectral response as a function of wavelength, spatially resolved doping profiles and minority carrier lifetimes for the cells. The tests available in the VPL are further described in section 6.3. In order to achieve its educational objectives, VPL also contains assignments which isolate some of the process and processing parameters from the rest of the sequence and require the user to optimise either average device performance, best device performance or
191 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
manufacturing yield. The assignments are further described in section 6.4.3. It also contains extensive help files, which explain in detail each of the processes, the impact of each parameter and parameter interdependencies. Videos allow students to get a feel for the type of equipment used in commercial production lines.
192 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
6.2. Processing Virtual Solar Cells with the Virtual Production Line
The VPL software allows the user to create a batch of up to 100 cells for processing. The user must progress in one of two standard sequences through each of the processes in the manufacture of a screen-printed solar cell. In the first sequence the front and rear metal contacts are fired separately, and the order of processing is as follows: Saw Damage Etch b Rinse b Texture b Rinse/Acid Clean b Diffusion b Diffusion Oxide Removal b Al Screen-Printing and Firing b Plasma Edge Isolation b Antireflection Coating Deposition b Silver Screen Printing and Firing
In the second sequence, the contacts are fired together (co-fired), and the order of processing is as follows: Saw Damage Etch b Rinse b Texture b Rinse/Acid Clean b Diffusion b Diffusion Oxide Removal b Plasma Edge Isolation b Antireflection Coating Deposition b Silver Screen Printing b Al Screen-Printing b Co-firing of the Contacts
When a virtual batch of wafers is created, a set of parameters, which are altered as the wafer goes through each of the processing steps is initialised. The most important of these parameters are listed in Table 6-1. They are used to characterise the solar cell collection of parameters describing the physical properties and device structure of a solar cell is created and varied according to the parameters in each of the processing steps in the VPL.
193 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Table 6-1: Parameters Provided to PC1D by the Virtual Production Line Symbol Description Default A Area 1cm2 BBR Front broad band reflectance 0% IT AR coating layer thickness 80nm RS Rear surface reflectance 64% RE Internal series resistance (emitter) 1ohm RB Internal series resistance (base) 1ohm RSH Shunt resistance 0.002siemens SD Shunt diode current 1×10-7 T1 Region 1 thickness 0.01um BD Background doping resistivity 1ohmcm PD1 Region 1 front diffusion peak doping 3×1020 cm-3 JD1 Region 1 front diffusion junction depth 0.4um FSRV Effective minority carrier front surface recombination velocity - low level 1×105 cm/s injection T2 Region 2 thickness 250um BR Net low level injection excess carrier lifetime 20us JD2 Region 2 junction depth 0.2um
The VPL allows the user to specify the type of silicon wafer (single crystal or multicrystalline), its thickness, resistivity, the wafer slicing process used (wire sawn or inner diameter diamond tipped saw) and the quality of the wafers to be processed (Table 6-2). The wafer thickness, bulk resistivity, surface damage to the wafers during the sawing process and minority carrier lifetime determined by the wafer quality are stored in variables which describe the characteristics of the wafers and are used to determine how they respond to the remainder of the processing.
Table 6-2: Incoming Wafer Options Variable User Input Units Default Uncertainty WaferType Wafer type single crystal SWT Wafer Thickness Microns 300 +/-2% (BD) Wafer Resistivity ohms/cm3 1 +/-2% (SWWD) Wafer Slicing Process diamond tipped 15/10 saw Quality Wafer Quality (poor, standard, good good, semiconductor, float zone)
For each of the manufacturing processes, the user is able to control a number of parameters for the execution of each of the processes via slider bars which appear in a panel on the right hand side of the user interface, which is shown in Figure 6-1. The details of these parameters will be described for each process in the following sections.
194 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Figure 6-1: VPL User Interface for the Acid Clean and Rinse Process, which is performed after Texturing of the Wafers
When the user presses the button to run a process, a brief video of the process occurring in a commercial production line is played. At the conclusion of the video, the batch processing history is updated in the left hand panel of the user interface, and the right hand panel becomes ready for the next process in the sequence. There are some user-controlled variations to the standard sequence. After the etching process, the user is given the option to carry out texturing or not, after the diffusion, the user chooses between separate and co-firing of the contacts, and after the plasma edge isolation process, the user may choose to apply a titanium dioxide antireflection coating, a silicon nitride (SiNx) antireflection coating, or not to apply an antireflection coating (note that the version of the VPL contained in Appendix 1 does not include the SiNx antireflection coating, which is in the final stages of development). On completion of a batch, an IV curve (see Appendix 3 for a description of the IV curve) is created for each cell, and the important statistics describing the performance of the virtual solar cell are displayed in the right hand panel of the user interface.
6.2.1. Saw Damage Removal Etch A saw-damage removal etch is used to remove surface damage and to clean contaminants from the wafer prior to processing. In this process, the VPL allows the user to control the concentration of the etchants in the baths, the temperature of the solution, and the time the wafers spend in the baths (Table 6-3). A random number generator is used to vary the
195 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
etch temperature between ± 2ºC of the user specified bath temperature in order to introduce some uncertainty to the process which reflects conditions in the operation of a real production line.
Table 6-3: Saw Damage Removal Etch Variables Variable User Input Units Min Max Default Uncertainty (ET) etchTemp °C 0 90 40 ± 2°C EC etchChemConc % 20 40 30 Et etchTime mins 0 40 20 SSB ssBatches
The user specified variables, along with the amount of sodium silicate in the bath, which is determined by the number of batches previously etched in the solution, determine the amount of material etched from the surface of each the virtual wafers according to the algorithm contained in Appendix 4. The thickness of the wafer is adjusted, and stored for use in future processing and characterisation of the virtual solar cell. Insufficient etching may leave surface damage or contaminants on the wafer, while too much etching removes some of the useful silicon and makes the wafer thinner and therefore more vulnerable to breakage during the remainder of processing. A variable stores the probability of breakage of the wafer in each of the remaining processes and a random number generator is used to determine whether the wafer has broken or not.
6.2.2. Texturing Texturing reduces reflection form the front surface of single crystal wafers. The preferred crystal orientation for monocrystalline wafers is 1-0-0, which allows an anisotropic etch to preferentially etch the silicon material in the directions of the planes that have less strength and leaves the 1-1-1 planes exposed, as they have the highest bond density and thus are stronger than the other planes. The upright tetrahedral pyramidal structure that results (Figure 6- 2a) enables more light hitting the surface of the solar cell to be absorbed, rather than reflected, as it offers multiple opportunities for the light to be reflected onto the wafer surface. Reflection can be reduced from 33% to around 12% by texturing monocrystalline wafers (Wenham et al., 2006). While still useful, texturing processes are not as effective on multicrystalline wafers due to the random nature of the crystals within such wafers. Alterative texturing processes that achieve similar effects and do not depend on the crystal orientation have also been developed for multicrystalline wafers.
196 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Figure 6-2: (a) Pyramids formed by texturing of single crystal silicon wafers, and (b) Solar cells are dipped into chemical baths in Teflon or Polypropylene cartridges.
(a) (b)
Texturing of single crystal wafers is carried out by etching the wafers in a solution of sodium hydroxide and isoproponal as illustrated in Figure 6-2b. Texturing is a complex process, as the pyramids must have good uniformity, a size of between 3-8 microns across the base, and be etched deeply enough to cover the entire wafer with sharp apexes on the pyramids in order to have maximum effect. The size and shape of the pyramids is determined by the concentration of isopopropanol and sodium hydroxide in the bath, the temperature and time wafers spend in the bath, the evaporation rate of propanol (determined by exhaust rate and cover on bath), the wafer surface finish (determined by wafer etch process), the time between batches, the sodium silicate in bath (determined by previous batches processed in the solution), the number of wafers in each batch and the turbidity of mixture. The algorithms governing these dependecies are contained in Appendix 4.
Table 6-4: Texturing Variables Variable User Input Units Min Max Default Uncertainty Tt textureTime mins 0 30 20 TT textureTemp °C 50 100 70 ± 2°C TC1 textureChemConc (NaOH) % 1 3 2 TC2 texturePropanol % 0 10 5 (TEFR) textureEfr (Effect of Soldium 0 4 2 Silicate)
6.2.3. Rinse and Acid Clean An acid clean and rinse is performed after texturing the wafer surface. Dilute hydrochloric acid is used to react with metal atoms, which if not removed can cause penetrate into the silicon material when the wafers are exposed to temperatures approaching 1000°C, causing charge carriers to recombine and reducing the performance of the solar cell. Hydrofluoric acid is used to remove any residual sodium silicate. The user controlled variables for this process are listed in Table 6-5.
197 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Table 6-5: Post Texture Acid Clean and Rinse Variables Variable User Input Units Min Max Default AHCL hclChemConc % 0 30 10 hclt hclTime mins 0 10 4 ARt1 racRinseTime1 mins 0 10 3 AHF hfChemConc % 0 30 10 hft hfTime mins 0 10 5 ARt2 rinseTime2 mins 0 10 5
If the acid cleans are insufficient, or the wafers are not rinsed properly the minority carrier lifetime of the wafer will be adversely affected according to the equations in Appendix 4. If the wafers are not rinsed sufficiently, the user also is warned that the operator is in danger.
6.2.4. Diffusion Solar cells (and microelectronics devices) are based on a p-n junction. The starting wafer is normally p-type and the n-type layer is usually created by diffusion. The wafer is placed in a chamber of gas containing the dopant atoms (usually phosphorous), or covered in a thin layer of material containing the dopant. With the application of high temperatures, the dopant atoms ‘diffuse’ a few microns into the material. At the high temperatures in the diffusion furnace, impurities can easily diffuse into the silicon since they become very mobile at these temperatures, degrading electrical performance. The environment, the wafers and the phosphorous source must therefore be extremely clean.
Figure 6-3: (a) Phosphorous Deposition, and (b) Diffusion
(b) (a)
User controlled variables such as the temperature in the diffusion furnace, the concentration of the phosphorous source, and the amount of time the wafers spend in the furnace will determine the doping density and the depth of the junction (Table 6-6). It is also important that the phosphorous source dries completely before entering the high temperature diffusion step. The equations which describe these dependencies can be found in Appendix 4.
198 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Table 6-6: Diffusion Variables Variable User Input Units Min Max Default Uncertainty DBS diffusionBeltSpeed cm/sec 0.1 1 0.5 DT1 diffusionDryingTemp °C 100 300 200 ± 4°C diffusionZone1Temp °C 500 800 600 DT2 diffusionZone2Temp °C 700 1000 1000 ± 4°C diffusionZone3Temp °C 500 800 700 diffusionGasFlow l/sec 0 50 25 (DDF) diffusionDilutionFactor % 10 100 100
Excess diffusion of phosphorous damages the silicon surface, which leads to an effective loss of any light absorbed close to the surface. Insufficient diffusion leaves the silicon with poor electrical conductivity, causing high resistive losses within the n-type silicon and potentially poor electrical contact to the metal that is applied later in the fabrication sequence.
6.2.5. Diffusion Oxide Removal A diffusion oxide removal is performed after the diffusion process in the VPL. In commercial production lines, it may be carried out either before or after plasma edge isolation. The diffusion oxide removal ensures that antireflection coatings are uniform and that the passivation benefits of SiNx antireflection coatings are obtained.
Table 6-7: Diffusion Oxide Removal Variables Variable User Input Units Min Max Default rac2CleanTime mins 0 10 5 rac2ChemConc % 0 30 10 rac2RinseTime mins 0 10 5
6.2.6. Plasma Edge Isolation Plasma edge isolation removes the sides of the wafer which have been exposed to the diffusion source during the diffusion process and therefore contain n-type material which acts as a shunt path for the generated current to flow to the back of the wafer, where it does not contribute to collected current. The wafers are first stacked so that only their edges are exposed. High voltages in the presence of gases creates a plasma of electrical discharge, which etches away the edges of the wafers. In this process, the user controls the power supplied to the plasma and the time that the wafers are in the plasma chamber (Table 6-8). If the power is too high, the wafers can be damaged. However, if the diffused region at the side of the wafer is not removed completely due to insufficient power or time spent in the plasma chamber, shunting will occur at the edges of the wafer.
199 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Table 6-8: Plasma Edge Isolation Variables Variable User Input Units Min Max Default Uncertainty (PEP) plasmaEtchPower int W 150 1500 700 (PEt) plasmaEtchTime int mins 2 20 10
6.2.7. Antireflection Coatings Antireflection (AR) coatings are comprised of a thin layer of dielectric material with a refractive index and thickness that is suitable to cause incoming light to the surface of a solar cell to be refracted, rather than reflected. They are particularly beneficial on multicrystalline wafers that can not benefit from the chemical texturing etch that forms pyramids beneficial to light trapping in the case of single crystal silicon wafers. An antireflection coating will usually be optimised for the most useful wavelengths of light: green, yellow and red. Blue light is not well utilised in solar cells, due to the dead layer cased by emitter formation, hence the light reflected from the AR coating is usually blue, giving the solar cell a blue appearance. The combination of AR coatings and texturing can result in reflection losses of around 1% in monocrystalline wafers (Wenham et al., 2006). Applying an AR coating prior to deposition of the metal contacts can also reduce the penetration of the top-contact silver into the wafer, hence allowing a shallower junction to be used, which is beneficial since the doping of the emitter creates a dead layer in the solar cell, particularly if the emitter is heavily doped. The thinner junction also allows for a more lightly doped emitter to be used, so that the negative effects of emitter doping are reduced.
Titanium dioxide (TiO2) is often deposited as a spray process. As well as affecting optical performance, non-uniformity in an AR coating is visible and may carry the implication of lower quality. If the dielectric is used with fire-through silver contacts, non-uniformity of the dielectric layer across the wafer surface may also lead to poor electrical performance of the solar cell. Atmospheric pressure chemical vapour deposition (APCVD) provides better uniformity and is hence a preferable method of deposition, particularly in the case of fire through contacting. In the VPL, the user controls the speed of the belt which carries the wafers into the AR coating equipment, the amount of gas supplied and the temperature in the chamber (Table 6-9).
Table 6-9: TiO2 Anti-Reflection Coating Variables Variable User Input Units Min Max Default ARBS arBeltSpeed cm/sec 1 10 5 ART arTemperature °C 0 300 150 ARGF arGasFlows l/sec 1 5 2
The thickness and refractive index of the resultant AR coating is determined by the gas flow and the belt speed, while the hardness of the coating is determined by the temperature. The
200 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
hardness of the coating impacts the penetration of the fire-through metal contacts when they are applied after the AR coating, as described in the following section.
Figure 6-4: (a) Non-uniform AR Coating, and (b)Atmospheric Pressure Chemical Vapour Deposition of a TiO2 AR Coating
(a) (b)
Over the past decade, SiNx AR coatings have gradually replaced TiO2 coatings. SiNx AR coatings have the additional benefit of providing surface passivation. The plasma-enhanced chemical vapour deposition (PECVD) method is of particular interest to developing countries, as it allows the fabrication of SiNx coatings at low temperatures (around 400 °C) so clean rooms are not necessarily required (Kuepper, 2006). However, PECVD requires continuous access to highly pure gases and the equipment is complex to run and understand. Silicon nitride AR coating is controlled by the VPL operator in a similar way to the TiO2 AR coating, except that there are two types of gases used in the process (Table 6-10).
Table 6-10: SiNx Anti-Reflection Coating Variables Variable User Input Units Min Max Default ARTM arTime mins 1 30 15 ART arTemperature °C 0 400 150 ARGF1 gasFlow1 l/sec 1 5 2 ARGF2 gasFlow2 l/sec 1 5 2
The VPL implements the surface passivation effect of SiNx AR coatings, which in the case of multicrystalline wafers also passivates dangling bonds at the crystal interfaces and therefore improves the minority carrier lifetime according to the equations contained in Appendix 4.
201 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
6.2.8. Screen Printing and Firing of the Metal Contacts Solar cells usually have metal contacts attached at the front and back surfaces to collect the current and carry it to an external circuit (Figure 6-5). In the commercial manufacturing environment, most companies screen-print metal pastes to form metallic contacts to silicon wafers. Screen printing of the metal contacts of the solar cell was developed for the solar cell industry, replacing vacuum deposition used in the electronics industry (Menanteau, 2000).
Figure 6-5: (a) Front metallization of a screen-printed multicrystalline solar cell with AR coating, (b) Screen printing of Al back contacts in a commercial production line
(a) (b) The VPL user is able to specify the design of the screen used for the screen printing process, including the density of the metal mesh, the thickness of the wire in the metal mesh, the distance between the screen and the surface of the wafer and the distance between the screen and the squeegee (Table 6-11). The thickness of the strands in the wire mesh used for screen printing is important as thick strands can restrict the amount of paste that is squeezed through, while thin strands can break easily.
Table 6-11: Al Screen Setup Variables Variable User Input Units Min Max Default ALMD AlMeshDensity strands/cm 50 250 100 AETA AlEmulsionThicknessAbove microns 0 30 15 AETB AlEmulsionThicknessBelow microns 0 30 15 ALSD AlStrandDiameter microns 20 120 50 AlPattern Solid
The Al screen setup variables interact with the Al screen print variables, which include the pressure applied by the squeegee, the viscosity of the metal paste and the speed that the squeegee moves across the wafer surface (Table 6-12). Together they determine the thickness of the paste deposited, according to the algorithms in Appendix 4.
202 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Table 6-12: Al Screen Print Variables Variable User Input Units Min Max Default ALPP AlScreenPrintPressure (low-high) 1 10 5 ALV AlScreenPrintViscosity (low-high) 1 10 5 ALPS AlScreenPrintSpeed (low-high) 1 10 5
If insufficient Al paste is deposited in the printing process, all of the paste will be converted to an alloy during firing, leaving no Al metal, which reduces conductivity and is not suitable for soldering interconnect tape to. Too much Al paste will take a long time to dry and may not be entirely converted to Al metal (by driving out the solvents and binders) during the firing process. In this case, the rear contact will not have good conductivity or solderability.
Table 6-13: Al Firing Variables Variable User Input Units Min Max Default AFBS AlFiringBeltSpeed cm/sec 1 10 5 AFDT AlFiringDryingTemp °C 100 300 250 AlFiringZone1Temp °C 500 800 600 AFZ2 AlFiringZone2Temp °C 700 1000 800 AlFiringZone3Temp °C 500 800 800 AFGF AlFiringGasFlow (Nitrogen) l/sec 0 20 10 AFOP AlFiringOxygen % 0 100 21 cellsBrokenInAl cellsBrokenInAl
During the Al firing process, the Al paste is converted to molten metal and the solvents and binders are evaporated. Additionally, at high temperatures, the Al forms an alloy with the silicon in the solar cell, which reduced recombination at the back surface of completed solar cells. While higher temperatures provide better solid solubilities and therefore better alloying, too much alloying consumes all of the aluminium. High temperatures combined with fast belt speeds are therefore generally preferable. If wafers have been broken during the screen printing or transportation of cells prior to drying, Al paste may be unintentionally deposited on the front of the solar cells following. During firing, this is alloyed through the junction, leading to shunting of the solar cell. The likelihood of solar cell breakage is modelled according to the squeegee pressure and the previous processing of the solar cell, such as the etching process. Screen printing of the front metal contacts is similar to the printing of the Al back contacts. Because it is important to reduce the shading losses, while obtaining good coverage of the solar cell in order to collect the generated current, silver pastes, which have good conductivity, are commonly printed onto wafers in a grid pattern. The metal grid on the top of a solar cell must be designed to minimise reflection. Thin lines of metal, however, will have a higher resistance to the flow of current. There are limits to how accurately metal can be applied in very thin, closely spaced lines during manufacture, and the emulsion thickness issues are even more critical for front surface contacts than for rear contacts. 203 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Table 6-14: Silver Screen Setup Variables Variable User Input Units Min Max Default (SMD) SilverMeshDensity strands/cm 50 250 100 SETA SilverEmulsionThicknessAbove microns 0 30 20 SETB SilverEmulsionThicknessBelow microns 0 30 20 SFS SilverFingerSpacing mm 1 6 2 (SSD) SilverStrandDiameter microns 20 120 80 (MW) SilverFingerWidth microns 50 250 200
Table 6-15: Silver Screen Print Variables Variable User Input Units Min Max Default (SSP) SilverScreenPrintPressure (low-high) 1 10 5 (SPV) SilverScreenPrintViscosity (low-high) 1 10 5 SilverScreenPrintSpeed (low-high) 1 10 5
The formation of the silver contacts with the top diffused region of the solar cell is one of the most critical steps. Adequate alloying must be achieved while avoiding driving the silver too deeply into the solar cell, where the emitter region has less phosphorous doping and therefore the silver-silicon boundary has less conductivity and also to allow electrons to flow in the thin p-type region and enter the silver contact. Additionally, it is critical not to penetrate the junction, short circuiting the solar cell. The highest performance screen printed solar cells have shallow, heavily diffused emitters with silver contacts penetrating a small amount into the emitter. The junction depth and doping density obtained during the diffusion of the emitter, and the thickness and hardness of the antireflection coating in the case of fire-through metal contacts will be factors in determining the temperature and duration required in firing of the top metal contacts
Table 6-16: Silver Screen Firing Variables Variable User Input Units Min Max Default (SFt) SilverFiringBeltSpeed cm/sec 1 10 5 SilverFiringDryingTemp °C 100 300 250 SilverFiringZone1Temp °C 500 800 650 SFT2 SilverFiringZone2Temp °C 600 1000 850 SilverFiringZone3Temp °C 500 800 650
High temperatures in the firing of metal contacts limit the application of screen-printing to clean environments, as impurities could be driven into the silicon.
204 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
6.3. Simulation of Quality Control Tests
After completing each process in the sequence, relevant tests become available in the tools menu. Some of the tests return parameters that are stored in the wafer data within the VPL. Others involve running a PC1D simulation, in which case, a graph will be available for display.
6.3.1. Simple Tests which Return Stored Parameters The following is a list of the tests which simply return the parameters which are stored as part of the wafer data file and updated when different processes are performed: Wafer thickness Resistivity Minority Carrier Lifetime Pyramid Size and Coverage AR Coating Thickness AR Coating Refractive Index Silver Paste Thickness Silver Contact Resistance Series Resistance Shunt Resistance
Three of these tests are now described in more detail in order to provide an indication of the value of these tests. The remainder of the tests are more fully described in the help files accessible from the VPL software.
6.3.1.1. Wafer Thickness Thickness measurements are used following chemical etching to ascertain whether the right amount of silicon has been removed so as get rid of all the saw damage on the surface resulting from the wafering process. Excess silicon removal will lead to wafers being thinner than desirable while insufficient silicon removal will lead to residual saw damage of the surface and likely subsequent deterioration in the minority carrier lifetimes.
6.3.1.2. Minority Carrier Lifetime Minority carrier lifetime is the time necessary for the generated minority carrier concentration to decay to 1/e of its original value. It is a measure of the average length of time that an electron-hole pair will survive prior to recombining. The parameter is initially determined by wafer type, quality and doping concentration within the wafer. For example, a
205 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
more heavily doped wafer will have shorter minority carrier lifetimes because a higher number of majority carriers are available for participation in the recombination process. The open circuit voltage and short circuit current are affected by the minority carrier lifetime, since it determines the characteristic dark saturation current of the diode and because it determines the distance the carriers can travel, it affects the current that can be collected. On a real production line, Voc decay measurements can be used for minority carrier lifetime testing, but only after a junction has been formed. Photocondutivity decay measurements, using an electric coil adjacent to the wafer which is excited by a light source can be used without a junction, but the surface must be passivated in order to obtain good measurements. The minority carrier lifetime test in the VPL returns the value stored for each wafer. The test can be used after high temperature steps in order to determine if contamination has damaged the wafer, or after processes which involve gettering, such as the firing of the aluminium rear contacts, to determine how much they have improved the minority carrier lifetime.
6.3.1.3. Pyramid Size and Coverage Visual inspection after texturing can be performed with the use of an optical microscope to determine the typical pyramid size, shape and the percentage of the wafer surface effectively covered by pyramids. A graticule on the eyepiece of the microscope facilitates reasonable accuracy (to within about 10%) of the base dimensions of the pyramid. The pyramid size and coverage tests in the VPL return the values that are created during the texturing process.
6.3.2. Tests Which Call PC1D PC1D (Basore, 1990), a numerical modelling package, which is able to simulate the performance of the solar cell, is used to perform analysis on the virtual solar cells for the following tests:
6.3.2.1. Doping Profile The n and p type doping concentrations can be determined by measuring the capacitance as a function of applied voltage, or by using mass spectrometry. The depth of the junction and the doping concentration in the diffused layer can be determined, as illustrated in Figure 6-6.
206 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Figure 6-6: Doping Profile Displayed by the VPL
6.3.2.2. Sheet Resistivity In the VPL, the sheet resistivity is calculated by PC1D on the basis of the known doping concentrations and lifetimes of the materials. A four point probe tester can be used to measure sheet resistivity. Two probes are used to feed current into a layer of material. The other two measure the voltage, allowing the resistivity in the layer to be determined. The test can be used to determine the sheet resistivity of the undiffused wafer, but is particularly useful following the formation of the emitter, when the test measures the sheet resistivity of the emitter. The sheet resistivity of the emitter indicates the losses due to resistance to the lateral flow of current in the emitter of the finished device, as it travels to the silver contact. It is also a direct measure of the amount of phosphorus that has been diffused within the emitter, but only in a single layer, not a diffusion profile.
6.3.2.3. Voc Contour Mapping The open circuit voltage test probes the open circuit voltage at a variety of points across the cell surface area. A contour map of Voc can be generated, identifying localised shunting which may occur as a result of contamination from aluminium paste in the screen printing process. Because the VPL does not contain spatially resolved information about the wafer, the Voc test uses PC1D to calculate the Voc for the wafer at various points. Different values for shunt resistance are given to PC1D for the purposes of these calculations. The first value is determined on the basis of the wafer characteristics. The second point is for the case at the edge of the wafer, and used a shunt resistance calculated on the basis on the surface resistivity, which will be approximately ten times the shunt resistance prior to edge isolation. If the wafer has been shunted via contamination during processing, a third value is calculated for a shunt
207 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
resistivity randomly generated between 0.5 and 0.2. Depending on the difference between the best and worst values calculated for Voc, up to eight Voc values between the two are generated, which are then used to plot contours, as shown in Figure 6-7.
Figure 6-7: Voc Test Results
6.3.2.4. Pseudo IV Curve The pseudo IV curve can be measured prior to the application metal contacts by using photoconductance in a similar way to that described for the minority carrier lifetime test. The curve does not include the series resistance losses in the metal contacts or the metal silicon interfaces, but can identify losses due to junction recombination or shunting effects.
6.3.2.5. Reflection and Spectral Response The current produced by the solar cell in response to specific wavelengths of light is measured. The quantum efficiency is the percentage of photons of incoming light that create electrons that contribute to collected current. The internal quantum efficiency is the percentage in relation to photons that enter the cell, whereas the external quantum efficiency is calculated in relation to the total number of photons, including those that are reflected from the surface of the wafer. Reflection can be measured from the top surface of the wafer before and after texturing in order to determine the effectiveness of the texturing process.
208 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Figure 6-8: Reflection and Spectral Response Test Results
6.3.2.6. IV Curve The IV (current-voltage) curve displays the electrical properties of the cell, as described in Appendix 4. The curve provides information about the series and shunt resistances of the device and the performance of the junction. The maximum power and the overall efficiency of the device can be calculated. The test is conducted by using a flash simulator, which uses standard test conditions: 1000W/cm2 of light of spectrum according to AM1.5 at a temperature of 25 degrees C. The test allows cells to be sorted into similar lots for encapsulation into modules, which is important because mismatched cells cause energy generated in the best cells to be wasted as the current flowing through a string of cells is limited by the worst performing cell.
Figure 6-9: IV Curve Test Results
209 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
6.4. Educational Features of the VPL
6.4.1. Saved Settings When a batch of wafers is processed through the VPL, all of the processing parameters used and the results of all the tests performed are saved as part of the batch data. This enables the user to look back on previously processed batches and compare the results of varying parameters. The user can print the batch info, including test data to a text file by pressing the ‘export batch’ button on the toolbar. Text files can then be opened in a spreadsheet utility, where cells within a batch or different batches can be compared by, for example, graphing the data produced on the same axes. If the user does not specify differently, the settings for each process remain the same as those used the previous time the process was run (these are known as ‘previous settings’). This means that the user can easily vary only a small number of parameters and observe the effects. There is a set of reasonable, but not optimised parameters that are stored as ‘default settings’ within the VPL. The user can also choose to have these settings used at any time, which provides a useful starting point for any new students. In addition, the user can specify ‘preferred settings’, so that all the processing parameters in the VPL can be returned to these values at any time. Preferred settings can be edited through the ‘edit settings’ menu item in the VPL user interface. Default, previous or preferred settings can be used either for a single process, or can be selected as the settings for all processes, which can then be altered for any particular process.
6.4.2. Help Files Help files exist for each of the processes and each of the tests available in the VPL. The help files explain the principles behind the interactions of the processing parameters and the resulting performance of the solar cell. Practical difficulties in relation to the processing and testing techniques and equipment are also described. Because there are no textbooks available on commercial solar cell production, these help files contain a unique and invaluable training resource.
6.4.3. Assignments The VPL contains nine in-built assignments which test the user on the optimisation of a single process or a combination of processes by keeping all the parameters in the production line as default values apart from some of those that are relevant to the learning objectives of the current assignment. The currently available assignments are: Wafer Etch Texturing
210 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Acid Clean Phosphorous Diffusion Screen Printing and Firing of the Aluminium Contacts Plasma Etching Screen Printing and Firing of the Silver Contacts Diffusion and Silver Screen Printing and Firing
Because of the complex interactions of solar cell processing, learning is likely to be more effective if there is initially a focus on only a few of the variables at one time. As an understanding of the variables develops, more complexity can be added. The assignments assess the user on achieving the best average performance for the cells in the batch as well as the best single cell, in order to teach users about the importance of production line yield. In some cases, the users attention is focused on achieving the best results for a specific test, such as optimal pyramid coverage in the case of the texturing assignment.
211 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
6.5. The Effectiveness of the Virtual Production Line
6.5.1. Software Simulations as Learning Tools Simulations have been found to be effective in promoting learning. Experiments compared the effectiveness of simulation software with traditional pen and paper and lego model building in relation to learning how to solve a mechanical problem (Reamon & Sheppard, 1997). It was found that software simulations assisted learners in building a more complete mental model, which they were better able to relate to abstract symbolic variables and expressions that the learners who relied on the other two modes. While simulations helped learners develop a good intuition or feel for the mechanisms being studied (‘a well developed mental model’), they were best complemented by abstraction through the study of equations, since this abstraction better allows the transfer of the learning to other situations. When combined with hands-on experience, theory and abstraction, it was found that software simulations can be highly effective learning tools. (Nahar et al., 2001) used a case study approach to evaluate the use of and advantages of IT tools in international technology transfer. They found that software based training tools, including simulation tools reduced the travelling needs of experts from the technology supplier, reduced the need for on the job training, made training possible at the convenience of the recipient firm, and increased the training and technology transfer capacity of the supplier, because they allowed the user to practice and repeat limitlessly and allowed users to see almost the same plant parameters and control the same equipment as in the real plant. The ability to monitor the training results was also improved.
6.5.2. Measuring the Effectiveness of the Virtual Production Line at UNSW The VPL has been used as the basis of a fourth year engineering subject “Photovoltaic Technology and Manufacturing”, within the Photovoltaic Engineering degree at the University of New South Wales in Australia. The course has been run three times, in 2002, 2004 and 2006 with enrolment of approximately 60 students on each occasion.
212 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
Figure 6-10: A class of 55 students studying solar cell manufacturing through the VPL
Feedback from student surveys indicate that the VPL is an effective tool for teaching about the manufacturing environment and the importance of the various processing parameters on device performance. Although the lecturer in charge was not able to be present for a significant number of weeks during the semester, the UNSW Course And Teaching Evaluation and Improvement process (CATEI) in 2006 revealed that the overall satisfaction in the course was rated 3 in a range 1-4, where 4 is strongly satisfied and 1 is strongly dissatisfied (83% of the students either agreed or strongly agreed with the statement “Overall, I was satisfied with the quality of this course”). Table 6-17 contains the ratings for a number of other relevant questions in the survey.
Table 6-17: Course and Teaching Evaluation Scores for a Course that used the VPL Question Mean Rating (1-4) The course was challenging and interesting 3.2 The course provided effective opportunities for active student participation in learning 3.1 activities The course was effective for developing my thinking skills (e.g. critical analysis, 3.2 problem solving) The assessment methods and tasks in this course were appropriate given the course 2.9 goal The course advanced my ability for independent learning and critical analysis 3.3
By comparison, the average CATEI rating for an engineering course at UNSW is 3.0, so this course is perceived to be of above average effectiveness as a mode of education delivery as assessed under the criteria.
213 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
6.5.3. The Effectiveness of the VPL at Suntech Power The VPL has also been used in the training of engineers at Suntech Solar in China, as described in more detail in the following chapter. As a result of processing and monitoring production through the VPL, experience has been gained in process control and a better understanding of the complex interactions of processing parameters has been developed. Training through the Virtual Production Line at Suntech has ensured that some of the manufacturing knowledge embodied in experienced individuals such as the CTO, is diffused throughout the organisation. The senior production line engineers at Suntech are now in a position to train the more junior staff. The training can be carried out at low cost, without disrupting the operation of the production line. The use of the VPL has also reduced the time and travelling requirements of the CTO, enabling him to spend more time in the training of research personnel. In Figure 6-11, the framework is used to highlight the role of production line software simulations in assisting learning for PV manufacturers.
Figure 6-11
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
incentives & incentives, resources PRODUCTION direction INNOVATIVE resources CAPABILITIES for search CAPABILITIES and direction for search ROUTINES Learning by Doing Learning by Searching new production technique DOING R&D Improvement Capabilities
RECONFIGURATION
Simulating the production line environment allows individuals to learn by doing, so this process is highlighted in red in the diagram. Through the interactions of senior and junior engineers during the training routines, knowledge is shared throughout the organisation and the total production capabilities are increased, so routines are also shaded red. This type of simulation can also be used to identify problems in the real production line by comparing its performance with the simulation. Where the enterprise wishes to make changes in the fine tuning of processes on the real production line, the simulation can also be used to predict the
214 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
effect of these changes and therefore inform production in a similar way to quality control routines. The virtual production may have a limited role to play in providing information that can inform innovation or investment. Coordination and integration is therefore shaded orange. However, the VPL is not useful in simulating search-based learning because it is based on the conventional technology. Other types of software simulations such as PC1D, which has been described earlier (Basore, 1990) can be used to model new types of devices and may better supplement the understanding gained through R&D.
215 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
6.6. Conclusion
The development and use of a simulation software package designed for the education and training of solar cell production line engineers has been described. The in-line testing available in the package enables the user to monitor the impact of varying processing parameters and to learn about the use of these tests in commercial production. The software includes features which enable users to learn about manufacturing concepts such as quality control testing, yield optimisation, and parameter tolerances. The VPL has found to be an effective tool for the education of student engineers in a university environment and for training production line engineers. Its value lies mainly in acting as an alternative for doing-based learning, where it can work to improve production capabilities. Other types of software simulations may also be useful in promoting search-based learning and building R&D capabilities. There is potential for the VPL and software simulations more generally to assist manufacturers in developing countries in overcoming problems related to the shortage of skilled personnel in the industry by enabling experiential learning without the need to access the production line equipment and by reducing the cost of and the time required for training. It may also enable them to take advantage of the relatively low labour costs in developing countries by reducing the need to hire staff internationally.
216 Chapter 6. A Virtual Production Line for the Manufacture of Screen-Printed Solar Cells
References
Basore, P.A. (1990), Numerical Modelling of Textured Silicon Solar Cells Using PC1D, IEEE Transactions on Electron Devices, ED-37 (2), p 337. EPIA and Greenpeace (2007), Solar Generation IV 2007: Solar electricity for over one billion people and two million jobs by 2020, European Photovoltaics Industry Association, Greenpeace. Haase, J. (2005), High Volume Solar Cell Production Lines - Concepts and Cost Aspects, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. Koch, W., Lucas, P. and Viaud, M. (2005), Cooperation Opportunities Between the Semiconductor and the Photovoltaic Industries, 20th European Photovoltaic Solar Energy Conference, Barcelona, Spain, 6-10 June. Kuepper, N. (2006), The Production of Silicon Solar Cells in Developing Countries, Bachelor of Engineering Honours Thesis, School of Photovoltaics and Renewable Energy Engineering, The University of New South Wales, Sydney, Australia. Menanteau, P. (2000), Learning from Variety and Competition Between Technological Options for Generating Photovoltaic Electricity, Technological Forecasting and Social Change, 63 (1), p 63. Nahar, N., Savolainen, V. and Huda, N. (2001), IT Aided Training in International Technology Transfer, in Kocaoglu, D.F., Anderson, T.R., Milosevic, D.Z., Daim, T.U., Niwa, K., Gulledge, T.R., Kim, C. & Tschirky, H. (eds), "Technology Management in the Knowledge Era (CD-ROM)", PICMET, Oregon, U.S.A. Reamon, D.T. and Sheppard, S.D. (1997), The Role of Simulation Software in an Ideal Learning Environment, DETC'97, Sacramento, California, September 14-17, 1997. Wenham, S., Green, M., Watt, M. and Corkish, R. (2006), Applied Photovoltaics - Second Edition, UNSW Centre for Photovoltaic Engineering, Sydney, Australia. Wenham, S.R., Bruce, A. and Cotter, J. (2002), Manufacturing of Screen Printed Solar Cells through the Virtual Environment, 8th International Symposium on Renewable Energy Education (ISREE8), Orlando, Florida, August 2002.
217 218 CChhaapptteerr 77.. CCaappaabbiilliittyy BBuuiillddiinngg aatt SSuunntteecchh PPoowweerr
The purpose of this chapter is to apply the analytical framework developed in chapter 5 to the case of Suntech Power (Suntech), a photovoltaic cell manufacturer in China, in order to identify how capability building occurred in this case, and the ways in which the Chinese technological system supported or constrained the enterprise. Suntech was chosen as a revelatory example, since high levels of capabilities, including innovative capabilities for cell manufacture, have been rapidly and successfully built. First producing photovoltaic cells in 2002, Suntech was profitable in its first full year of production and grew rapidly to become the 10th largest producer of cells worldwide by 2004, the 8th largest in 2005 and the 4th largest in 2006. The enterprise is implementing innovative high-efficiency, low-cost technologies and has continually improved production yields and reduced the use of silicon material in the cells. This was achieved when the technological system for photovoltaics manufacture in China was in its infancy. The case study research comprised a number of structured and semi-structured interviews with the chief technical officer of Suntech, Professor Stuart Wenham and the use of secondary sources of information that have documented technological milestones and progress of Suntech. The historical development of the PV industry in China, and the origin and start up of Suntech Power is described in section 7.1. Section 7.2 details the capabilities developed at Suntech and section 7.3 describes the capability building strategies adopted by Suntech. In section 7.4, the framework developed in chapter 5 is used to identify the roles of different types of learning at Suntech and the organisational and technological system factors that have influenced the effectiveness of each type of learning. The discussion in chapter 10 of this thesis draws upon the insights from this case study and the existing literature on learning in the PV industry to propose typical requirements and constraints of modern sector PV manufacturing enterprises in developing countries, appropriate capability building strategies and policy interventions to assist these manufacturers.
219 Chapter 7. Capability Building at Suntech Power
7.1. Background to the Case Study
7.1.1. PV Manufacturing in China China began the 21st century with a collection of underperforming PV manufacturers who used outdated equipment and produced cells with low efficiencies. Since 2000, a number of new Chinese PV manufacturers have been established. China is now the third largest producer of solar cells, exporting most of the production. The Chinese PV industry in the period up to 2000 and post 2000 are now described.
7.1.1.1. The period up to 2000 Research on photovoltaics in China began in 1958, and small pilot plants were launched in the 1970s (Yang et al., 2003). In the late 1970s, three state-owned semiconductor plants in Ningbo (Ningbo Solar Cell Factory), Yunnan (Yunnan Semiconductor) and Kaifeng (Kaifeng Solar) were converted into monocrystalline cell manufacturing plants (Dai et al., 1999). In the late 1980s, the existing manufacturers upgraded their production lines and four more enterprises were established (Dai et al., 1999; Marigo, 2006): Non-Ferrous Academy, a state owned enterprise in Beijing, which was affiliated with Beijing General Institute for Non-Ferrous Metals and folded prior to 1999; Qinhuangdao Huamei in Qinhuangdao, which folded in 2003; Yu Kang Solar in Yunnan, which was a joint venture between Yunnan government, Korean and International Finance Company, which folded in 1997 because of marketing failures; and Harbin-Chronar in Harbin, a joint venture with a US enterprise, which folded in 2003. All of the enterprises were state-owned, apart from the joint ventures Harbin-Chronar and Yu Kang Solar, which were state-foreign joint ventures. All production lines were financed through government R&D programmes or international aid programmes (Li, 2004b). Either the whole production line or all of the major equipment of these early Chinese manufacturers was sourced from mainly US suppliers, including equipment for monocrystalline silicon ingot pulling, ingot squaring and wafer slicing, cell fabrication, and module lamination (Shi, 2005; Yang et al., 2003). China did not produce capital equipment for photovoltaics manufacturing until 2000 (Yang et al., 2003).
220 Chapter 7. Capability Building at Suntech Power
Table 7-1: Chinese Cell Manufacturers, Equipment Source, Technology and Production, 1995 Enterprise Starting Date Equipment Technology Production Production Capacity 1995 1995 Harbin- 1991 All imported a-Si 1 MW 200 kW Chronar Non-Ferrous 1987 All imported mc-Si 100 kW 20 kW Academy Qinhuangdao 1990 Key imported sc-Si 1 MW 200 kW Huamai PV Kaifeng Solar 1975 Key imported sc-Si 300 kW 180 kW New line 1988 Ningbo Solar 1976 Key imported sc-Si 300 kW 300 kW Cell Factory New line 1988 Yunnan 1983 All imported sc-Si 500 kW 300 kW Semiconductor New line 1987 Total: 3.2 MW 1.2 MW Source: (Dai et al., 1999)
According to Table 7-1, the production capacity of the six cell manufacturers in 1995 was about 3.2 MWp per year, but the actual production was 1.2 MWp (37.5% of capacity), due to ‘serious equipment bottlenecks in different parts of these production lines’ resulting from lack of finance to invest in the necessary imported equipment (Dai et al., 1999, p 4). The technology used in the older plants in the 1990s was basically the same as that used in the late 1980s (Dai et al., 1999; Yang et al., 2003). Since this first generation of Chinese manufacturers were state owned and funded, they were dependent on the government’s decisions and had neither the mandate nor the incentive to carry out R&D or to innovate (Liu & White, 2001). The entrance of foreign enterprises in the mid 1990s increased the pressure on domestic producers. Dai et al. (1999) observed that ‘some domestic manufacturers still work well, some run under difficulty and some are even going to close or [are] already closed.’ The efficiency of Chinese manufactured solar cells averaged 10-12% in 1999, while the maximum achieved was 13.5%. In contrast, foreign manufacturers were producing cells that were on average 14-15% efficient (Dai et al., 1999). Given fixed manufacturing costs per cell, the approximately 20% lower efficiency achieved by Chinese manufacturers translates into 20% lower production output for the same input. Chinese manufacturers were therefore struggling to make a profit and were consequently unable to invest in capital equipment. Zhao (2001) reported that production capacity in 1998 was 4 MWp and production was 2.3 MWp (an improvement to a still unsatisfactory 57.5% of capacity). Dai et al. (1999) reported that production capacity was 5 MWp in 1999. Zhao et al. (2006a, p 27) describe the 1990s in the Chinese PV industry as a period of “importation, digestion, absorption and innovation”, resulting in capacity increases and capability improvements.
221 Chapter 7. Capability Building at Suntech Power
7.1.1.2. The post-2000 period There was only one more new entrant (Trony) prior to the early 2000s, when a number of new cell and module producers started up, including Shagahi Topsol (2000), Suntech Power (2001), Baoding Yingli (2001), Shenzen Topray (2002), Soltech (2003), Tainjin Jinneng (2003) and Nanjing PV Tech (2004). In 2000, Topsol (Shangahai Topsol Green Energy Co. Ltd.) collaborated with the Institute of Solar Energy of Shanghai Jiaotong University (SJTU) to install a production line for c-Si solar cells, with some of the equipment, including a furnace and laminator being designed in-house (Yang et al., 2003). Topsol has accessed experts from SJTU, national research institutes and well-known enterprises. Baoding Yingli (Baoding Tianwei Yingli), formed in 2001, and first producing cells in 2003, was the first solar cell and module manufacturer in China to produce its own multicrystalline ingots (Schmela, 2005a). Baoding Yingli purchased wire saws from Switzerland and module production equipment from Italian manufacturer Helios. Almost all the materials, including interconnect ribbon, EVA and even silicon carbide used to cut wafers was imported from Europe or the US. The enterprise was slow to get the production line working. It took until 2006 before production reached 37 MWp. In 2004, Nanjing PV Tech (CEEG), a joint venture between the Chinese Electrical Equipment Group in Jiangsu and a group of Australian experts was established. CEEG have a strategic partnership with the University of New South Wales (UNSW) to jointly research and develop technology. Doctor Zhao Jianhua, who was an Associate Professor at UNSW, is the General Manager. By the end of 2005, CEEG’s capacity was around 300 MWp and production around 150MWp. New start-ups Jing Ao, Jiang Ying Jetion, Shanghai Chaori, SMIC, Solarfun and Big Sun have entered production in 2006, and a number of other enterprises are preparing production facilities. In Jing Ao, JingLong Industry has 55% shares, Australia Solar Development Company has 30% shares, and Australia PV Science & Engineering Company has 15% shares. JingLong Industry manufactures Cz-Si ingots, wafers, and manufacturing equipment. It was established by Australians who had been connected with UNSW, including Dr Dai Ximing, Dr Bruce Beilby, Mr Ted Szpitalak and Mr Yang Huaijin. Whole production lines were imported from Europe and the United States. The enterprises with the highest published cell efficiencies include new entrants Suntech (17.6%), Jiang Ying Jetion (17.5%) and Jing Ao Solar (17.2%).
222 Chapter 7. Capability Building at Suntech Power
Table 7-2: Current Chinese Cell Manufacturers and Production Data
Date Planned Capacity Production Production Production Production Formed Productio 2006 2006 2005 2004 2003 n 2007 Kaifeng Solar 1975 - - - - - 0.3
Ningbo Solar Cell Factory 1976 60 50 40 25 3 1
Yunnan Semiconductor 1983 25 35 6 4 2 2.5
Harbin-Chronar 1988 - - - - - 0.8
Hua Mei PV Device 1990 - - - - - 0.5
Trony 1993 5.8 0 5.8 4.5 0.8 0
Suntech Power 2001 265 300 160 82 35 8
Shenzhen Topray Solar 2002 70 35 18 20 4 0.9
Baoding Yingli 2003 150 60 37.02 3 3 0
Shanghai Topsol Green 2003 5 0 5 5 3 0 Energy Soltech 2003 10 10 1 0 0 0
Tianjin Jinneng 2003 5 2.5 2 2.2 1 0
Nanjing PV Tech Co. Ltd 2004 130 192 48 5 0 0
Jing Ao Solar 2005 29.5 75 29.5 0 0 0
Jiangsu Linyang 2005 0 0 0 0 0 0
Jiang Ying Jetion 2005 50 25 5 0 0 0
Shanghai Chaori Solar 2005 25 0 1 0 0 0
SMIC Corp. 2005 3.5 5 3.5 0 0 0
Solarfun 2005 26 60 26 0 0 0
Trina 2005 25 0 0 0 0 0
Big Sun 2006 21 0.6 0.6 0 0 0
CSI 2006 30 0 0 0 0 0
Wuxi Shangpin Solar 2006 25 0 0 0 0 0 Energy TOTAL 960.8 850.1 388.42 150.7 51.8 14
Sources: (Dai et al., 1999; Hirshman et al., 2007; Jäger-Waldau, 2006; Schmela, 2005d, 2006) and enterprise websites
223 Chapter 7. Capability Building at Suntech Power
Figure 7-1: Chinese Solar Cell Manufacturers, Production, Capacity and Production Plans
300
250
200
150 Cells (MW) Cells
100
50
0 I y d r gli ar ce on lar h ng or n i i or er CS ev et tec w ne ony Yi g Sun Solar i So larfun o in Trina Tr i D o P J ng B Chron g J h Co. Lt Sol u Linyangn eng S h in n- f SMIC Corp. ec iconduct bi Yi Tec ei PVJing Aongs Solar Baodi M Kai V Tianj Har ang Sunt Sem Jia Ji P Hua ing Shanghai Chaor ingbo Solaranj Cell Fact Shenzhen Topray Sola Yunnan N N Wuxi Shangpin Solar Energy Shanghai Topsol Green Energy
Planned Production 2007 Capacity 2006 Production 2006 Production 2005 Production 2004 Production 2003
224 Chapter 7. Capability Building at Suntech Power
7.1.2. Inputs to the PV Industry in China
Production Equipment Since 2000, manufacturing of silicon feedstock, ingot and wafer production has developed in China, as well as most of the production equipment and materials for cell and module manufacture (Zhao et al., 2006b). Wafer etching, diffusion furnaces, drying furnaces, plasma etching machines and testing and sorting machines are now produced in China and widely used by local manufacturers. The cost of the Chinese equipment is on average 30% of the cost of imported equipment. Some of this equipment performs well and is of good quality. About half of the equipment now sold to Chinese manufacturers is domestically produced (Zhao et al., 2006a). Key equipment is still imported, mainly from Europe (Marigo, 2006). For example, parallel plate PECVD machines and automatic screen printers cannot be made domestically, and the cell handling for automated production lines cannot be produced locally. The materials such as aluminium paste, wet chemicals and slurry however, can also be produced domestically. Locally manufactured equipment for all but the most complex production tasks can now be purchased for 30% of the cost of imported equipment (Zhao et al., 2006a).
Silicon supply Silicon shortages are extremely problematic in China. The growing number of enterprises compete for the limited supply and the prices have soared. Lack of feedstock has constrained the production of solar cells in China for the past few years, and it is reported that some feedstock of low quality has been used due to lack of monitoring (Zhao et al., 2006a). Li (2006) says that the figures given by Chinese manufacturers are exaggerated, because they do not want to reveal low capacity factor and that the silicon shortages are causing most to produce at 30% capacity. Much of the silicon production in China is monocrystalline silicon, since the plant is cheaper and the techniques are mature, so the quality is equivalent to imported ingots and the equipment can be manufactured locally (Zhao et al., 2006a). Domestically produced furnaces achieve high quality and are one third to half the price of imported ones (approximately US$150,000 each) (Pichel, 2006). The plants are smaller and investment and construction times are also less. This production relies on imported feedstock. China had capacity for the production of 2386 tons of monocrystalline ingots in 2005, equivalent to 200 MWp of solar cells. The polycrystalline silicon manufacturers are using an adapted Siemens process, which is more energy intensive than the most advanced techniques and the scale of operations in China is also too small to achieve low costs (Zhao et al., 2006a). All the polycrystalline ingot furnaces are imported. In 2005, Chinese manufacturers supplied 80 tons of polycrystalline silicon, while the demand from the PV industry was 1596 tons.
225 Chapter 7. Capability Building at Suntech Power
Despite large investments in new monocrystalline and polycrystalline silicon capacity, China will not be able to provide enough silicon to supply all its cell manufacturing capacity for a number of years. Small Chinese cell manufacturers are likely to continue to have problems sourcing material at reasonable cost, and even large ones will have difficulty unless they have supply contracts. Some cell manufacturing enterprises have invested in Chinese silicon production either directly or via joint ventures, to avoid the type of long-term supply contracts with controlled prices that silicon manufacturers are now demanding as a prerequisite for expanding their production capacity.
Physical Infrastructure Although large parts of China are not well served by infrastructure, the large cities, and in particular the industrial districts all have reliable electricity supplies and transport networks. Industrial districts receive preferential access to electricity when there is a shortage. Hydro- electric schemes in some parts of China make large amounts of electricity available cheaply for very energy intensive activities such as silicon ingot production. China has a well developed transportation and ports infrastructure to serve its large and growing manufacturing industries.
Human Capital China has a fast growing PV industry and therefore a shortage of people skilled in production and innovation (Zhao et al., 2006a). China does, however, have many people with related competencies from the electronic industries. Formerly, the government assigned students to a work unit, or danwei, and a job for life after graduating. China's economic reforms have rendered the old system obsolete, and students are now able to search for jobs on their own. Since the late 1970s, China had a policy of sending students to study in overseas universities (Broaded, 1993; Li, 2004a). With booming economic growth, Chinese enterprises have been able to attract back some of these nationals who have been trained abroad in foreign enterprises, universities, and R&D institutes. Chinese nationals have the advantage of local connections, understanding the way business is done, as well as being exposed to the international technology frontier, western ways of doing business, and speaking English. As a result of the 130 million rural migrant workers that take low paid jobs in the cities, and the low value of the Chinese currency, labour costs in China remain amongst the lowest in the world (Li, 2004b). Unskilled wages can be less than US$200/month, or about US$1/hour, one tenth of the wages paid in industrialised countries (Li, 2006; Pichel, 2006).
226 Chapter 7. Capability Building at Suntech Power
7.1.3. Recent Industry Developments
Manufacturing Cost Advantages In addition to the savings available to Chinese PV manufacturers via the purchase of domestically produced equipment and materials and the low cost of labour, there are a number of other cost advantages to manufacturing in China. For example, sales and administration and R&D costs are lower than in industrialised countries, as are tax rates (Pichel, 2006). Construction costs are also 10-20% of the typical cost in Europe (Li, 2006). The average cost of industrial land is also very cheap: US$0.2-0.4 million / hectare, while electricity is the equivalent of $US 6c/kWh. As a result, some smaller manufacturers can afford to offer modules for 5-10% less than established brands in order to access export markets (Li, 2006), while larger manufacturers can expand, invest in R&D or obtain profits.
Clusters Geographical clusters of PV enterprises have emerged in China (Figure 7-2). Zhejiang and Jiangsu provinces, near Shanghai hosts solar cell manufacturers Suntech, Nanjing PV Tech, Solarfun, Ningbo and a number of others, as well as wafer manufactuers Trina, NREI, Zhejiang Sunda and Jingong. A smaller cluster exists near Beijing and in Hebei province, where there are six module equipment manufactures and cell manufactures Tianwei Yingli and Jing Ao and Ingot manufacturers Jinglong.
Figure 7-2: Geographical Clusters in the Chinese PV Industry
Figure has been removed due to copyright restrictions.
227 Chapter 7. Capability Building at Suntech Power
Key to Figure 7-2 Manufacturer Product Types 1. CSI Technologies Modules 2. EMEI Semiconductor Materials Silicon 3. Harbin-Chronar 4. Hope Industry and Trade Co. Modules 5. Kyocera (Tianjin) Modules 6. LDK Solar Hi-Tech Wafers 7. Luoyang Silicon 8. Nanjing PV Tech Cells 9. Ningbo Wafers, Cell, Modules 10. Ningjin Songgong Wafers 11. Shanghai Solar Cells, Modules 12. Shanghai Topsola Cells, Modules 13. Shenzhen Jiawei Modules 14. Shenzhen Nenglian Modules 15. Sichuan Xinguang Silicon 16. Soltech Modules 17. Suntech Cells, Modules 18. Tianjin Jinneng Modules 19. Tianwei Yingli Wafers, Cells, Modules 20. Xi’An Modules 21. Xinri Wafers 22. Yunnan Tianda Wafers, Cells, Modules 23. Zhejiang Sino-Italian Wafers 24. Zhong Lian Cells, Modules
Source: (Marigo, 2006)
Links with Research Institutes Most current Chinese manufacturers have links with universities and or research organisations. Collaborations include product and process improvement for immediate commercial applications as well as new device design and equipment development (Marigo, 2006). Chinese PV research is now well funded, and research institutes and universities, through interactions with manufacturers are likely to achieve further progress.
International Interactions Since the entrance of many new PV manufacturers after 2000, including a number of joint ventures and enterprises that have employed foreign experts, the intensity of Chinese enterprises’ interactions with both international and domestic research institutes and enterprises throughout the supply chain have increased. Chinese enterprises are now a visible presence on the international PV scene. For example, 38 of 334 exhibitors at the 2006 European PV conference were from China and Taiwan (Jäger-Waldau, 2006).
Vertical Integration In order to achieve competitiveness by insulating themselves from the high prices of silicon and wafers, many Chinese manufacturers are moving towards vertical integration through purchases, joint ventures and vertical expansion. This has become possible recently, as China has developed the capability for good quality ingot (particularly monocrystalline) manufacture.
228 Chapter 7. Capability Building at Suntech Power
7.1.4. The Chinese Market for PV In 1992, there were less than 10 enterprises retailing, installing and maintaining PV systems in China, all affiliated with government sponsored R&D institutes. In 1998, the number had risen to over 50, including many privately owned enterprises (Dai et al., 1999). At this time, state-owned enterprises occupied the majority of the market.
The Domestic Market China’s PV market in 2003 was comprised of 51% remote rural applications, 36% remote industrial and communications, 9% consumer products and 4% grid connected (REDP, 2004). The Chinese government has implemented large scale rural electrification programmes over the past decade (see Box 7-1), with a doubling of installed capacity from 20MWp to 40MWp in 2002 alone. The target of the Brightness programme is to install 300 MW by 2010 (Schmela, 2005b). The fast market growth and high competition resulted in a price drop from US$4.50/Wp to $3.50/Wp between 1996 and 1998, which prevented most local manufacturers from competing for market share (Li, 2004b). BP Solar, Shell, Siemens Solar, Sharp, Sanyo, SEC and Photomatt all had a presence in the Chinese PV market from the late 1990s (Dai et al., 1999).
Box 7-1: PV Market Programmes in China
Main current PV activities 1996–2010: Brightness Programme: goal is to provide 100 watts of capacity per person to 23 million people with de-centralised energy systems based on solar and wind. 2002–2004: Township electrification programme (part of the Brightness Programme). Rural electrification based on PV, wind and small hydro. Subsidy (208 millions in total) on the capital cost of equipment. Total installed PV capacity at the end of the programme in 7 western provinces is about 20MWp.
Other PV activities 2006: Village electrification programme (a follow up of the Township programme). Electrification of 20,000 villages in China’s off-grid western provinces. 300MWp are expected to be installed. Total budget about 20 Billion RMB (2 Billion $US) 2006. China Renewable Energy Promotion Law comes into effect: Possible PV feed-in tariff. 2006. On-grid roof-top plans in some cities/municipalities (i.e.100,000 roofs in Shanghai and 1,000- rooftop PV programme in Wuxi). Subsidy on the installation equipment. On-grid PV in the Gobi desert (Gansu province): feasibility study under way for 8MWp to be installed PV for the 2008 Beijing Olympic Games. Road lamps, lawn lighting facilities, lamps for public lavatories and irrigation
Installed PV capacity 2004: 60MW Expected installed capacity in 2010: 400MW Cell production capacity 2004: 64MW
Source: (Marigo, 2006)
The small grid-connected PV market, is expected to take at least five years to develop due to many barriers (Shi, 2005). However, local governments have begun to promote this 229 Chapter 7. Capability Building at Suntech Power
application for PV. In October 2005, the Shanghai municipal government endorsed the 100,000 Roof Project for which a feasibility study was commissioned in August 2004 (Suntech Power, 2006f), while Jiangsu province is planning a scheme for 1,000 PV rooftops. The NPC passed a renewable energy law for China on February 28th 2005, which was implemented on January 1st, 2006, and includes a goal to generate 10 percent of energy from renewable energy sources by 2020, and 17% by 2020 (Jäger-Waldau, 2006). There have been some reports (Lewis & Wiser, 2005; Marigo, 2006) that China intends to introduce an aggressive feed in tariff under the renewable energy law, whereas others suggest PV will be considered on a project basis (Photon, 2006). The total installed capacity in China in 2004 was 60MWp, and is projected to be 400MWp by 2010 (Marigo, 2006). However, the market is small compared to cell production and despite commitment to the renewable energy law, the growth trajectory for the Chinese PV market is also much smaller than that for production. For example, projections for PV installations in China in 2010 are 130 MWp and for 2020, 200 MWp (Wang, 2006), while Chinese cell production was already 390MWp in 2006, capacity in 2006 was already 850 MWp, and planned production for 2007 is 960 MWp (see Table 7-2).
Figure 7-3: Three Scenarios for Future Cumulative PV Installations in China
Figure has been removed due to copyright restrictions.
Source: (Marigo, 2006)
Chinese manufacturers will therefore continue to rely mainly on exporting their products to international markets to achieve growth. This leaves them vulnerable to the risk of changes in the level of support for PV in other countries.
Export Markets Until about five years ago, Chinese PV modules suffered a reputation for poor quality internationally. Most Chinese-produced modules did not have quality certificates, or had counterfeit ones (Hug & Schachinger, 2006). Even faked module junction boxes were detected
230 Chapter 7. Capability Building at Suntech Power
in Europe, purporting to be produced by a Swiss manufacturer (Schmela, 2005b). The worldwide shortage of silicon, however, has forced European PV companies to look to China for solar cells. In many cases, German entrepreneurs with a booming home market sought out Chinese products, including via entrepreneur trips arranged for German PV enterprises to China (Hug & Schachinger, 2006; Schmela, 2005b). 97% of Chinese modules were exported in 2005 (Wang, 2006). Most of the exports have gone to Germany, the US, and Japan.
OEMs Chinese OEM enterprises have been making solar cells and modules for sale under the brands of system integrators in Europe. Some of these European enterprises are now purchasing the Chinese OEM enterprise, or engaging in joint ventures. Conversely, some of the OEM manufacturers may forward integrate by purchasing system integration enterprises in Europe and the US (Pichel, 2006).
Downstream Value Chain Interactions Although many linkages have been established between cell and module manufacturers and their suppliers, some of the lucrative value-added downstream parts of the value chain have had less attention. There is no investment or interest in research into built environment PV products, such as BIPV on the part of Chinese R&D programmes (Marigo, 2006). There are, however, a few large scale projects in planning, for example related to the Beijing Olympics in 2008 (Li, 2006). Consumer products manufacturers are also moving into PV street lights and park light products (Shi, 2005).
7.1.5. Support for the PV Industry in China
Fiscal and Financial Support The PV industry is seen as a strategic high technology industry by the Chinese government. The government has therefore provided a range of fiscal and financial incentives to supplement China’s competitive manufacturing advantages and stimulate the growth of the industry. In particular, China has encouraged the export industry and the influx of foreign investment in domestic manufacturing. Materials for exported products are exempted from import duties, which range between 8.5-14%, but the exemptions do not apply when the product is sold domestically (Li, 2004b). Joint ventures and FDI manufacturers are eligible for the import duty exemptions regardless of where the product is sold, an inducement for foreign investment in manufacturing. The usual company tax rate in China is 33%, but some municipal governments also give 5-8 year tax holidays to encourage high-technology projects and foreign investment.
231 Chapter 7. Capability Building at Suntech Power
Government Funded R&D In China, the Ministry of Science and Technology (MOST) is responsible for implementing R&D projects. MOST has been responsible for the National R&D Project, the 863 Project and the 973 Project (Marigo, 2006). The National R&D Project has supported PV since 1981, and the support has included projects such as the upgrading of manufacturing lines at Ningbo and Kaifeng, R&D projects on crystalline, amorphous and polycrystalline silicon solar cells and materials, and the development of equipment for measuring solar cells (Zhao et al., 2006a). The 863 Project (2000-2005) has supported CdTe and CIGS solar cell research, while the 973 Project (2000-2005) has supported research into low cost and long life thin film, dye-sensitized and polymer solar cells. The State Development Planning Commission (SDPC) and the State Economic and Trade Commission are responsible for the industrial development of the industry. The innovation fund for Middle/Small-Scale Enterprises provides investment matching grants to support the development of innovative products and mass commercialisation (Zhao et al., 2006a). The government also supports universities and research institutes in R&D, including the state-run research institute, the Chinese Academy of Sciences. Dai et al. (1999) reported that the PV industry consensus was that too many government agencies were involved in the implementation of PV support programs, resulting in inefficient use of resources, poorly organised and under-achieving R&D projects, and lack of commercialisation of research. There were more than 40 institutes, universities and manufacturers carrying out R&D at that time, primarily focused on improving cell efficiency and developing new device structures. PV R&D in China is now supported at a fairly high level through the R&D departments of some manufacturers (including Suntech, Jiangsu Linyang Solarfun and CEEG), and through national and local governments. The MOST budget for PV R&D in the 11th year plan (2006- 2010) is expected to be around 120 million Yuan (US$17 million) (Marigo, 2006). In comparison, in 2005, Germany spent US$30.3 million, Japan spent US$60.5 million and the US US$86 million on PV R&D (IEA PVPS, 2006). Zhao et al. (2006a) believe that although the problem of lack of commercialisation has been overcome and there are now good industry- research linkages, R&D now suffers from lack of human capital (technical expertise and training), within enterprises and within research organisations. Table 7-3 shows that China has greatly improved their research outcomes with laboratory solar cells between 2002 and 2004, but still lags behind the world’s best research.
232 Chapter 7. Capability Building at Suntech Power
Table 7-3: Highest Efficiency Laboratory Cells Produced in China and Worldwide, 2002 and 2004 China 2002 China 2004 World 2002 World 2004 Technology Highest Area Highest Area Highest Area Highest Area Efficiency (cm2) Efficiency (cm2) Efficiency (cm2) Efficiency (cm2) Silicon Mono-Si 20.4 4.00 20.4 4.00 24.7 4.00 24.7 4.00 Poly-Si 14.5 4.00 16.0 4.00 19.8 1.09 20.3 1.002 Si (thin film) 13.6 1.00 16.6 4.017 16.6 4.017 III-V Cells GaAs 20.1 1.00 21.9 1.00 25.1 3.91 25.1 3.91 (crystalline) Thin Film Chalcogenides CIGS 9 1.00 12.1 1.00 18.4 1.04 18.4 1.04 CdTe 7 0.03 13.36 0.5 16.5 1.032 16.6 16.0 Amorphous Si Si 8.6 100 8.6 100 10.1 1.199 10.1 1.199 (amorphous) Sources: (Green et al., 2003, 2005; REDP, 2004; Yang et al., 2003)
High Technology Development Zones National and provincial governments in China have established high technology development parks to attract foreign investment and to promote export industries and interactions between foreign and domestic enterprises. Suntech’s main production facilities, for example are located in Wuxi New District (WND), a national high-technology development zone, approved by the state council in 1992. Within the 22 square kilometre district, over a thousand foreign-funded enterprises, including Sony, Panasonic, GE and Kodak; five national technological innovation and industrialisation bases, including high-technology incubation services; and a number of non-state-owned high technology enterprises, including Suntech, have been established.
World Bank Support for Quality Improvement In 2004 and 2005, the technical improvement component of the World Bank / GEF China REDP project supported about 10 manufacturers to achieve international standard IEC- 61215 certification (ter Horst & Zhang, 2005). Almost all Chinese module manufacturers now have their products certified under international standards such as IEC (performance), TUV (safety) and sometimes CE (EU conformance) (Marigo, 2006). REDP (2004) attributes this change to the programmes, whereas other authors (Hug & Schachinger, 2006; Schmela, 2005b) suggest the impetus came from the requirements of the German market.
233 Chapter 7. Capability Building at Suntech Power
7.2. Suntech Power
The following sections detail the development of capabilities at Suntech Power.
7.2.1. Origin and Start Up of Suntech Power Suntech, established in 2001, was the initiative of the Chinese government. Based on experience accumulated in the manufacture of photovoltaic space cells, the government had attempted unsuccessfully to establish PV manufacturing for decades. It was recognised that the low efficiency of cells produced and the inability to operate cost effectively in Chinese PV plants was primarily due to the use of old human-embodied technology and equipment. The Chinese government looked worldwide for nationals who had become highly trained in photovoltaics at overseas institutions. In 2000, Wu Xi regional government officials offered Dr Shi Zhengrong, an Australian citizen who was born in Jiangsu Province, the opportunity to set up a new PV manufacturing enterprise in China, using conventional technology, with capital of $US6 million. Dr Shi received his PhD degree from the University of New South Wales (UNSW) on multicrystalline silicon thin film solar cells in 1992. He had since worked as a senior researcher at UNSW and was the deputy research director at Pacific Solar, a enterprise set up to commercialise the thin film crystalline silicon on glass technology initially developed at UNSW. In 2001, there were several PV enterprises in China that were all performing poorly and losing money. Dr Shi therefore had reservations about the potential for establishing a successful PV enterprise in China, and was also reluctant to leave Pacific Power, which was at an interesting stage of its technology development. During a two week visit to China, however, he saw that China’s infrastructure had improved and that, with the right technological expertise and capital, the potential for successful PV manufacture in China existed. Dr Shi approached Professor Stuart Wenham, the Director of the ARC Centre of Excellence for Advanced Silicon Photovoltaics and Photonics at UNSW, for assistance with production technology in the proposed new enterprise, including setting up production lines and optimising the processing parameters. Professor Wenham had previously been involved in setting up and fine-tuning production lines at Tideland Energy (acquired by BP Solar) in Australia and Eurosolare in Italy. Dr Shi obtained agreement from Professor Wenham in 2001, before committing to involvement in the new enterprise. On the 9th September 2001, Suntech Power Co. Ltd was officially established in Wu Xi, in Jiangsu Province, near Shanghai.
234 Chapter 7. Capability Building at Suntech Power
Box 7-2: Suntech Vision:
” Suntech is committed to becoming the "lowest cost per watt" provider of PV solutions to customers worldwide. By focusing on technical leadership through leading R&D and a culture based on innovation, cooperation and integrity, Suntech is working daily to realize its vision to be a global energy leader, providing efficient solar solutions for a green future.
At Suntech we have a vision of becoming one of the world's largest solar energy providers. By producing low cost per watt solar solutions through ongoing investment in R&D combined with our low-cost China- based manufacturing”
(Suntech Power, 2006f)
The government offer to attract Dr Shi included a 25% share of ownership of the enterprise. Once Suntech was established, the government sold their share to enterprises such as Jiangsu-based Little Swan Group, which subsequently sold its major shareholding prior to the public float in 2005. Suntech produced its first modules from purchased cells in March 2002 and began producing cells in September 2002.
7.2.2. Production Capabilities at Suntech Production capabilities have been defined as the ability to produce products of an appropriate quality and cost. Suntech’s ability to produce good-quality products can be measured in part by the efficiency of the cells it produces, and can also be observed in its ability to achieve quality certification of its products. Its ability to produce cells at low cost is a function of cell efficiency and high yields on the production line, and the use of less silicon through thinner wafers. The ability to offer products at an appropriate quality and price has enabled Suntech to sell increasingly large amounts of products, while the ability to do so at low cost has enabled it to move to profitability rapidly. This section describes Suntech’s progress in the areas of efficiency, certifications, production line yield, wafer thickness, production volume and profitability.
7.2.2.1. Efficiency Suntech was able to produce 14% efficient cells soon after the first production line was established, and was producing cells of up to 14.5% efficiency within the first few of months of production in late 2002. International PV enterprises, many of which had been operating for many years and had developed their product over a long period of time, were producing on average about 15.5% efficient cells at that time. The 14.5% efficiency, helped to enable Suntech to be profitable within the first 12 months of production. Each year, the efficiency of cells has been increased by between 0.5 and 1% (Figure 7-4). Efficiencies of 15.8% have been achieved through the development of innovative processes and device designs (described in section 7.3).
235 Chapter 7. Capability Building at Suntech Power
The first cells produced on the newest production line (using innovative technology) are 18% efficient.
Figure 7-4: Increasing Cell Efficiency at Suntech 2002-2007
18.5
18
17.5
17
16.5
16
15.5 Efficiency sc-Si cells (%) cells sc-Si Efficiency 15
14.5
14 Sep-02 Mar-03 Sep-03 Mar-04 Sep-04 Mar-05 Sep-05 Mar-06 Sep-06
7.2.2.2. Certifications The first modules produced by Suntech were found to be inadequate for demanding European customers, so Suntech initially sold modules only into local markets until the production process was operating satisfactorily. Within the first year, quality sufficient for international markets was achieved. Suntech’s modules achieved UL (Underwriters Laboratories) certification on May 1st 2006. Suntech has also attained quality certificates including ISO 9001:2000, TüV and CE certificates and international test standards including IEC61215: 1993 (Suntech Power, 2006b).
7.2.2.3. Thin Wafers & Yield When Suntech first began production, yield was around 80%, but within the first year, processing and equipment problems were resolved and 95% yield was achieved. Yield in late 2006 was 98.2%, well above the industry average, which is estimated to be between 90-95% (Hegedus & Luque, 2003; Lüdemann, 2005). With extremely low labour and equipment costs in China, and very high prices for silicon feedstock and wafers worldwide, the cost of silicon makes up approximately 80% of Suntech’s cell manufacturing costs. Starting with 270 m wafers, Suntech reported shifting production to 210 m wafers by the end of 2005 (Suntech Power, 2006c), helping to reduce production costs even while silicon prices continued to rise. Successful manufacture of cells on 180 m thick wafers with 98% yield was reported in mid 2006 (Suntech Power, 2006d).
236 Chapter 7. Capability Building at Suntech Power
Although Suntech is now able to use 180 m wafers, most suppliers are currently only producing 210 m.
7.2.2.4. Rapid Production Expansion Suntech began producing solar cells in September 2002, with an initial production capacity of 2 MWp. It produced 8 MWp of cells in 2003, increased production to 35 MWp (20 MWp monocrystalline and 15 MWp multicrystalline) in 2004, 82 MWp in 2005 and 160 MWp in 2006, always fulfilling production targets set the previous year. Suntech has expanded rapidly to achieve production scales on par with the leading international enterprises. Each increase in capacity has involved improved technology and the use of a higher percentage of locally produced equipment. The first factory was located in Wu Xi new district, a national high- technology development area in Jiangsu. The production from this factory grew to 60MW capacity by May 2005 when new facilities, built a few minutes from the first factory, began operating in September 2005. The latest planned increase includes manufacturing and R&D facilities in Caohejing high-technology development area in Shanghai. The facilities are expected to begin operations in early 2008 (Suntech Power, 2006e).
Figure 7-5: Suntech Cell Capacity and Production 2002-2006
300
250
200
150
100 Annual Production Rate (MWp) Rate Production Annual
50
0 30-Nov-02 25-Nov-03 19-Nov-04 14-Nov-05 09-Nov-06 28-Feb-03 23-Feb-04 17-Feb-05 12-Feb-06 01-Sep-02 27-Aug-03 21-Aug-04 16-Aug-05 11-Aug-06 29-May-03 23-May-04 18-May-05 13-May-06
Capacity Production
Throughout this rapid expansion, Suntech has maintained almost full capacity utilisation. Capacity utilisation varies a few percent from line to line, but overall is about 92%, and is highest for the newest lines, which are more reliable.
237 Chapter 7. Capability Building at Suntech Power
7.2.2.5. Profitability Despite rapid expansion necessitating continuous new investments in physical capital, Suntech began making a profit in its first full year of production and has maintained profitability since (Figure 7-6). As it has secured contracts for the supply of silicon, Suntech has needed to rely less on deals which involve the reciprocal sale of cells, and has been able to convert more of its cell production into modules. Hence the profits from the sales of cells have decreased.
Figure 7-6: Suntech Annual Revenues Q4 2004 - Q2 2006
200
180
160
140 )
120 Total Modules 100 Cells Integration 80 Net Revenues (US$million Revenues Net 60
40
20
0 Q4 2004 Q1 2005 Q2 2005 Q3 2005 Q4 2005 Q1 2006 Q2 2006 Q3 2006 Q4 2006
Stock exchange scouts spotted Suntech’s potential and invited the company to list on the New York stock exchange in 2005. Dr Shi bought out the other investors in the enterprise and listed Suntech on the New York Stock Exchange in December, retaining 45% of the shares. The IPO raised between US$300-350 million, and facilitated further expansions. The share price rose from $US15 to $US21.20 at launch. Since peaking at $US43.40 in January it dropped to $US26.48 by November 2006 (Figure 7- 7).
238 Chapter 7. Capability Building at Suntech Power
Figure 7- 7: Suntech Share Price December 2005 - 2006
Figure has been removed due to copyright restrictions.
Source: (Seeking Alpha, 2006)
7.2.3. Innovative Capabilities Innovative capabilities have been defined as the ability to introduce new products or processes to an enterprise. Suntech’s initial emphasis was on learning manufacturing competences using mature technology, but it began building R&D capabilities as well as research linkages from its inception. This section describes Suntech’s progress in the development and implementation of new processes, quality control routines and device designs.
7.2.3.1. New Processes A number of innovative processes have been developed at Suntech, in order to achieve better performance from standard technology. For example, through internal R&D, a new method involving selective diffusion and texturing of multicrystalline wafers via chemical processes was developed at Suntech in 2003 (Li et al., 2003). The selective emitter design was made up of a heavily doped area under the metal contacts in order to improve electrical contact and lateral conductivity, and a lightly doped area between the gridlines, to avoid high rates of recombination throughout the emitter and a consequent ‘dead layer’ in the region of the cell where long (blue) wavelengths are absorbed. The acid-based chemical etching reduced reflectance to 5% for wavelengths 300-1000 nm. The theoretical potential for the device was calculated to be 16.5%, and it resulted in average cell efficiencies of 15.8% on the production line. Suntech have also made significant progress in collaboratively developing innovative silicon purification technologies to substantially reduce silicon costs. More detail on these
239 Chapter 7. Capability Building at Suntech Power
collaborations is given in section 7.4.2 Suntech has also published research on module lamination for large area solar cells (Yuan et al., 2005) and balance of systems components (Zhaoyuan & Lin, 2005).
7.2.3.2. New Quality Control Routines Suntech has published a number of papers on various types of innovative measurements used for process optimisation, such as Photoconductivity Decay (PCD) measurements (Chen et al., 2005), which provide information about carrier lifetimes. The work for this research was conducted by Suntech staff who were on exchange at UNSW. Other published work has included the characterisation of parasitic resistances (Zhu et al., 2005a, b) and the study of material quality and its impacts on processing (Li et al., 2005). A novel quality control technique (Sustained Luminescence Testing) (Trupke et al., 2006; Trupke et al., 2007) recently developed at UNSW is also being implemented at Suntech and has already identified various previously unknown processing problems. The technique gives information about carrier lifetimes and uniformity of wafers at any stage in the production sequence, which can be used for crack detection, spatially resolved series resistance monitoring, quality control of raw material and process control of individual key processing steps such as the emitter diffusion; and is hence an extremely valuable tool in improving production techniques.
7.2.3.3. New Device Designs Collaborations with the ARC Centre of Excellence for Advanced Silicon Photovoltaics and Photonics at UNSW have included the development of ‘semiconductor finger’ technology (Wenham et al., 2005a; Wenham et al., 2005b), which has achieved around 18% efficient cells in pilot production. The technology, pictured in Figure 7-8, finds a solution to one of the major trade-offs in conventional cell design. In conventional screen printed cells, a heavily doped emitter is used to achieve good ohmic contact between the emitter and the metal contacts, and to reduce the resistivity of the emitter, through which current must flow laterally in order to reach the metal contacts. High rates of recombination of charge carriers, however, occur in the heavily doped region, reducing the current collected by the solar cell. In the ‘semiconductor finger’ design, heavily doped laser scribed grooves (semiconductor fingers) allow good lateral conductivity for charge carriers to reach the metal lines which are screen printed perpendicular to the fingers (Figure 7-9). The front diffusion of the cell is a lightly doped area, with less emitter recombination than in conventional cells. Because the semiconductor fingers do not shade the cell, they can be located close together, reducing the problem of high resistivity in the lightly doped emitter.
240 Chapter 7. Capability Building at Suntech Power
Figure 7-8: Cross section of the Semiconductor Finger emitter design developed through collaboration between UNSW and Suntech
Figure has been removed due to copyright restrictions.
Source: (Wenham et al., 2005b)
The heavily doped semiconductor fingers make good ohmic contact with the metal contacts, which can be spaced much further apart than in conventional cells, since the fingers carry the current to the contacts. Reflection of sunlight from metal contacts on the front surface of the solar cell is therefore reduced.
Figure 7-9: Screen printed lines perpendicular to heavily diffused semiconductor fingers
Figure has been removed due to copyright restrictions.
Source: (Wenham et al., 2005b) A dielectric/AR coating passivates the top surface and isolates the metal from the lightly diffused top surface. Suntech have been able to achieve up to 18% efficiency with this
241 Chapter 7. Capability Building at Suntech Power
new technology in production. The new structure can produce cells at comparable cost to conventional processes and is therefore expected to be more profitable. After a pilot phase in 2006, the technology has entered mass production in 2007. Suntech and UNSW are currently developing another technology that is expected to produce cells with 20% solar conversion efficiency on commercial production lines. Pilot production is planned to begin in 2007. The new technology involves an innovative rear metal contacting scheme to overcome the high rear surface recombination velocities associated with conventional screen-printed solar cells. Initial experimentation again shows significant promise with this work being based on an earlier UNSW patent that predated this collaborative research agreement.
242 Chapter 7. Capability Building at Suntech Power
7.3. Capability Building Strategies at Suntech
This section describes the capability building strategies adopted by Suntech, which have included: A strong commitment to internal R&D, Hiring of international experts, International research collaborations, Collaborations with suppliers of equipment and materials, The implementation of training, quality control and maintenance routines, Continuous upgrading of production lines, and The use of manual handling to reduce costs and allow flexibility in upgrading production facilities.
Internal R&D Suntech has set up an R&D laboratory from scratch, which is particularly difficult for a new enterprise in a developing country, given the specialised equipment and high standards of cleanliness required. As a percentage of total operating expenses, Suntech’s R&D budget has increased from around 10% to around 30% between 2004 and the end of 2006 (Figure 7-10). 56 full-time researchers are employed, which is a large number relative to the industry average. The cost of employing these researchers in China, however, is equivalent to employing about 8 full time researchers in Germany. In addition, much of the work Suntech carries out through collaborative research (discussed in the following sections) is not measured as part of the internal R&D budget, so the figures may not accurately reflect their innovative effort, but still indicate an increasing focus on building R&D capabilities. Suntech was also the recipient of the largest single government grant for commercialisation in 2004-2005. The grant was 4 million Yuan ($US530,000) for R&D on key technologies for the industrialisation of crystalline silicon solar cells. As a result of internal R&D, Suntech has implemented innovative processes that have increased cell efficiency to 15.8%. Production costs have also been reduced through the use of thinner wafers and improved quality control, which have relied on R&D. Suntech has progressively built the capabilities to carry out increasingly advanced R&D, and now has the capabilities to do all its own basic research.
243 Chapter 7. Capability Building at Suntech Power
Figure 7-10: Suntech Operating Expenses, General & Admin, Sales and R&D Expenses 2004-2006
20000
18000
16000
14000
12000 Total Operating Expenses General & Admin 10000 Sales R&D expenses 8000
6000 Operating Expenses (US$,000) Expenses Operating
4000
2000
0 Q4 2004 Q1 2005 Q2 2005 Q3 2005 Q4 2005 Q1 2006 Q2 2006 Q3 2006 Q4 2006
Total Operating General & R&D Expenses Admin Sales expenses R&D% Q4 2004 1808 825 803 180 9.955752 Q1 2005 2014 1098 737 179 8.887786 Q2 2005 5955 4719 966 270 4.534005 Q3 2005 6563 4720 1017 826 12.58571 Q4 2005 11038 8008 946 2084 18.88023 Q1 2006 6992 4904 1010 1078 15.41762 Q2 2006 7957 4751 1574 1632 20.51024 Q3 2006 11977 7050 2889 2038 17.01595 Q4 2006 18759 11569 3564 3626 19.32939
Hiring Suntech has recruited and trained in PV technology many people with skills in related industries, such as electronics, automation and process engineering. It has not, however, been able to satisfy its requirements for people with PV-specific expertise within the China, and has therefore needed to hire staff internationally. Suntech has recruited into key roles a number of Chinese nationals who have been trained and have gained experience overseas in order to gain industry specific knowledge. For instance, Dr Shi is one of the inventors on the key patents for the Pacific Solar technology that is now being commercialised by CSG solar in Germany and was an executive director of Pacific Solar. Other Chinese nationals include vice general manager of R&D, Dr. Tihu Wang, who previously worked at NREL in the USA; director and senior research scientist, Dr. Jingjia Ji, who previously worked at UNSW in Australia; and deputy research director Mr. Guangchun Zhang, who was previously at UNSW and Pacific Solar in Australia. These staff came to the enterprise with significant international research and production experience (Box 7-3).
244 Chapter 7. Capability Building at Suntech Power
Box 7-3: Key Staff at Suntech
CEO Dr. Zhengrong Shi is Suntech’s founder, chairman of the board of directors and chief executive officer. Prior to founding Suntech in 2001, he was a research director and executive director of Pacific Solar Pty., Ltd., an Australian PV enterprise engaged in the commercialization of next-generation thin film technology, from 1995 to 2001. From 1992 to 1995, he was a senior research scientist and the leader of the Thin Film Solar Cells Research Group in the Centre of Excellence for Photovoltaic Engineering at the University of New South Wales in Australia, the only government-sponsored PV industry research centre in Australia. Dr. Shi is the inventor for 11 patents in PV technologies and has published or presented a number of articles and papers in PV-related scientific magazines and at conferences. Dr. Shi received a bachelor's degree in optical science from Jilin University in China in 1983, a master's degree in laser physics from the Shanghai Institute of Optics and Fine Mechanics, the Chinese Academy of Sciences in 1986, and a Ph.D degree in electrical engineering from the University of New South Wales in Australia in 1992.
Chief Technical Officer Dr. Stuart Wenham has been the chief technical officer since July 2005. He is also currently a Scientia Professor and the Director of the Centre of Excellence for Advanced Silicon Photovoltaics and Photonics, at the University of New South Wales in Australia. From 1995 to 2004, he was the co-director of Research at Pacific Solar Pty. Ltd. From 1999 to 2003, he was the head of School for Photovoltaic Engineering and the director of the Key Centre for Photovoltaic Engineering at the University of New South Wales. From 1996 to 1998, he was the head of the Electronics Department and from 1991 to 1998, the associate director of the Photovoltaics Special Research Centre, also at the University of New South Wales. In 1999, Dr. Wenham received The Australia Prize for Energy Science and Technology and in 1998, the Chairman's Award at the Australian Technology Awards, in both cases jointly with Martin A. Green. Dr. Wenham received his Ph.D. degree in electrical engineering and computer science from the University of New South Wales in Australia in 1986.
Senior Research Scientist Dr. Jingjia Ji is a director and senior research scientist at Suntech and has been with the enterprise since 2003. From 1995 to 2002, Dr. Ji worked as a senior research scientist at Pacific Solar Pty., Ltd.. From 1991 to 1994, he worked at the University of New South Wales as a senior research assistant. From 1985 to 1990, he worked in the Shanghai Institute of Organo-Fluorine Materials in China as the head of the department of chemical engineering. Dr. Ji received his bachelor's degree in chemical engineering from the East China Institute of Chemical Technology in China in 1983, and a Ph.D degree in industrial chemistry from the University of New South Wales in Australia in 1994.
Manager R&D Mr. Yichuan Wang is a manager of Suntech’s PV cell research and development department and has been with our enterprise since 2001. From 1979 to 2001, he worked at Yunnan Semiconductor Co., Ltd. on the research, development and manufacturing of PV products. From 1996 to 2000, he worked on sci-tech planning projects organized by the PRC Ministry of Science and Technology. In 1984, he participated in the introduction of the PV cells manufacturing line. Mr. Wang received his bachelor's degree in physics from Yunnan University in China in 1968.
Dr. Tihu Wang joined Suntech as vice general manager of R&D. Dr. Wang has 23 years of experience in leading and conducting advanced scientific research on high-efficiency solar cells, semiconductor crystal growth, material characterizations, and physical metallurgy. Prior to joining Suntech, Dr. Wang worked at the U.S. National Renewable Energy Laboratory where he acquired extensive expertise in the entire line of silicon photovoltaic technology, from silicon feedstock production, crystal growth and solar cell manufacturing, to thin- film silicon fabrication.
Director R&D Mr. Guangchun Zhang is Suntech’s deputy research director of research and development and has been with the enterprise since November 2005. He specializes in research on high-efficiency solar cell design. From January 2003 to October 2005, Mr. Zhang was a professional officer at the Centre for Photovoltaic Engineering and the School for Photovoltaic Engineering at the University of New South Wales. From 1997 to 2002, Mr. Zhang was a research engineer at Technology Development Group and was seconded to Pacific Solar Pty. Limited from the University of New South Wales. From 1994 to 1996, he worked at the Photovoltaics Special Research Centre and the Centre for Photovoltaic Devices and Systems, also at the University of New South Wales. From 1982 to 1994, Mr. Zhang taught and researched at the School of Electronic Engineering at Shandong Polytechnic University in China, first as an assistant lecturer, then as a lecturer and finally as an associate professor. Mr. Zhang received his bachelor’s degree and his master’s degree in 1982 and 1988, respectively, from the School of Electronic Engineering at Shandong Polytechnic University.
Source: (Suntech Power, 2006c)
The recruitment of these PV experts, in particular Dr Shi as CEO and Professor Wenham as CTO, has been critical to Suntech’s success. Their knowledge and experience of the 245 Chapter 7. Capability Building at Suntech Power
PV industry and the direction of technological change within it has allowed the enterprise to make optimal technology decisions in relation to production, R&D and strategic investments. Suntech has been able to attract Professor Wenham to the enterprise as a result of its flexible approach to employment and IP arrangements. Professor Wenham is employed on a contractual basis, and spends only some of his time at Suntech, while he carries out other roles at UNSW in Australia.
Research Collaborations Much of Suntech’s research is carried out through collaborations with the University of New South Wales (UNSW) in Australia. In December of 2002, Suntech signed a technical cooperation agreement with the Centre for Photovoltaic Devices and Systems (now the Centre of Excellence for Advanced Silicon Photovoltaics and Photonics) at UNSW. During the 2nd quarter of 2006, the agreement was extended until the end of 2010 (Suntech Power, 2006d). The collaborations have included the full spectrum of technology development, from basic research to commercialisation. The research is of particular value to Suntech, as it provides access to leading edge research, but is also closely aligned with production goals, compatible with manufacturing using existing infrastructure and equipment and involves technologies which are close to commercialisation. Despite having built a strong base of internal innovative capabilities, which enable production to be adjusted and improved, Suntech is still looking to collaborative research with UNSW to acquire new cutting edge technology. Suntech has contributed 50% or more to a number of research projects in conjunction with UNSW. For example, the semiconductor finger research previously described in section 7.2.3.3 was developed through a Sutech-UNSW collaboration which was wholly funded by Suntech, including research conducted both in China and at UNSW. The cash contribution alone for the research conducted at UNSW has been A$500,000 for 2006 ($US375,000). Suntech also funded all the living, travel and accommodation expenses for the UNSW staff and students who worked at Suntech on the collaboration. Two factors facilitate Suntech’s participation in these research collaborations. The first is the strong existing links between key Suntech staff and UNSW. Having spent fifteen years studying and working at UNSW and Pacific Power, Dr Shi has strong links with the Australian organisations. Suntech's Chief Technology Officer, Stuart Wenham, who is also Scientia Professor and Director of the UNSW PV Centre, leads the research collaborations. The second factor that facilitates Suntech’s participation in collaborations with UNSW is its flexible approach to IP arrangements. Suntech does not aim to own the technologies that are developed outright, but rather to access them for its own production. Joint patent ownership arrangements have been negotiated between UNSW and Suntech, such that Suntech has the right to use the technology, while UNSW retains the publication rights it requires as an academic institution and
246 Chapter 7. Capability Building at Suntech Power
the IP rights to license the technology to others, in which case Suntech and UNSW share the resultant income. Suntech-UNSW research collaborations have also provided opportunities for the exchange of personnel for research and training activities. Suntech also intends to send students to complete the photovoltaics degree at UNSW. There are informal learning and training opportunities through these exchanges, as well as potential recruitment opportunities. Suntech also collaborates with enterprises in other parts of the supply chain (described in the following sections), and with several universities in China to conduct materials research such as experimenting with lower purity silicon.
Supporting Local Production Equipment Suntech’s strategy has been to assist the development of low-cost, high quality Chinese manufactured production equipment by working with equipment manufacturers. Its first production line was predominantly made up of second-hand equipment purchased cheaply from the US, because the PV manufacturing equipment industry in China was in its infancy, the quality of equipment was unknown and Suntech couldn’t afford the time and expense to test it. The only things initially purchased in China were some of the chemical benches. Suntech has progressively moved from imported to local equipment as it has become feasible to source equipment locally. In the first expansion, some of the equipment was purchased in China, but the major plant came half price from a Japanese manufacturer, along with some second hand pieces of equipment from Italy. Production equipment of good quality is now made very cheaply in China. Suntech now purchases all of the non-critical equipment in China, including furnaces for diffusion and high temperature processes, and all of the chemical benches. Critical equipment is still imported, usually from Europe, because Chinese suppliers have not been able to achieve the quality required for sensitive processes. For instance, Suntech purchases PECVD equipment from Centrotherm in Germany and screen printers from Baccini in Italy. When Suntech has developed proprietary processes, it has worked closely with and assisted local suppliers to design equipment to satisfy its requirements. For example, it developed a new method of texturing multicrystalline wafers via chemical processes in 2003 (Li et al., 2003). The baths for the texturing are relatively sophisticated and have particular capabilities. These baths have been custom designed and made in China. The lasers for the new ‘semiconductor finger’ technology previously described, which was being put into large scale production in late 2006, are also being custom made in China. A laser has been designed that can achieve a sufficiently high throughput, with the capabilities required for this technology, but without additional unnecessary capabilities that would have increased the price.
247 Chapter 7. Capability Building at Suntech Power
Suppliers of Pastes, EVA, Glass and Chemicals Suntech initially attempted to use locally available materials, but it was found that the local pastes, EVA and chemicals were inadequate to achieve the quality required in export markets. Suntech stopped using local materials in its production while development work was carried out to improve their quality so that they matched the quality of materials available internationally. Suntech progressively switched to the Chinese materials once the quality was suitable. In order to achieve the quality required, Suntech worked with materials suppliers, giving them iterative feedback until the materials gave good results. Problems initially found with locally made pastes included: Insufficient conductivity Insufficient adhesion Reaction with the encapsulation material Bowing of wafers during firing of the contacts
In general the Chinese chemicals for the wet chemistry process were more suitable from the start, but in a minority of cases the purity needed to be improved before Suntech could use them. Suntech also worked with Chinese manufacturers to improve their EVA formulation. It has been importing glass to date, but has been evaluating Chinese glass and trying to improve the suitability of PV glass in China. As these problems are progressively resolved, Suntech is able to access lower cost, good quality local products, increasing their competitive advantage.
Silicon Suppliers Suntech has worked closely with Chinese supplier Luoyang Silicon to improve its proprietary technology for purifying silicon in order to produce feedstock suitable for making Czochralski ingots. Suntech carried out extensive research on the performance of wafers with different impurities after different processes, such as lasering and high temperature processes and under different processing conditions, as published in (Li et al., 2005). Suntech provided feedback on the suitability of the silicon being manufactured and also gave Luoyang some financial support to do some of the R&D work, and to move to commercial scale production of the technology. Suntech and Luoyang have now set up a joint venture to manufacture solar cells using all of the silicon that is produced from the new Luoyang silicon plant. The new factory is adjacent to the original Luoyang plant. Suntech owns 89% of the joint venture and the cells will be branded Suntech. By late 2006, 0.2 .cm p-type 200 m silicon wafers of 99.99% purity with a minority carrier lifetime of around 0.5 s had been manufactured at a quarter of the price
248 Chapter 7. Capability Building at Suntech Power
of the usual high purity wafers. It is predicted that by reducing wafer thickness from 200 m to 150 m the cost could be further halved. Suntech has also invested US$5 million in the Emei Semiconductor Material Factory, which produces solar-grade silicon. The investment has allowed Emei to triple production to 300 metric tons annually (Hirshman, 2005). The investments in Chinese silicon production have allowed Suntech to access low cost silicon, but also to avoid the long-term supply contracts with controlled prices that silicon manufacturers elsewhere are currently demanding as a prerequisite for expanding their production capacity.
Training and the Virtual Production Line While the CTO, Professor Wenham, initially trained Suntech’s production line engineers, the capability for the continuing training of these engineers has been developed internally and is facilitated by the use of the Virtual Production Line (VPL) co-developed by the author as a training tool for production line engineers, and further discussed in chapter 6. 1-2 week visits by Professor Wenham to Suntech during 2002 were dedicated to tuning the production line. Training of staff running the production line, who were not from a PV background, started immediately after the production line was operating acceptably, because Professor Wenham was only able to be in China intermittently, and the people making decisions on the line needed to have the skills to keep the line working well, including the monitoring, analysis and fault diagnosis. This training focused on optimisation of processing parameter interdependencies throughout the production line and the use and interpretation of quality control tests. Professor Wenham visited Suntech twice in 2002, four times in 2003, and six times in 2004, 2005 and 2006 for 1-2 weeks at a time. Each time, approximately three 2-3 hour training sessions were held, in which groups of 20 engineers were trained. Much of the teaching was carried out using the VPL, which can be conducted at low cost and without disrupting the operation of the production line. Professor Wenham has written an exam to strengthen the training. Professor Wenham no longer trains the production line operators, who are now trained by senior plant operators. Suntech has built a critical mass of people with the knowledge to do the training. Professor Wenham now spends more time managing and directly training the research team, which he indirectly leads in his capacity as CTO. This training is related to technology development and running of the pilot lines, and ensures that the knowledge embodied in Professor Wenham is transferred to the research team.
Monitoring and Quality Control Suntech has focused on developing a comprehensive quality control system, and has been able to invest more person-hours into quality control than manufacturers in industrialised
249 Chapter 7. Capability Building at Suntech Power
countries, due to the low cost of labour in China. Production line engineers have been educated to understand and use many simple tests to maintain the quality along the line through the Virtual Production Line software. Some innovative monitoring techniques, such as the Sustained Luminescence test described previously, have also been implemented. Operators are trained in these techniques and are motivated by yield and throughput targets associated with pay bonuses. Graphs of performance against these measures are often displayed on the wall. The monitoring of production through these quality control tests has provided information for the continuous revision of production line tuning and standard operating procedures.
Manual Handling and Upgrading Production Lines Through experience of implementing automated production, Suntech has found that not only is manual processing cheaper in the context of low labour costs in China, but has resulted in higher yields and has allowed production lines to be upgraded more easily. Yields are generally higher for automated processes in a mature manufacturing industry, and it has been assumed that the same would apply to the PV industry (Eberhardt, 2005; Lüdemann, 2005; Swanson, 2004). However, since PV manufacturing technologies are still evolving, equipment needs to be periodically redesigned to handle the wafers in a different way for a different type of processing. There has therefore been insufficient experience gained with automated cell handling equipment to achieve the high yields that would emerge when a stable process has been used for many years. Suntech has implemented the latest technology developed through R&D as it builds new production lines. Manual handling has also allowed Suntech to upgrade existing production lines rapidly and at lower cost. Using expensive highly automated equipment with a 10 year depreciation schedule commits a manufacturer to a process, with limited room for variation, delaying the adoption of innovations. Conversely, manual operations have allowed Suntech to reconfigure their assets and bring new technologies to commercialisation rapidly.
Maintenance and Operation of Equipment In countries where the labour costs are very high, it is often most cost-effective to commission technicians from the supplier to repair equipment. In contrast, Suntech has a team of trained maintenance technicians standing by in case of equipment failure. When equipment is imported, Suntech usually negotiates a deal with the supplier to train its technicians to maintain the equipment. Suntech’s rapid expansion has given it good leverage in negotiations with equipment suppliers who would normally prefer a maintenance contract. Good performance of equipment and therefore high yields has resulted from effective monitoring of the equipment and routine preventative maintenance at Suntech. The knowledge built through experience in maintenance has also enabled Suntech to collaborate with equipment suppliers in the design of new production line equipment.
250 Chapter 7. Capability Building at Suntech Power
OEM Manufacture, Horizontal Expansion and Downstream Expansion Suntech has expanded its markets and acquired new capabilities through OEM manufacture and acquisitions into the downstream parts of the PV value. In 2005-2006, Suntech produced modules under an OEM agreement with the German enterprise SolarWorld, which supplied specifications and technological know-how (Schmela, 2005c). In 2006, Suntech purchased the Japanese BIPV manufacturer MSK, giving Suntech access to the Japanese market via MSK’s well developed sales and distribution network; and to MSK’s product development capabilities, systems integration, commissioning and maintenance expertise. Suntech has moved MSK’s manufacturing to China and plans to integrate the two enterprises and combine many overlapping areas of the operations, including manufacturing, sales, purchasing, R&D and back office functions (Suntech Power, 2006a). The acquisition has the potential to provide great synergies by combining the complementary capabilities of the two enterprises. It represents an interesting new development in industrialising country takeovers of enterprises from industrialised countries. Suntech does some module assembly, but also provides cells to other OEM enterprises that it has helped to establish. The OEM enterprises benefit from a guaranteed supply of cells and marketing opportunities, while Suntech continues to expand and increase market share. In the first quarter of 2006, Suntech won system integration contracts in China of 2.2MW (Suntech Power, 2006c). In August 2006, Suntech announced the founding of Suntech America Inc., a subsidiary that will help Suntech access US markets (Hirshman, 2006). The enterprise also established Suntech Power (Hong Kong) Co., Ltd. in the first half of 2006, in order to improve international purchase and sales linkages (Suntech Power, 2006d). Shenzhen Suntech was set up in 2006 to carry out solar power grid integration projects in Southern China. New manufacturing, R&D facilities and a sales office were also established in Shanghai to develop sales in the domestic PV market (Suntech Power, 2006e).
251 Chapter 7. Capability Building at Suntech Power
7.4. Analysis of the Case Study Using the Framework
The preceding sections of this chapter have identified the capabilities achieved and the capability building strategies adopted by Suntech between 2002 and 2006. The framework developed in chapter 5 will now be used to identify the factors which have supported or constrained different types of learning at Suntech and the strategies that have been used to build capabilities and overcome deficiencies in the technological system. The discussion in chapter 10 of this thesis will the PV industry literature previously reviewed to suggest the extent to which these findings can be generalised. This case study also provides useful verification of the explanatory power of the framework against actual developments at a national and an enterprise level.
7.4.1. The Development of Chinese PV Technological System The part of the framework dealing with the Chinese technological system is shown in Figure 7-11, (a) as it was prior to 2000 and (b) in 2007. In each case, functions where the technological system was effective in supporting enterprise-level learning are represented by red lines and shading. Blue is used to indicate areas where the technological system has partially supported learning, while functions left black have not been successfully performed by the system. Each of the functions of networks in the technological system is now discussed, and the ways that the institutional environment and the technological trajectory has influenced the operation of networks are identified. It should first be noted that prior to 2000, enterprises had little influence over institutional arrangements, which were decided by the state on the basis of broad industrialisation policies. Private enterprises are becoming more influential as China attempts to reach the forefront of industrialisation and their importance is increasingly recognised by the state.
252 Chapter 7. Capability Building at Suntech Power
Figure 7-11: The status of the Chinese technological system viewed through the framework (a) before 2000, and (b) in 2007.
TECHNOLOGICAL TRAJECTORY
influence perceptions of the provide direction for technological paradigm search
Chinese Technological System
Networks
OTHER ACTORS influence policy and institutions
MARKET & NON-MARKET INTERACTIONS influence operation of markets influence connectivity INSTITUTIONS
knowledge creation & exchange resources for production and innovation investment opportunities provide incentives to invest and improve alter allocation of resources direction & incentives
ENTERPRISE
(a)
TECHNOLOGICAL TRAJECTORY
influence perceptions of the provide direction for technological paradigm search
Chinese Technological System
Networks
OTHER ACTORS influence policy and institutions
MARKET & NON-MARKET INTERACTIONS influence operation of markets influence connectivity INSTITUTIONS
knowledge creation & exchange resources for production and innovation investment opportunities alter incentives to invest and improve alter allocation of resources direction & incentives
ENTERPRISE
(b)
253 Chapter 7. Capability Building at Suntech Power
Investment Opportunities Early Chinese manufacturers produced primarily for small domestic markets, which provided limited opportunities for learning by doing. Institutional arrangements reinforced market monopolies by state-owned enterprises, while market expansion was not strongly enough supported by policy to warrant large investments in PV manufacture. There were therefore few opportunities for non-state-owned manufacturers to participate in the technological system. The primarily state-owned Chinese manufacturers also lacked funds or a mandate to invest in innovative activities and were therefore unable to conduct effective R&D or innovate in order to upgrade the technology used on their production lines. Investment opportunities are considered to have been insufficient in 2000, and the function is therefore not highlighted in Figure 7-11a. Chinese policy and institutional arrangements since 2000 have allowed, and in fact encouraged Chinese manufacturers to access export markets, and have created opportunities for foreign investment, which has allowed enterprises such as Suntech to expand and has stimulated growth in the number of actors in the Chinese PV industry. In addition to the entry of new cell manufacturers, the success of Chinese module and cell manufacture in global markets has spurred investment in equipment and materials manufacture in China. China has also provided investment opportunities by expanding domestic markets. Influencing operation of markets is therefore shaded red. Since the technological system was able to successfully provide investment opportunities in 2007, it is highlighted in red in Figure 7-11b.
Resources for Production and Innovation Prior to 2000, the Chinese PV technological system was not able to supply personnel with PV production or R&D experience, or materials and equipment of sufficient quality required for export or cost effective PV production. Existing public R&D programmes were underperforming and did not support commercialisation. The institutional arrangements did not allocate sufficient resources to PV R&D or manufacture to compensate for deficiencies in the technological system in order to encourage strong investment in innovation or production. Low labour costs, and the existence of basic infrastructure for manufacturing and high technology industries, however, gave China a good basis for competitive advantages in PV manufacturing. Resources for production and innovation are therefore marked blue in the diagram for 2000. The number of materials and equipment suppliers in the technological system and their capabilities has developed over the past five years, and all but the most complex and sensitive pieces of equipment can now be supplied at low cost and good quality within China. Tax holidays, import duty exemptions and investment grants have allocated resources towards the industry, encouraging investment. The funding of Chinese research institutes in the PV area has also been increased to become comparable with that of the leading PV countries, but the level of
254 Chapter 7. Capability Building at Suntech Power
research still lags well behind in terms of laboratory cell efficiencies achieved, for instance. It will take some time for China to build up a critical mass of expertise in PV research and commercialisation such that it can compete with the leading countries. In the meantime, Chinese enterprises will continue to rely on international linkages for cutting edge research. In addition, the supply of skilled personnel has not been able to keep pace with the growth of the industry. Since the technological system is now able to supply virtually all the resources required for production and innovation, this factor is highlighted red in the diagram for 2007.
Direction and Incentives for Search State-owned Chinese PV manufacturers historically had little incentive to innovate, since they were not influenced by markets or the international technological paradigm, but were more concerned with keeping the old production lines running in order to fulfil production targets. Due to the failure to address the need for technological improvements, the production cost was quite high and thus these enterprises could not operate profitably. This factor is therefore unshaded. Since the 1990s, producers have been operating independently of the government, and since 2000 have been driven mainly by export markets. These export markets have stimulated growth and have provided incentives for cost reductions and quality improvements, indicated by a red line connecting the technological trajectory and the Chinese network of manufacturers. The institutional arrangements have facilitated connectivity to export markets and international best practices by allowing non-government owned enterprises to enter the market, encouraging foreign investment and export production. Direction and incentives for search are shaded red for the case of 2007. Although the standards and testing facilities encouraged improvement in module manufacture, most of the incentives came from the international technological trajectory through export markets (indicated by red text), while the incentives provided by institutions within China do not demand such high quality for domestic markets and are therefore blue.
Knowledge Creation and Exchange State-owned Chinese PV manufacturers were disconnected with international markets, suppliers and research organisations prior to 2000. The institutional arrangements that governed the national research organisations also kept them isolated from manufacturers. The small number of actors in the Chinese technological system and their lack of connectivity to each other and to international markets or research activities limited the extent of knowledge exchange and learning by interacting, this factor is therefore unshaded for the case of 2000. Since 2000, the entry of new non-state owned PV manufacturers, with international technical expertise and good R&D linkages has increased the interactions with international industry actors and best practices. Learning by interacting with suppliers has resulted in an increase in capabilities within PV manufacturers and supplier enterprises. The development of
255 Chapter 7. Capability Building at Suntech Power
locally produced, low cost equipment and materials have enabled manufacturers to reduce their production costs, while maintaining good quality products. Manufacturers and suppliers have developed knowledge, skills and complementary capabilities in relation to the properties of materials and their impact on cell performance and equipment design and operation through these interactions. Both public and private R&D expenditure has increased, and Chinese manufacturers have improved their production technology in order to access export markets with good quality and low cost products. China now not only accesses, but influences the international trajectory, introducing a new paradigm of low cost manufacturing. Knowledge creation and exchange in China is given red status for 2007.
256 Chapter 7. Capability Building at Suntech Power
7.4.2. Learning at Suntech Figure 7-12 shows the part of the framework concerned with learning at Suntech (a) at its formation in 2002, and (b) in 2007. The capabilities and factors that influenced learning in each case are highlighted red where they strongly supported learning, blue where they were partially functioning and not highlighted where they were insufficient.
Figure 7-12: Learning at Suntech (a) in 2002, and (b) in 2007.
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION incentives & resources PRODUCTION direction INNOVATIVE incentives & CAPABILITIES for search CAPABILITIES resources ROUTINES Learning by Doing Learning by Searching new production techniques DOING R&D Improvement Capabilities
RECONFIGURATION
(a)
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION incentives & resources PRODUCTION direction INNOVATIVE incentives & CAPABILITIES for search CAPABILITIES resources ROUTINES Learning by Doing Learning by Searching new production techniques DOING R&D Improvement Capabilities
RECONFIGURATION
(b)
257 Chapter 7. Capability Building at Suntech Power
Progress in each of the three types of learning in the framework will now be discussed.
7.4.2.1. Learning by Doing Figure 7-13 shows the part of the framework concerned with learning by doing at Suntech (a) at its formation in 2002, and (b) in 2007. As with any new enterprises, production capabilities and improvement routines such as quality control and training were initially limited, and mainly confined to the capabilities of the key staff, who had prior experience running PV production lines. These capabilities are therefore highlighted blue.
Figure 7-13: Learning by Doing at Suntech (a) in 2002, and (b) in 2007
production resources production resources investment opportunities investment opportunities
PRODUCTION PRODUCTION CAPABILITIES CAPABILITIES incentives & incentives & resources ROUTINES resources ROUTINES Learning by Doing Learning by Doing
DOING DOING
(a) (b) By 2007, Suntech had achieved better production capabilities (yield, efficiency and low cost) than most PV manufacturers using the mature technology for manufacturing wafer-based silicon solar cells. The production of many units has no doubt enabled Suntech to learn how to better carry out fine-tuning, process control and maintenance from the experience of production, and to obtain economies of scale in production. However, its higher than average cell efficiencies and yields, and its rapid achievement of these are better explained by a particular effort in the implementation of improvement routines. Quality control and maintenance routines have enabled Suntech to monitor and learn from its production experience. Internal training has ensured that the knowledge within the organisation, initially that of the CTO, and later of the experienced production line engineers, has been diffused throughout the organisation. Production capabilities and routines are therefore shaded red in the diagram representing 2007.
Investment Opportunities Suntech has been able to increase the scale of its production to become the third largest producer of solar cells in the world. While the Chinese market is not sufficient to absorb all of the production from Suntech, growing export markets have provided investment opportunities for Suntech. In this respect, Suntech was fortunate to begin production coincident with the beginning of the German market expansion after the introduction of the feed in tariffs in 2000, and investment opportunities are therefore shaded blue in the diagram for 2002. Since 2002, there has been increasing recognition of the quality of the Chinese products in export markets,
258 Chapter 7. Capability Building at Suntech Power
increasing the attractiveness of investment in Chinese solar cell manufacture. The opportunity to attract private investors and later list on the stock exchange has also give Suntech access to capital for expansion and investment in R&D. Investment opportunities are shaded red in the diagram for 2007.
Resources for Production Where the Chinese innovation system has not been able to supply the resources required, as indicated by no shading in the diagram for 2002, Suntech has adopted a variety of strategies to overcome these deficiencies, including: International hiring where human capital has not been available, The import of production equipment and later collaboration with suppliers of equipment and materials to accelerate the development of capabilities within the suppliers, Aggressive negotiation of international silicon supply contracts and investment in domestic silicon manufacturers.
The improved availability of production equipment and materials of appropriate quality at low cost has contributed to Suntech’s production capabilities, as indicated by red shading for 2007. This availability has arisen largely through the learning of suppliers via learning by interacting with cell manufacturers. Through aggressive financial and technical support of suppliers, Suntech has played a big part in the development of the Chinese PV value chain. Suntech has been instrumental in the learning by interacting process, and may be considered a prime mover in this respect.
7.4.2.2. Learning by Searching Beginning with only the innovative capabilities embodied in the key staff, Suntech has pursued a strategy of continuous improvement and upgrading of product and process technology, the development of improved equipment and materials and quality control tests. To this end, it has invested robustly in R&D facilities, staff and research collaborations. Experience in learning by searching has increased Suntech’s innovative capabilities to the point where it can now undertake its own basic R&D. While much of Suntech’s most important R&D activities have been conducted through research collaborations, the collaborative research depends on these internal capabilities. While learning by doing provides a partial explanation for the progress of Suntech, the introduction of innovative processes, device designs and quality control procedures has been responsible for many of its improvements. Suntech has been able to pursue a successful R&D strategy, largely because of the prior expertise of key staff and their continuing involvement in the research, and because of strong links with UNSW and flexible IP arrangements. There has
259 Chapter 7. Capability Building at Suntech Power
been a continuous effort in commercialising new technologies, which has been made possible through the flexibility of manual processing.
Figure 7-14: Learning by Searching at Suntech (a) in 2002, (b) in 2007.
innovation resources innovation resources investment opportunities investment opportunities
COORDINATION & COORDINATION & INTEGRATION INTEGRATION
incentives, incentives, direction INNOVATIVE resouces direction INNOVATIVE resouces for search CAPABILITIES & direction for search CAPABILITIES & direction for search for search
Learning by Searching Learning by Searching
R&D R&D
new production new production RECONFIGURATION RECONFIGURATION technique technique (a) (b)
Reconfiguration and New Production Techniques Suntech has rapidly introduced innovative technology to their production lines. The ability to reconfigure production successfully has depended on rapid expansion, whereby new plants use the latest technology; and on manual processing, which has allowed Suntech to upgrade the older production lines at much lower cost than if they were dependent on automated cell handling equipment, which can only be upgraded at high cost. This factor is therefore shaded red.
Innovation Resources and Knowledge There is a shortage of people with PV R&D experience and expertise in China, and solar cell research in China is well behind that of the world leaders. It has been observed that while government funding for PV R&D in China has increased to be comparable with that of the leading PV nations, research outputs from Chinese institutes still lag behind the world’s best. There is a shortage of personnel in China with PV R&D expertise and experience. Although innovation resources in China have improved, they still lag behind the leading industrialised PV nations (indicated by blue shading for 2007). While Chinese manufacturers have previously been constrained by these factors, new manufacturers, including Suntech, have sought out international research collaborations and hired personnel with international R&D experience in order to access the technology and personnel required. New manufacturers, which are not state–owned, especially those with foreign ownership or contacts, have been able to attract investment capital which has facilitated investments in R&D. Despite being a new firm with emerging innovative capabilities (blue), Suntech has been able to introduce new production techniques (red).
Direction and Incentive for Search With exposure to international markets, search at Suntech has been driven by the global
PV technological paradigm that sees high costs per Wp production as the primary problem for
260 Chapter 7. Capability Building at Suntech Power
the technology, and reduced use of silicon material, higher efficiencies and lower cost production techniques within the range of solutions. Although the PV industry China was initially very isolated, so external direction and incentives were unshaded, export and research collaborations have improved the international linkages, indicated by blue. Production line monitoring has also strongly informed the direction of search at Suntech which has therefore been coordinated with production priorities, indicated by red.
7.4.2.3. Learning by Interacting Figure 7-15 shows the part of the framework concerned with learning by interacting at Suntech.
Figure 7-15: Learning by Interacting at Suntech (a) in 2002, (b) in 2007.
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
(a)
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
It is clear from the previous discussion that much of Suntech’s success in building production and innovative capabilities has depended on interactions with other actors, including research collaborations, interactions with suppliers and hiring. As previously discussed, in 2002, links between actors in the Chinese PV innovation system were limited, indicated by a blue arrow but by 2007, connectivity between Chinese manufacturers and other domestic and international actors was much improved, indicated by a red arrow. Suntech has been a prime mover in bringing about learning in the Chinese PV supply chain by investing in and collaborating with local suppliers. From its inception, Suntech has engaged in learning by interacting through collaborative research, which has enabled Suntech to
261 Chapter 7. Capability Building at Suntech Power
access expertise and facilties at low cost, where its internal innovative capabilities have been inadequate. Interactions and vertical integration with downstream actors in the supply chain have given Suntech access to new markets and the capabilities for manufacturing BIPV products and carrying out some of the functions of downstream delivery of products to markets. The hiring of international experts has enabled Suntech to overcome the inability of the innovation system to supply human capital and has been critical to Suntech’s production and innovation capabilities.
Coordination and Integration Research collaborations at Suntech have been coordinated to satisfy priorities on the production line. Close links between research and manufacturing have enabled the technology to be commercialised within a short period of time. Coordination and Integration is therefore red.
Investment and Linkage Capabilities The experience and familiarity of Dr Shi and Professor Wenham with the latest developments in PV R&D and close links with UNSW have given Suntech access to the leading edge of the international technological paradigm, and therefore the knowledge of what technologies to invest in, what people, equipment and materials to buy and what parts of the value chain to protect and/or invest in. In addition to the strong links between Suntech and UNSW that arise through Dr Shi and Professor Wenham, Suntech’s linkage capabilities are enhanced by a strategy which places more value on continuing access to innovative technologies that can give it immediate competitive advantage, rather than the ownership and protection of IP rights. Investment and Linkage capabilities are shaded red in both 2000 and 2007.
262 Chapter 7. Capability Building at Suntech Power
7.5. Conclusion
In this chapter, the capability building strategies and the factors in the Chinese technological system that have influenced learning at Suntech have been identified through the use of the framework developed in chapter 5. The ability of the framework to explain capability building has been verified, as it has readily identified key aspects at the Chinese national level and at the Suntech enterprise level which are now being done well, as well as those which remain inadequate. The discussion chapter of this thesis will refer to this case study, two other case studies and the pre-existing literature on the PV industry in order to further verify the framework and then to draw out the implications of the findings for capability building in PV manufacturers in developing countries and for the PV industry more broadly.
263 Chapter 7. Capability Building at Suntech Power
References
Broaded, C.M. (1993), China's response to the brain drain, 37 (3), p 277. Chen, R., Fisher, K., Li, H., Ji, J., Wang, Y. and Shi, Z. (2005), Using PCD method to optimize standard BCSC processes, 15th International Photovoltaic Science and Engineering Conference (PVSEC), Shanghai, China, 10-15 October. Dai, Y., Shi, Z. and Xi, X. (1999), Technological Innovation of the Chinese Photovoltaic Industry, Prepared for the Center for the Integrated Study of the Human Dimensions of Global Change, Carnegie Mellon University, Beijing, China. Eberhardt, K. (2005), PV Production Facilities adapted to Technology Requirements - a Prerequisite for cost effective Mass Production, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. Green, M.A., Emery, K., King, D.L., Igari, S. and Warta, W. (2003), Solar Cell Efficiency Tables (Version 21), Progress in Photovoltaics: Research and Applications, 11 (1), pp 39–45. Green, M.A., Emery, K., King, D.L., Igari, S. and Warta, W. (2005), Solar Cell Efficiency Tables (Version 25), Progress in Photovoltaics: Research and Applications, 13 (1), pp 49–54. Hegedus, S.S. and Luque, A. (2003), Status, Trends, Challenges and the Bright Future of Solar Electricity from Photovoltaics, in Luque, A. & Hegedus, S.S. (eds), "Handbook of Photovoltaic Science and Engineering", John Wiley & Sons. Hirshman, W.P. (2005), Is there room in China‘s boom for PV?, Photon International, April 2005, pp 54-58. Hirshman, W.P. (2006), Suntech sets up US subsidiary as CEO joins NYSE international group, Photon International, September 2006. Hirshman, W.P., Hering, G. and Schmela, M. (2007), Gigawatts - the measure of things to come: Market survey on global solar cell and module production in 2006, Photon International (March 2007), pp 136-166. Hug, R. and Schachinger, M. (2006), Chinese Solar Modules Penetrating the German Market, Solar Reports, Solarserver Forum for Solar Energy. IEA PVPS (2006), IEA Photovoltaic Power Systems Programme International Statistics, Accessed from: http://www.iea-pvps.org, on: February 2007. Jäger-Waldau, A. (2006), Research, Solar Cell Production and Market Implementation of Photovoltaics, PV Status Report, European Commission. Lewis, J. and Wiser, R. (2005), Fostering a Renewable Energy Technology Industry: An International Comparison of Wind Industry Policy Support Mechanisms, Environmental Energy Technologies Division, Ernest Orlando Lawrence Berkeley National Laboratory. Li, C. (2004a), Bringing China’s Best and Brightest Back Home: Regional Disparities and Political Tensions, China Leadership Monitor, 11. Li, H., Ji, J., Wang, Y. and Shi, Z. (2005), Effects of High Temperature Process on CZ Materials From Different Suppliers, 15th International Photovoltaic Science and Engineering Conference (PVSEC), Shanghai, China, 10-15 October. Li, H., Wang, Y., Ji, J., Chen, R., Xu, Y., Shao, A., Tong, C. and Shi, Z. (2003), Implementation of selective diffusion in large-scale production of mc-Si solar cells, in 3rd World Conference on Photovoltaic Solar Energy Conversion, Osaka, Japan, pp 967-970 Vol.961. Li, Z. (2004b), Made In China, Renewable Energy World, 7 (1), pp 70-80. Li, Z. (2006), From Myth to Magic: PV investment in China, 21st European Photovoltaic Solar Energy Conference, Dresden, Germany, 4-8 September 2006. Liu, X. and White, S. (2001), Comparing innovation systems: a framework and application to China's transitional context, Research Policy, 30 (7), pp 1091-1114.
264 Chapter 7. Capability Building at Suntech Power
Lüdemann, R. (2005), Experience and Expectation of Silicon Solar Cell Mass Production - Requirements for Next Generation Equipment, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. Marigo, N. (2006), The Chinese silicon photovoltaic industry and market: a critical review of trends and outlook, Progress in Photovoltaics: Research and Applications, In Press, Corrected Proof. Photon (2006), Chinese official says reposrts of PV feed-in tariff are incorrect, Photon International, January 2006, p 38. Pichel, J. (2006), Consider Investing in Solar-Grade Polysilicon Companies, March 2006, Accessed from: RenewableEnergyAccess.com, on. REDP (2004), China PV Industry Development Report, Wang, S. (ed), NDRC/GEF/The World Bank China Renewable Energy Development Project Management Office, Beijing. Schmela, M. (2005a), Baoding Yingli: from crystal to solar modules, Photon International, September 2005, pp 58-61. Schmela, M. (2005b), Green light for solar power: China‘s PV industry awakens, Photon International, September 2005, pp 56-57. Schmela, M. (2005c), Suntech Power: a global brand made in China, Photon International, September 2005, pp 64-69. Schmela, M. (2005d), Super Sonic Solar Market - Worldwide Market Survey - cell & module production, Photon International (March 2005), pp 66-82. Schmela, M. (2006), Silicon Shortage - so what! Market survey on cell & module production 2005, Photon International (March 2006), pp 100-124. Seeking Alpha (2006). Shi, Z. (2005), Big PV Boom in China?, Photon International, August 2005, p 96. Suntech Power (2006a), Suntech Closes Acquisition of MSK Corporation, (Last Updated: 14 August 2006), Accessed from: http://suntech.wm.net.au/Investors/FinancialReleases, on: November 2006. Suntech Power (2006b), Suntech Power Receives UL Certification for Photovoltaic Modules, (Last Updated: 1 May 2006), Accessed from: http://suntech.wm.net.au/Investors/FinancialReleases, on: November 2006. Suntech Power (2006c), Suntech Power Reports First Quarter 2006 Financial Results, (Last Updated: 18 May 2006), Accessed from: http://suntech.wm.net.au/Investors/FinancialReleases/SuntechPowerReportsFirstQuarter 2006Financial/tabid/315/Default.aspx, on: November 2006. Suntech Power (2006d), Suntech Power Reports Second Quarter 2006 Financial Results, (Last Updated: 15 August 2006), Accessed from: http://suntech.wm.net.au/Investors/FinancialReleases/SuntechPowerReportsSecondQua rter2006Financia/tabid/306/Default.aspx, on: November 2006. Suntech Power (2006e), Suntech to Establish Manufacturing and R&D Facilities in Shanghai, (Last Updated: 21 August 2006), Accessed from: http://suntech.wm.net.au/Investors/FinancialReleases, on: November 2006. Suntech Power (2006f), Website, Accessed from: http://www.suntech-power.com, on: November 2006. Swanson, R.M. (2004), A Vision for Crystalline Silicon Solar Cells, Sunpower White Papers (Last Updated: June 2004), Accessed from: http://www.sunpowercorp.com/html/Resources/TP_index.html, on: November 2005. ter Horst, E. and Zhang, C. (2005), Impacts of Technology Improvement and Quality Assurance in the WB/GEF China Renewable Energy Development Project on PV industry and market development in China, 15th PVSEC. Trupke, T., Bardos, R.A., Schubert, M.C. and Warta, W. (2006), Photoluminescence imaging of silicon wafers, Applied Physics Letters, 89 (044107 ). Trupke, T., Pink, E., Bardos, R.A. and Abbott, M.D. (2007), Spatially resolved series resistance of silicon solar cells obtained from luminescence imaging, Applied Physics Letters, 90 (093506).
265 Chapter 7. Capability Building at Suntech Power
Wang, S. (2006), Current Status and Future Expectation of PV in China, China-EU Energy Collaboration Forum, Shanghai, China, Feb. 20-21, 2006. Wenham, S.R., Mai, L. and Tjahjono, B. (2005a), Transparent Conductors for Silicon Solar Cells, in Provisional Patent Specification FBR Ref 123719. Wenham, S.R., Mai, L., Tjahjono, B., Jia, J. and Shi, Z. (2005b), Innovative Emitter Design and Metal Contact for Screen-Printed Solar Cells, 15th International Photovoltaic Science and Engineering Conference (PVSEC), Shanghai, China, 10-15 October. Yang, H., Wang, H., Yu, H., Xi, J., Cui, R. and Chen, G. (2003), Status of photovoltaic industry in China, Energy Policy, 31 (8), pp 703-707. Yuan, X., Li, H. and Shi, Z. (2005), Large Area Solar Cells Module Laminating Technology Research, 15th International Photovoltaic Science and Engineering Conference (PVSEC), Shanghai, China, 10-15 October. Zhao, Y. (2001), The present status and future of photovoltaic in China, Solar Energy Materials and Solar Cells, 67 (1-4), pp 663-671. Zhao, Y., Wang, S., Li, X., Wang, W., Liu, Z. and Song, S. (2006a), Report on the Development of the Photovoltaic Industry in China (2004ዉ2005), China Renewable Energy Development Project Office. Zhao, Y., Wu, D. and Li, X. (2006b), The Status of Photovoltaic Industry and Market Development in China, Great Wall World Renewable Energy Forum Beijing, China, October 24-26, 2006. Zhaoyuan, M. and Lin, W. (2005), New Energy Storage Method For Stand-alone PV System, 15th International Photovoltaic Science and Engineering Conference (PVSEC), Shanghai, China, 10-15 October. Zhu, F., Ji, J., Wang, Y. and Shi, Z. (2005a), Characterization of Fill factor of solar cells by Measuring RS and RSH, 15th International Photovoltaic Science and Engineering Conference (PVSEC), Shanghai, China, 10-15 October. Zhu, F., Ji, J., Wang, Y. and Shi, Z. (2005b), Investigation of minority carrier lifetime for wafers passivation, 15th International Photovoltaic Science and Engineering Conference (PVSEC), Shanghai, China, 10-15 October.
266 CChhaapptteerr 88.. CCaappaabbiilliittyy BBuuiillddiinngg aatt GGrruuppoo FFéénniixx
The purpose of this chapter is analyse learning in the case of Grupo Fénix, an NGO engaged in supporting small scale module and BOS manufacturer in Nicaragua, in order to discover typical constraints for small scale PV manufacturers, how capability building occurred in this case, and the ways that the technological system supported or constrained learning. Grupo Fénix is an NGO engaged in rural community development through the promotion of renewable energy in Nicaragua. Grupo Fénix supports small scale enterprises that assemble PV modules and manufacture BOS components. The case study provides an opportunity to look at learning for both module and BOS manufacture in both urban and rural settings, and is of particular interest because of the attempts by Grupo Fénix to forge strong links with universities and communities. The case study was carried out during March-May 2003. The data was collected via: Semi-structured individual and group interviews with key staff of the Grupo Fénix organisations; Observation of the operation of the organisations, production processes, quality control processes, sources of technology and the roles of staff; Interviews with the users of the systems of Suni Solar, Fénix Norte and another PV system supplier; Product testing; and Collection of documents such as manuals, training materials, price lists, accounts and records of installations, specifications and circuit diagrams, and published papers.
The status of rural electrification, the PV industry and small scale enterprises in Nicaragua are reviewed in section 8.1. The history, philosophy and activities of Grupo Fénix are described in section 8.2. Section 8.3 details the capabilities of the Grupo Fénix enterprises, as measured by their ability to produce and sell products of an appropriate cost and quality. The framework developed earlier in this thesis is then used to examine the roles of different types of learning at Grupo Fénix and the organisational and technological system factors that have influenced the ability of the enterprise to learn is discussed in section 8.4.
267 Chapter 8. Capability Building at Grupo Fénix
8.1. Background on Nicaragua
This section provides background about the status of PV rural electrification and small enterprise support in Nicaragua.
8.1.1. Rural Electrification in Nicaragua Most of the population of Nicaragua is concentrated in urban areas along the west (Pacific) coast of the country, particularly around the capital, Managua. Until 2000, the Nicaraguan state-owned electricity company ENEL (Empresa Nacional de Electricidad) was solely responsible for providing electricity services in Nicaragua. ENEL’s primary strategy for serving rural areas was via extensions of its high-voltage transmission lines (Mathieu et al., 2001). However, the low population density, mountainous terrain and long distances involved made the line extensions very costly, and progress was slow. During the 1990s, the electricity sector was restructured, and ENEL was separated into six separate companies. Between 2000-2001, five of the companies were sold to private interests. The two distribution companies, which owned concessions over the entire area of the country covered by the grid, were sold to the same buyer. Within the concession area, the supplier is obliged to provide electricity to customers within 150m of the grid (World Bank, 2003), but there are many villages where a few houses on the main road are electrified and the rest of the village relies remains unconnected. The concession holder is disputing these obligations and the tariffs for rural customers. In addition, more than 50% of the geographical area of the country (the entire east of the country) is not covered by the concession area. CNE estimates that only 60% of the population within the concession area and 21% of the population outside the concession area have electrical energy (Comisión Nacional de Electricidad, 2000). According to ENITEL (a Nicaraguan telecommunications provider), electrical power supply in Nicaragua, particularly in rural areas, is unreliable and very unstable, while fault repair time has increased, since there are generally no maintenance technicians in rural areas (Ernberg & Arce, 2002). The area not covered by the National Interconnected Electric Grid is called the ‘open area’. This area remains open to be divided into smaller concession areas on a case by case basis. The open area is characterized by the very low density of its population. 15,584 households inhabit an area of 124, 433 km2, equal to 6 inhabitants/km2, while density in the two concessions areas reaches approximately 80 inhabitants /km2 (Mathieu et al., 2001). Access is often difficult in the open area (e.g. only by river for most of the Atlantic Region). There are therefore large areas of the country which require decentralised electrification solutions. Since the privatisation, ENEL continues to provide electricity via diesel mini grids to larger villages in
268 Chapter 8. Capability Building at Grupo Fénix
remote areas. There have also been a number of small PV projects implemented in Nicaragua in the past that have left a scattering of systems with no arrangements for ongoing maintenance or replacement of parts. These projects have tended to be demonstrations and have been heavily subsidised. There is some political will to improve the situation, as the government sees electricity as key to rural economic activities (World Bank, 2003). The Comisión Nacional de Energía (CNE) has identified off-grid solutions as the most cost-effective in rural areas. The World Bank, UNDP and GEF are supporting the implementation of rural electrification pilot projects, including photovoltaics, in areas where the population is dispersed and load density is low. This intervention is described in more detail in the following section. A National Electricity Development Fund (FODIEN) was established in 2000 to finance the rural electrification programme. This was to be funded via the privatisation of the electricity sector, but the contributions to FODIEN have been very small and are uncertain, since they are decided on a yearly basis. The World Bank (2003) reports that the fund was inoperative by 2003.
8.1.2. Markets for PV in Nicaragua A World Bank Private-Public Infrastructure Assistance Facility (PPIAF) study (Mathieu et al., 2001), carried out prior to the World Bank / GEF Nicaragua project found that only 2% of remote rural households without access to the grid had PV systems. The majority of the remote rural households surveyed (94.2%) believed that electricity is important for education. Reading, productive activities and water supply were also felt to rely on electricity by a majority of those surveyed. However, only 14% felt that a PV system was an investment priority for their family and 56.9% said that they would prefer to wait for the grid to arrive. Prior to the World Bank / GEF Nicaragua project, sales of PV systems were primarily to government or aid programmes, with some on one-off sales to relatively affluent private customers such as coffee growers or other farmers with larger holdings. In the National Plan of Rural Electrificación (PLANER), the CNE calculated a willingness to pay of the rural population of between US$5 to 8 per month (Jochem, 2005), while the cost of a 25W system was approximately US$350 in 2003. The PPIAF study therefore (Mathieu et al., 2001) found that the off-grid delivery of SHS was not viable without subsidies, since the density of demand is too low. The World Bank / GEF project calculated that subsidies of 25-50% were required to stimulate a market of sufficient density to support a network of maintenance staff in the project’s pilot regions of El Cua, El Ayote and Nueva Guinea (World Bank, 2003). A market of 8000 SHS over 5 years was seen as ambitious but achievable in the context of the subsidies. Although it is difficult to estimate the market size prior to the World Bank project, it can be observed that by far the largest retailer of SHS in Nicaragua (Technosol) has sold almost
269 Chapter 8. Capability Building at Grupo Fénix
20,000 PV modules in Nicaragua between 1998 and 2007 (Environmental Resources Trust, 2007), including sales of 203 modules in 2003 (Suni Solar, 2003). There are four other suppliers of PV systems in Nicaragua, including Suni Solar. The total market could not have been more than 500 SHS per year. This situation would be expected to continue without intervention.
Market Constraints The cost of PV systems in Nicaragua is higher than in many other developing countries (Table 8-1). The high cost is largely a result of the poor economies of scale achieved in the purchase of components and in installation and maintenance infrastructure. As a result, PV systems are out of the range of what most poor rural Nicaraguans are willing to pay. Also contributing to the high cost are custom duties on imported SHS equipment of 5% and sales tax (IGV) of an additional 15 %. While renewable energy equipment, such as SHS components, are in theory exempt from import duties, this is not reflected in current practice for equipment classification. An electricity law which had given a three year exemption period from sales tax for PV systems ended in March 2001. By contrast, diesel is exempt from sales tax.
Table 8-1: Comparison of SHS prices in Indonesia, Sri Lanka and Nicaragua Nicaragua Sri Lanka Indonesia Capacity Watts 43 42 40 Cost of goods sold (COGS) US$ 643 234 255 Solar module US$ 250 134 160 Battery US$ 130 30 25 Controller US$ 90 15 30 Fluorescent lamps US$ 120 30 20 Accessories US$ 53 25 20 Gross margin US$ 150 234 137
% 30% 50% 35% Retail price US$ 918 469 392 Indicator US$/Wp 21 11 10 1. Taxes and interest cost have been assumed to be included in the gross margin. In Sri Lanka several companies have tax exemption. 2. Gross margin includes operating expenses such as depreciation, salaries, wages, rent utilities, interest, advertising, taxes
Source: (World Bank/UNDP, 2002)
Nicaragua has a well developed microfinance industry, including NGOs, credit unions and microfinance / business development services (BDS) providers, but Suni Solar reports that there is no credit available for people wishing to buy PV systems. The World Bank verifies that Nicaraguan microfinance institutions only provide short term working capital, and identifies an unsupportive legal and regulatory framework and an undisciplined credit culture as key constraints (World Bank, 2003). At the time of the case study, Suni Solar’s competitor, Technosol, was negotiating credit through an international specialized PV fund in order to
270 Chapter 8. Capability Building at Grupo Fénix
provide 4 years financing at 14 % interest rate to its clients. Suni Solar has thus far been unsuccessful in securing credit from Banks and Credit Unions for customer financing. Low knowledge about PV, low willingness to pay in relation to the high cost of PV systems, lack of finance for PV and the lack of sales, service and maintenance infrastructure constrain the market. The wider use of renewable energy technologies for off-grid electrification is also constrained by existing fossil fuel subsidies and faulty application of import duties.
Markets and the World Bank/GEF Nicaragua Project The PV part of the World Bank/GEF Nicaragua project aims to create a local industry with sufficient density and market stability to be sustainable, through the provision of 5 years of financial support and the improvement of consumer confidence through standards development. The project is to be implemented through a cash-sales approach, with a goal of 150 kW of PV capacity, or around 8000 SHS, over a 5 year period. Users are to purchase systems from accredited dealers, who install and maintain the systems. The user is to pay 5-10% of the system costs upfront and to pay the balance in monthly instalments over three years via micro-finance organised through the project. The project provides grants and subsidies to bring the cost of systems down by 25-50%, as detailed in Table 8-2. These are to be paid to the dealers upon proof of purchase and verification of the installation.
Table 8-2: Financing of PV Systems through World Bank / GEF Nicaragua Project
PV System Size (Wp) 20 36 50 Estimated Unit Cost ($US) 425 488 600 Estimated Unit Cost (Cordobas) 6072 6981 8576 Downpayment (10% for 20W, 20% for larger systems) 304 698 858 GEF grant (53.63 Cd /Wp) 1073 901 772 Government Grant (Cd) 1970 1500 1375 Loan Balance (Cd) 2726 3882 5571 Monthly Payment for 3 years at 18% interest (Cd) 99 140 201 % of GEF and Government Subsidy to Capex 50% 34% 25% Source: (World Bank, 2003)
The criteria for dealer accreditation included demonstrated capability and previous experience in PV distribution and/or retail business, an acceptable business plan and ‘other criteria’ (World Bank, 2003). Despite the World Bank’s recognition that monopoly operation could result in the demise of existing small local dealers, the main project background research project document identifies only one PV supplier, and subsequent project documents mention only two, although there are at least five identified by Jochem (2005). In addition, the project prefers the bulk purchase of goods, with bidding packages of greater than $US150,000 preferred, excluding small producers and suppliers.
271 Chapter 8. Capability Building at Grupo Fénix
8.1.3. Small PV Enterprises in Nicaragua
Small Business Support in Nicaragua About 42% of all Nicaraguan households conduct some type of non-farm enterprise. Two thirds of these enterprises are one person operations. Three quarters are located in urban areas and the other quarter are rurally based (Coordinadora Regional de Investigaciones Económicas y Sociales, 2001). The government places importance on rural development, in particular agriculture and small and medium enterprises, because of their labour intensive nature (Government of Nicaragua, 2001). BDS provision in Nicaragua is primarily through NGOs. The service is supply driven, and biased toward urban areas. There is a tendency to focus on high- tech solutions such as e-commerce and computerising accounting, production and management (World Bank, 2003). There is very little focus on basic business training. There is a stated intention to invest in rural infrastructure, support rural production technology and implement demand driven programs for small and medium enterprises. The Inter-American Development Bank contributed a soft loan of $6,790,000 to finance technological innovation for small and medium enterprise in 2001. The government administered funding was intended to support the private sector, NGOs and universities in providing technology services, and to strengthen national research laboratories (Inter-American Development Bank, 2001). However, the government is prioritising funding towards the fishing, milk and tourism sectors ”…and developing a legal apparatus for [intellectual property rights]. Scientific research is something we may think of in the future, if we have time, energy and money.” (Velho, 2002, p 12)
Small Enterprise Support and the World Bank/GEF Nicaragua PV Project There is provision within the World Bank/GEF Nicaragua project for BDS, technical assistance (including installation and maintenance training), market studies and assistance with promotion to the dealers. However, only accredited dealers qualify for involvement in the program.
Human Resources for PV Enterprises While there are many unemployed university educated engineers available in Nicaragua, there is no specific engineering training in photovoltaics, no engineers experienced in the design or manufacture of BOS or PV modules, and no vocational training available for PV installation and maintenance technicians. There is therefore likely to be a shortage of skills and knowledge related to all parts of the PV value chain. Similarly to many developing countries, the Nicaraguan university system suffers from brain-drain, lack of resources and a low focus on research, reducing the quality of the education. In 1990, there were 25 000 Nicaraguans with university degrees who had officially emigrated to
272 Chapter 8. Capability Building at Grupo Fénix
the United States. Nicaraguan university pay is very low, and lecturers are civil servants, promoted on the basis of seniority, rather than research activities. 75% of the teaching staff are full-time teachers (Velho, 2002). Academics in Nicaraguan Public universities also lack electronic libraries (Ernberg & Arce, 2002).
273 Chapter 8. Capability Building at Grupo Fénix
8.2. Background to Grupo Fénix
This section describes the history, organisation and activities of Grupo Fénix.
8.2.1. History and Philosophy of Grupo Fénix Susan Kinne, the founding director, started Grupo Fénix (Kinne & Komp, 2001), a Nicaraguan NGO in 1995, with a group of Electrical and Industrial Engineering students at the National University of Engineering (UNI) in Managua, the capital city of Nicaragua. The initial objective of Grupo Fénix was the creation of a renewable energy industry in Nicaragua. Students were motivated by the belief that locally made products would provide jobs locally and be more affordable, hence proliferating throughout Nicaragua. The stated objectives of Grupo Fénix are (Grupo Fénix, 2007): To conduct practical research into appropriate energy technologies, To support community self-determination and local responsibility for each project, To increase the technical skill level and employability of local people, To improve the health and living standards of families and communities, To preserve natural resources.
The UNI students decided to manufacture PV system components in Nicaragua. Since Susan Kinne had previously worked for Cincinnati Milacron in the US, and had seen the reject doped silicon wafers thrown away, the group hoped that reject solar cells may be accessible. Searching for more technical expertise for her students, Susan contacted Richard Komp in Maine, USA (a proponent of solar energy since the 1960s) for technical advice, since she had a copy of his book “Practical Photovoltaics”, written in 1981, and a photovoltaic battery charger made by his company (Sunwatt). Richard Komp agreed to come to Nicaragua to give workshops and since then has become a director of, donor to and regular technical consultant for Grupo Fénix, making annual visits. Suni Solar S.A. (Suni), a small business, was formed by some of the UNI students and Grupo Fénix to manufacture photovoltaic modules. Suni began operations in a house in the barrio Edgar Munguia, a poor barrio in Managua. Legal status was obtained in 1998, in order to facilitate legal sales and apply for an importers license. Suni is engaged in assembling modules at a cottage industry scale and manufacturing in-house designed 12V DC lamp inverters. Suni also imports and retails charge controllers, PV modules, deep-cycle batteries, and designs and installs systems. The engineers employed by the enterprise also repair electronic appliances in order to bring in extra income.
274 Chapter 8. Capability Building at Grupo Fénix
Through Richard Komp, the Falls Brook Center (FBC) in New Brunswick, Canada became aware of Grupo Fénix, and decided to incorporate renewable energy into a program to help survivors of landmines from the Contra War in northern Nicaragua during the 1980s. FBC received a grant from the Canadian government for this purpose. The grant of C$80,000 ($US5500) funded a program, ‘Light and Limbs’, through which a group of landmine survivors were trained to install PV systems and make PV modules during 1999, for which Grupo Fénix provided technical input. The project was designed to give skills and employment to landmine survivors, who were also provided with prosthetics. Based on the success of this program, the grant was renewed for C$50,000 ($US3500). Most of this work took place in the mountainous Department of Madriz bordering Honduras, four hours north of Managua. Around 12 landmine victims were trained in cutting and soldering cells, assembling panels and the operation of solar cells. The first training courses were in Managua and were taught by Suni Solar. Two courses on system installation were also delivered by Suni Solar employees. Participants received a certificate for PV system installation. As a result of the Light and Limbs project, a second manufacturing enterprise (Taller1 Sabana Grande) supported by Grupo Fénix is now operating in Sabana Grande, a tiny rural settlement near Totogalpa in the Madriz department in the north of Nicaragua (see Figure 8-1). This enterprise is not incorporated as a business. Two landmine survivors are employed making PV modules, carrying out R&D into improved methods of PV module manufacture, and small LED lighting devices, such as lights for outdoor toilets and bicycles. Since 2002, Taller Sabana Grande has been financially independent of Suni Solar.
Figure 8-1: Map of Nicaragua showing Managua and the Totogalpa Municipality
Figure has been removed due to copyright restrictions.
1 Workshop in Spanish 275 Chapter 8. Capability Building at Grupo Fénix
Both Suni Solar and Taller Sabana Grande have maintained their relationships with Grupo Fénix. Grupo Fénix continues to maintain its central office and presence at the UNI, has worked consistently to promote renewable energy in the rural villages of Unile and Sabana Grande in Madriz Department, and has also installed solar energy systems for various applications in the Esteli, Rio San Juan, Ometepe and Jinotega Departments. The primary aim of Grupo Fénix’s work is to combine the technical and organizational expertise of university- trained people with the needs, skills and interests of local people in Nicaragua's low-income communities.
8.2.2. Grupo Fénix Activities Grupo Fénix has expanded into a small cooperative network of university-trained engineers and several local communities developing and implementing renewable energy technologies and finding cost-effective ways to implement these technologies. The organisations and activities which make up Grupo Fénix include: Proyecto de Fuentes Alternas de Energia (PFAE, Alternative Energy Sources Project). Founded in 1998, this is the group that was the genesis of Grupo Fénix within the National Engineering University (UNI). PFAE continues to organise seminars, workshops, short courses and annual renewable energy fairs, and has established working relationships with photovoltaics specialists and universities internationally. Suni-Solar, S.A. the small business that designs, produces and sells renewable energy products and services. The Asociación Fénix or AsoFénix, has Nicaraguan non-profit status since August, 2001, and allows Grupo Fénix to receive grants and carry out social and community development projects. Centro de Investigación, Promoción y Producción con Energía Renovable (CIPPER). The Centre for Research, Promotion and Production of Renewable Energy was established to facilitate technology interactions between the UNI and rural communities. The first CIPPER rural centre carrying out research and dissemination of Renewable Energy Technologies is in Sabana Grande, and has absorbed the PV manufacturing enterprise in Sabana Grande. Mujeres Solares (Women’s Solar Groups). Grupo Fénix works with community groups in the Madriz region to promote a solar culture. Solar groups have been formed in the villages of Unile and Totogalpa, within which women share knowledge and experiences of using solar cookers and crop dryers. A workshop has been built in Unile, also in the Madriz department, where the group builds solar cookers and carries out research and testing on new types of solar cookers using local materials. A demonstration photovoltaic system has also been installed at the workshop. Members
276 Chapter 8. Capability Building at Grupo Fénix
of the solar cooker groups in Madriz have participated in ‘solar fairs’ in Managua, where they disseminate information about the technology and its use in the rural communities. Grupo Fénix organises an annual solar culture course, where people from industrialised countries travel to the Madriz region in Nicaragua to learn about photovoltaics and solar thermal technologies and their applications in remote communities. Income is brought to the families in the communities who provide food and accommodation to the visitors. Taller Sabana Grande teaches the visitors how to make a small PV panel and battery charger, and the visitors build a solar cooker with the women’s solar cooker groups.
Figure 8-2: Unile Women’s Solar Cooker Group
8.2.3. Group Fénix Staff Suni Solar employs three people in the manufacturing side of the enterprise. Two electronic engineers educated in Managua designed the circuits for fluorescent lamp ballasts for their electrical engineering masters thesis at the UNI and manufacture lamps, photovoltaic modules and carry out R&D into product improvements. An uneducated single mother who lives in the barrio does much of the assembly of the modules and cleans the premises. Another woman who lives in the house that Suni Solar uses as its premises is also paid a small wage for answering phones and tending to customers when the office is closed. A qualified accountant is also employed on an ad hoc basis. Richard Komp, who is a Grupo Fénix board member and director and shareholder of Suni Solar, continues to make annual three monthly visits to provide technical assistance to Grupo Fénix.
277 Chapter 8. Capability Building at Grupo Fénix
The two workers in Taller Sabana Grande were participants in the Falls Brook Center “Light and Limbs” project in 1999. One of the workers was previously trained as a residential electrical technician, and has worked in television repairs. His eyesight was severely damaged by a land mine, which impairs his ability to perform tasks requiring a high degree of hand-eye dexterity. He has, however, adapted to his circumstances and is able to solder the interconnect tape to the busbars of the solar panels with accuracy. He also has a prosthetic leg, which hampers his ability to walk or stand comfortably. The second worker had no related training prior to involvement with Grupo Fénix. He is mainly engaged in cutting and soldering cells in preparation for encapsulation. He is also involved in maintenance on systems installed by Grupo Fénix. He has reduced mobility due to shrapnel damage to his leg, but still walks long distances to visit remote systems. There are also a handful of installation and maintenance technicians who are employed full or part time through Grupo Fénix and its projects, including some people with university engineering degrees.
278 Chapter 8. Capability Building at Grupo Fénix
8.3. Capabilities at Grupo Fénix
This section details the development of capabilities at Grupo Fénix. The following section then uses the framework developed in chapter 5 to analyse the learning processes that have enabled Grupo Fénix to build these capabilities and the factors that have facilitated or constrained these learning processes.
8.3.1. Production Capabilities and Activities at Grupo Fénix Starting with a basic technique for module manufacture provided by Richard Komp, of Maine, USA, as described in his book ‘Practical Photovoltaics’ (Komp, 1995), Grupo Fénix has built the capability to produce PV modules using a new ‘cottage industry’ technique. Suni Solar also have the capability to manufacture 15W (1.3A), 20W (1.6A), and 32W (2.6A) 12V DC lamps, based on the circuit the engineers designed for their electronic engineering masters thesis topic at the UNI. The conventional technology available for module assembly using commercial techniques is capital intensive and not appropriate for the resources of small scale enterprises. Commercial photovoltaic module assembly typically involves testing and sorting the cells into similar lots, soldering the cells into strings, forming a stack of (typically) glass-EVA2-cells- EVA-Tedlar™ and curing the EVA encapsulant by placing the stack in a vacuum lamination machine that carefully controls the temperature and pressure profile. Grupo Fénix lacks the financial resources to invest in the plant required to manufacture PV modules using the methods employed in commercial module manufacture. Grupo Fénix uses expensive RTV silicone as an encapsulant, instead of EVA, which is widely used commercially. EVA has good thermal and optical properties for use in silicon photovoltaics, is stable under UV radiation and durable in a wide range of climatic conditions. EVA, however, requires a commercial laminating machine that can produce the correct vacuum and temperature profile. Quality control issues, as well as high cost of materials, result from the use of RTV silicone. Although RTV silicone has demonstrated good UV-stability and durability in the field (Dross et al.), it is difficult to remove air bubbles formed during mixing and pouring by hand. The presence of air bubbles in the finished module increases the likelihood of corrosion when the air bubbles expand and contract due to thermal cycling, and may reflect light away from the cells. The encapsulation sequence used by Grupo Fénix in 2003 was as follows (Figure 8-3): A shallow basin is made by framing a sheet of glass with aluminium window framing. RTV
2 Ethyl-Vinyl-Acetate, a polymer specifically developed for solar cell encapsulation. 279 Chapter 8. Capability Building at Grupo Fénix
silicone3 encapsulant material is mixed and then poured into the basin. Pre-soldered strings of cells are then placed on the first layer of encapsulant, followed by more silicone and a backsheet. Standard window glass is used, since the special low-iron, tempered glass that is used commercially is not available in Nicaragua and even tempered glass is very expensive and difficult to obtain. This method is described in more detail in Appendix 5.
Figure 8-3: Grupo Fénix Module Assembly Sequence
Cost and Price of Grupo Fénix Products Table 8-3 compares the costs of Grupo Fénix’s manufacturing with typical commercial costs, illustrating the difficulty in achieving prices that are competitive with imports. A breakdown of the costs and sources of materials is provided in Appendix 6. The RTV silicone and the interconnect tape used by Grupo Fénix are currently donated. The solar cells are sourced through Sunwatt (a not for profit organisation affiliated with Richard Komp’s US solar business), which donates the cells upfront and then requires $2/Wp (US$100 for a 50 Wp module) when a module is sold.
The cost in wages and materials to Suni of manufacturing a 50 Wp module is calculated from Table 8-3 at approximately US$204. If Grupo Fénix was to pay for all the materials it uses, including the donated encapsulant, the total price for the module materials and wages would be approximately US$229 for a 50 Wp module. These figures do not include the operating expenses such as depreciation of equipment or rent. In contrast, a module commercially manufactured in the US, including operating expenses typically cost US$122.50 to produce in 2000 (Frantzis et al., 2000). Transport, marketing expenses and after sales service are not included in any of the figures.
3 Room Temperature Vulcanising silicone, a two-part liquid encapsulant sold as Dow Sylgard 184 280 Chapter 8. Capability Building at Grupo Fénix
Table 8-3: Costs of Module Manufacture – Conventional Commercial and Grupo Fénix Small-Scale
Cells Encapsulant Other Wages Total (US$/Wp) material Materials (US$/Wp) Costs (US$/Wp) (US$/Wp) (US$/Wp) Commercial 1.46a 0.05b 2.45c Grupo Fénix 2.00d 0.50e $0.50f $1.58g 4.58h a Based on figures given for Cz crystalline modules in Frantzis et al. (2000) b Estimated from EVA purchase price by BP Solar, Australia of US$2.20 per metre in 2002 c Based on figures given for Cz modules in Frantzis et al. (2000). Transport to Nicaragua would add significantly to this figure. d Grupo Fénix pays $2/Wp to Sunwatt (who provides the cells upfront) when it sells a module. e The retail price of Sylgard 184 in the US is greater than US$60/l. Dross (2006) reports that Grupo Fénix are able to produce two 60Wp modules per litre of RTV silicone, or US0.50/Wp. f Calculated from costings given by Grupo Fénix as detailed in Appendix 6. g If Suni Solar could operate at its estimated full capacity of 1 X 50Wp module/ 2.5 days, given monthly wages of US$954 for two engineers, one full time manufacturing labourer, two part-time installation technicians and a part-time secretary and accountant (Suni Solar, 2003). h Excluding depreciation of equipment, rent and marketing expenses
Suni Solar were selling a 50 Wp module in 2003 at US$300. Given VAT of 15%, the net amount going to Suni from the sale would be $261. The profit (without subtracting marketing expenses and after sales service) would therefore be $32 (11% of the sale price) if the encapsulant was paid for. The manufacture of the modules would be marginal, but feasible, if the enterprise was operating at full capacity. Sales, however, are limited, since the cottage industry modules are being offered at a similar price to imported products, which are perceived by consumers to be higher quality. For example, one of Suni’s competitors was selling a 50 Wp Astropower module at US$305 in 2003 (Table 8-4). Suni is unable to reduce costs to expand market share, and has not yet been able to make a profit, despite the avoidance of transport costs for the local product.
Table 8-4: Advertised Prices of Modules from Suni Solar and Altertec 2003 ALTERTEC SUNI SOLAR S.A. Module Size Brand Price (US$) Brand Price ($US) 25 W Suni Solar $180 35 W - Suni Solar $225 50 W Astropower $305 Suni Solar $300 75 W Astropower $415 Suni Solar $440 120 W Astropower $660 Suni Solar -
The lamps manufactured by Suni require printed circuit boards, transformer and inductor coil wire, a pot core and a range of electronic components such as transistors, diodes, resistors and capacitors, as well as the lamp housing and reflector. The manufacturing of the lamps involves winding transformers and inductors, soldering the components onto the printed circuit board, and wiring the ballast into the lamp housing. The electronic components for a 10 W lamp cost Suni approximately $8, and the lamp housing and reflectors cost around $2.50. One lamp takes about half a person-day to produce (approximately $US7 in wages). The total cost of materials and wages is around $17.50. The lamps are sold for between US$23-30. A 10
281 Chapter 8. Capability Building at Grupo Fénix
W lamp is sold for US$25, leaving a profit of US$3.70 after VAT, but excluding marketing expenses and after sales service (17% of the sale price). The lamps sold by Suni are slightly cheaper than the US$25-40 offered by competitors for imported lamps (Foster & Gómez, 2005; World Bank, 2003), and are in more demand than the domestically made modules, perhaps because it is not such a significant purchase, and therefore less risky for both NGOs and individual customers.
Access to Materials for Production It has been established that one of the most important obstacles to Grupo Fénix manufacturing robust modules cost effectively is its inability to source materials at a reasonable cost. Grupo Fénix is unable to purchase cells at a competitive price in small quantities, even though it is buying reject cells. The high cost of the RTV silicone is a further obstacle to competitiveness, although thus far, the RTV silicone has been paid for out of landmine rehabilitation project funds, or from sources such as the Grupo Fénix solar culture course attended by visitors from the United States. The remainder of the materials for module manufacture (detailed in Appendix 6), with the exception of the glass and the backing sheet for the modules, are purchased mainly from hardware stores, which are stocked with poor quality, expensive products that are virtually all imported, since Nicaragua has very little manufacturing industry. The small size of Suni’s purchases of materials is not sufficient to motivate local suppliers to consistently stock the materials being used. The aluminium frame and the back sheet material used for module production have been changed many times because the material has no longer been available. Some of the electronic components for the lamp circuits produced by Suni Solar are available in Nicaragua, but others must be purchased via the internet from the US (Table 8-5). Charge controllers, deep cycle batteries and inverters which are sold as part of solar home system packages are purchased from Suni Solar’s competitors in Nicaragua who are importers and system integrators.
Table 8-5: Providers of Materials for Manufacture of Lamps
Name of Provider Materials Provided SINSA, hardware store, Managua Electrical parts Tecnirepuestos, Managua Transistors, capacitors, diodes CECA Electronics Adolfo Grover Electronics www.Jameco.com Electronics www.digikey.com www.allelectronic.com
Importation from the US has proved particularly problematic for Suni. Although Suni have a license to operate as an importer, its supplies of RTV silicone and reject solar cells are
282 Chapter 8. Capability Building at Grupo Fénix
delivered via another NGO’s intermittent loads of goods, such as second hand clothes, from the U.S. Concern over corruption in the customs operations (the possibility of paying bribes or losing the goods) was cited as the reason for not importing, but communication difficulties appear to be a primary reason. Suni uses the internet at telecentres nearby in Managua, but the speed of the internet connection makes it virtually unusable most of the time. Telecommunications are also extremely high cost, due to a monopoly market situation. Poor telecommunications infrastructure, combined with the language barrier, prevent Suni from gaining access to reliable supplies from international suppliers. Importation is also expensive. For example, tempered glass is available from San Salvador (El Salvador), but the transportation is prohibitively expensive for small quantities. Where international purchase of electronics components are made, purchase via the internet has been found to be more cost effective.
Quality of Grupo Fénix Products Grupo Fénix modules have been found to underperform in relation to their rated value.
IV curves were taken from 10 modules with ‘nominal’ ratings ranging between 25 and 75 Wp manufactured by Grupo Fénix organisations between 1999 and 2003, some directly after manufacture and some after extended periods in the field. The fill factors averaged 0.43 (see Table 8-6), whereas a Kyocera manufactured module measured for comparison had a fill factor of 0.72, which is typical of quality commercially manufactured modules.
Table 8-6: Fill Factors of Modules Manufactured by Grupo Fénix System Location Manufacturer Fill Factor Apatule School Taller Sabana Grande 0.29 Casa Base at Santo Domingo, El Jobo Suni Solar 0.50 Church Las Pintatas Taller Sabana Grande 0.33 Tamarindo 0.49 Cacao Taller Sabana Grande 0.47 Las Lomas Suni Solar or Taller 0.54 Sabana Grande Leon Suni Solar 0.41 Leon Suni Solar 0.46 Leon Suni Solar 0.38 Leon Suni Solar 0.41 Average 0.43 For Comparison: Kyocera 0.72
Bubbles in the encapsulant, and the use of backsheet materials that have not been subject to long term or accelerated testing, put the product at risk of corrosion, delamination and further performance degradation. Some modules were observed with corrosion, particularly around the solder joints (Figure 8-4), with soldering only contacting a small part of the busbar, and with cells so close together that the interconnect tape could potentially short circuit the back and front of the cells.
283 Chapter 8. Capability Building at Grupo Fénix
Figure 8-4: Corrosion in Solder Joints
The quality of the Suni Solar lamps is also inadequate. It was found that the lamp circuit drew current when the switch was on, even when the tube was faulty and did not illuminate, thus flattening the battery in the system. Blackened tubes and partially illuminated tubes were also observed (Figure 8-5).
Figure 8-5: Tube Partially Illuminated by Suni Solar Ballast
An example of the Suni Solar designed and manufactured lamp circuit was sent to Sandia National Laboratories in the US for testing and certification in 2001, but did not pass the testing. Suni Solar believes that there was some anomalous problem with the particular circuit sent to the U.S. What is clear is that the certification procedure does not support manufacturers in improving its product in order to meet standards.
284 Chapter 8. Capability Building at Grupo Fénix
Quality Control Routines There is inadequate monitoring of module quality in the Grupo Fénix production process. Prior to encapsulation into modules, Grupo Fénix reads the current through and voltage across completed strings of cells in the sun with a multimeter, but in the absence of IV testing of the cells, it seems quite likely that mismatch is occurring in the modules produced.
Figure 8-6: String testing
Since the solar cells used are usually reject cells from large scale manufacturing, they are likely to have poor electrical characteristics and are not sorted into batches or labelled according to their characteristics when delivered to Grupo Fénix. Grupo Fénix do not have the means to effectively sort the cells prior to encapsulation, because the resistive losses in a standard multimeter and leads preclude accurate measurement of the short circuit current, as described in Appendix 5. At the time the case study was conducted, IV curves of modules were also not being measured by Grupo Fénix. The ratings of the modules produced were instead estimated on the assumption that that the efficiency of the cells were the same as the cells in imported modules. The performance of the modules is therefore being overestimated and modules are being sold with a rating that does not reflect their true performance at the operating voltage. While the probable source of bad module fill factors is the poor performance of reject solar cells, the soldering technique or the cutting techniques could also result in high parasitic resistances. Failure to measure the IV curves of cells means that Grupo Fénix cannot determine the source of poor fill factors and improve their manufacturing process, even if it was known that the fill factors were poor.
285 Chapter 8. Capability Building at Grupo Fénix
Nevertheless, quality control has improved over time. Suni is now testing the IV curves of modules, and recognises the importance of cell testing after it was discovered that a number of modules did not produce as much power as they should have. It was subsequently found that they had very poor IV curves. It was necessary to replace the panels with imported ones, at great financial loss to the enterprise as well as detrimental impacts on its reputation. Although it now measures module IV curves, Suni still does not have the capacity to test the IV curves of cells.
8.3.2. Innovative Capabilities and Activities at Grupo Fénix Grupo Fénix has developed a new technique for the manufacture of PV modules at a cottage industry scale, without the use of commercial lamination equipment and has designed a DC fluorescent lamp circuit. It has also improved upon the module manufacturing techniques, adopted new materials and developed equipment for manufacturing. In response to the market, Suni produces smaller panels, and Taller Sabana Grande is experimenting with LED lighting products.
Design of the Lamp Circuit While Suni has designed functioning lamp ballast, it was not able to achieve quality certification for the lamp. The design did not adequately deal with the issues of crest factor, waveform symmetry, voltage and frequency regulation, which have found to be typical causes of tube blackening in lamps in the literature review in chapter 4. Successful designs for ballast circuits are also freely available in the literature, for example in (Vervaart & Nieuwenhout, 2000). Since the literature is primarily in English, however, Suni is not able to take advantage of the existing knowledge. After failing the certification testing, Suni has not been able to improve the circuit for retesting.
Improvement of the Module Encapsulation Process Experience has also led Grupo Fénix to prefer an encapsulation method that allows the module to be disassembled and reassembled in case of module breakage or damage, saving the valuable cells. This is a promising initiative for future recycling of module components.
Adoption of New Materials and Equipment The first modules were made with plumber’s silicone as the encapsulant, which was subsequently found to cause corrosion of the solder joints, so the RTV silicone was then subsequently imported from the US. Wire interconnects were initially used to join the cells in the modules. It was found to be very difficult to get a strong solder join with wire, so Suni had the wire rolled by a jeweller, since interconnect tape was not available. The wire formed a copper oxide over time, so eventually, Suni needed to source the interconnect tape via Richard Komp in the US. 286 Chapter 8. Capability Building at Grupo Fénix
Mylar ™ back sheet material for modules was initially brought from the US by visitors, but Grupo Fénix has experimented with other materials for the back sheet and found some suitable materials that are available locally. The first one was a flexible white plastic used for plastic raincoats which sticks to the RTV silicone where most others don’t. However, the supplier stopped stocking it, so Grupo Fénix tested more materials and switched to a white plastic with a nylon weave in the centre that is used to make advertising banners. Some equipment has also been adapted, such as the cell cutting tool used by Grupo Fénix, which is a diamond-tipped disc attached to a Dremel™ hand tool (purchased from the U.S. by Richard Komp) that has been attached to a hinge on a cutting board that guides cells so that a straight cut can be achieved.
Collaborations with Research Organisations Grupo Fénix has links with the national engineering university (UNI) in Managua, and often collaborates on solar projects with local and overseas research students, as well as independent volunteers. Students from the local UNI have carried out solar energy research projects in conjunction with Group Fénix. Students from the University of New South Wales (UNSW), the University of Washington, and the University of Dayton have also carried out projects on improving the quality control of solar panel and ballast production, low wattage LED light production, and business development. The author was involved in one such UNSW student collaboration, which resulted in some changes to the manufacturing and testing processes. The collaboration took place during 2002-2004. Research areas where Grupo Fénix wanted some technical input had been identified during a prior visit by the author and a UNSW professor to Grupo Fénix in Nicaragua in October-November 2001. Subsequently, students were engaged in investigating: Cheaper encapsulation materials, Finding suppliers willing to supply small quantities of specialist materials such as interconnect tape, Development of a circuit for testing solar cells prior to encapsulation, Identifying possible causes of low performance such as soldering and cell-cutting techniques and hence improved production methods.
During visits in February-March 2003 and February-March 2004, students and Grupo Fénix staff discussed and tried out new ideas for module manufacture, and new cell cutting methods were adopted, stress loops were included in the module strings and soldering the full length of the cell was adopted. Efforts to transfer an IV cell testing circuit, however, were not successful. It is suspected that the collaboration did not engender full buy-in of the staff. The problems perceived to be the most important to UNSW were not priorities for Grupo Fénix.
287 Chapter 8. Capability Building at Grupo Fénix
UNSW students also suggested that the interconnect tape may be used more effectively and cheaply without using additional solder, since it is already ‘tinned’. The soldering procedure would be simplified without the use of additional solder and it would be easier to obtain flat strings of cells that could be more easily encapsulated with less silicone. However, a high Watt soldering iron (greater than 80 W) with a large flat tip would be required. The suggestion has not been implemented, because of the lack of an appropriate soldering iron, and probably also the inability to see the value in the improvement. In 2005-6, Frederic Dross, from a research organisation (IMEC) in Belgium worked with Grupo Fénix on an improved encapsulation technique that used less RTV silicone and resulted in the development of new techniques that have reduced the use of expensive RTV silicone. Dross et al. (2006) report that the amount of silicone required has been reduced by half. The research collaborations between Grupo Fénix and domestic and international universities and PV industry people have allowed Suni Solar and Taller Sabana Grande to access significant amounts of knowledge and technology and to gain access to materials that they would not have been able to import independently. However, the short term and insufficiently participatory nature of some of these interactions has perhaps limited their effectiveness.
8.3.3. Investment and Linkage Capabilities at Grupo Fénix At the time of the case study in 2003, Suni Solar had carried out no formal feasibility studies or market research and did not have a formally developed business strategy. However, through the Grupo Fénix – UNI link, Suni Solar was at the time of the case study receiving assistance to develop a business plan from a master’s student in business studies and some volunteers from the US.
Investment Capabilities Suni Solar’s employees are engineering graduates with little business management experience. While the accounting records were kept up to date, despite running at a net loss, there were little records of inventory, strategic decisions were made on an ad hoc basis and no decisive changes were made to the business approach. Lack of financial capital constrained the investment in premises and equipment. The Suni Solar workshop in Managua is within a house in a poor barrio in the city, not easily visible to customers and not in good proximity to suppliers and commercial areas in Managua. The equipment used in the workshops is very basic. The most expensive and complex pieces of equipment are the oscilloscopes, which were donated by a Swedish group. Aside from the
288 Chapter 8. Capability Building at Grupo Fénix
oscilloscopes, multimeters and small hand tools are used for the manufacturing. Lack of financial capital also precluded the employment of consultants or experienced staff to give technical assistance or business assistance. At the time of the case study, Suni Solar was seeking investment funds and grants in order to secure better premises, carry out marketing activities, invest in production equipment and apply for product certification. To this end, Suni has unsuccessfully approached Banks and Credit unions. To be approved for a loan, Suni must demonstrate a good business model, a proven track record and assets as security. Taller Sabana Grande is a cottage industry workshop, located in the house of Marco- Antonio, one of the landmine victims who makes the panels. The enterprise sells primarily to local customers, with little marketing effort and employs only two people, making a small number of products with a limited product line, There is therefore limited capacity in or requirement for accounting, inventory or management.
Training Richard Komp (a US physicist, solar entrepreneur and member of the Grupo Fénix board) initially trained the staff of Suni Solar in the basics of solar cell and module operation and a basic technique for module manufacture. He now visits on a yearly basis and provides technical assistance, but this is primarily related to innovative activities, such as the development of new products or techniques. The link between Taller Sabana Grande and Suni Solar through Grupo Fénix has enabled the rural workshop to access materials and technology to manufacture modules and install and maintain PV systems. However, due to the small number of staff and the collaborative nature of the activities, formal internal training is not required to transfer knowledge within either Suni Solar or Taller Sabana Grande. Internal training at Grupo Fénix has therefore been limited to that initially provided to Taller Sabana Grande by the staff of Suni Solar.
Market Linkage Capabilities Due to depressed economic conditions in rural Nicaragua and low levels of awareness of photovoltaics, Grupo Fénix have worked to cultivate a "solar culture" in order to create a market for solar products and insure the long-term viability and success of the project. Both Suni Solar and Taller Sabana Grande primarily connect with potential customers through Grupo Fénix in the villages that it is active. Grupo Fénix, however, has little knowledge of or connectivity with most of the rural population of Nicaragua, and is active only in a handful of villages. Most of the people in these villages have a low awareness and knowledge of PV technology and do not have the capacity to pay for a PV system. The majority of the systems
289 Chapter 8. Capability Building at Grupo Fénix
have therefore been installed on community buildings and funded through donations or NGO projects.
Figure 8-7: A Grupo Fénix PV System on a Rural Health Clinic
The close links between Suni Solar and Grupo Fénix have allowed some feedback on modules and components from users to flow back to Suni Solar, providing a basis for it to improve its products, and to tailor the products it sells to the market requirements. Low levels of communications and transport infrastructure increase transaction costs in rural areas and the risks to customers associated with the purchase of PV systems. Suni Solar is urban based, and only has one agent selling its products in a rural region (in Cua Bocay, in the Jinotega region). It is therefore difficult for Suni to promote their products or to provide after sales service and the maintenance of systems. Suni’s marketing activities include an advertisement in the Yellow Pages, occasional radio announcements and frequent participation in rural fairs, some of which are sales tax exempt. Taller Sabana Grande also occasionally participates in fairs and has placed advertisements in the newspaper. While it is rurally based, it has poor connectivity with towns outside the department of Madriz. Despite marketing efforts, Grupo Fénix primarily relies on word of mouth to make cash sales and reports that cash purchase customers have generally already seen a working system. Customers are otherwise generally unwilling to take the risk.
290 Chapter 8. Capability Building at Grupo Fénix
Lack of available credit is a further barrier to market access. Some customers have asked Suni to provide credit, or payment by instalments, but it is unable to bear the risk or provide capital for manufacturing panels upfront. Suni’s small size and track record are barriers to accessing micro-credit agreements through financial institutions. One financial Institution (Asociación de Consultores para el Desarrollo de la Pequeña, Mediana y Microempresa (ACODEP), an NGO engaged in the development of the micro and SME sectors in Nicaragua through the provision of loans, management training and business consultancy and public advocacy) has expressed interest in providing customer credit for productive uses of PV, such as refrigeration systems for tiendas (small shops) and water pumping for vegetable irrigation. ACODEP requires Suni to provide a good business plan and evidence of the benefits of PV in community development, which was being undertaken at the time of the case study. Taller Sabana Grande is aiming to become involved in co-operatives which bulk buy products and then sell to members on credit. While Suni and Taller Sabana Grande have been able to access markets through NGO projects that have involved Grupo Fénix, and have been engaged by a few other NGOs who have wanted to use locally manufactured products, they do not have a prominent reputation, official accreditation or overseas contacts and are therefore usually overlooked by international development projects, including the large World Bank / GEF project described in section 8.1.2. Grupo Fénix was not accredited by the World Bank project and its market is likely to be consequently decreased as potential customers purchase subsidised systems through the project. Its competitors Technosol and ECAMI get most of the big projects because they are large, have the reputation and contacts and sell imported, certified products.
Sales Suni has the capacity to assemble up to one 35/50W module every 2-3 days, but can only afford to make modules in accordance with demand. It only sells about 10 of the locally manufactured PV modules per year, because it has only a small number of customers, most of whom prefer to buy imported modules, which are perceived to be better quality and are only marginally more expensive. In addition to the sales made through Grupo Fénix projects, Suni Solar has made some cash sales of PV SHS to more wealthy coffee growers and cattle farmers, usually in the Madriz region, where Grupo Fénix has a presence. These clients either purchase their systems in person from the office in Managua or over the telephone. Taller Sabana Grande has also made a few cash sales to households in their local area. Suni Solar makes regular sales of their lamps to the system integrators and retailers that are their competitors, including Technosol, Altertec and Seeizmo.
291 Chapter 8. Capability Building at Grupo Fénix
Figure 8-8 indicates the fluctuation in Suni Solar sales (including sales of imported components). After covering their expenditure on materials, wages and rent, they have not made a profit thus far.
Figure 8-8: Suni Solar Income 2001-2002
8000
7000
6000
5000
4000
3000
2000 Average = $2141
1000 Monthly Income $US 0 Jun-01 Jul-01 Aug-01 Sep-01 Oct-01 Nov-01 Dec-01 Jan-02 Feb-02 Mar-02 Apr-02 May-02 Jun-02
After Sales Service Suni Solar officially offers warranties of 20 years on panels, 1 year on controllers, 1 year on batteries, 1 year on lights and 3 months on wiring after installation and testing. However, despite relatively good connectivity with communities, poor institutional arrangements have still resulted in poor system maintenance. Despite links between Suni and Grupo Fénix, the Grupo Fénix technicians do not have the ability to repair the BOS in the field. Some systems or system components are sold by Suni purely for cash, with no installation or service. Where installation is included, clients are informed about the use and maintenance of their system, but there is usually no monitoring or organised maintenance of cash-sales systems, since the users are often remote from Managua or any technicians. Suni only has one local technician in a rural department, Cúa Bocay, who buys the systems from Suni and takes responsibility from there. In contrast, Technosol (a larger competitor), has a number of regional technicians who install and maintain systems and hold stocks of spares. When systems without maintenance agreements fail under warranty, customers may receive telephone assistance or may bring components to Managua for repair. Both Suni and Taller Sabana Grande are very proud of their products and their work, and have accepted components for repair outside of warranty periods and have sent technicians to repair systems where there was no maintenance agreement. On community buildings, there has been a lack of responsibility taken for the systems and lack of financial provision for maintenance. In some
292 Chapter 8. Capability Building at Grupo Fénix
cases, Group Fénix has repaired systems free of charge, but in others, they are effectively abandoned. Grupo Fénix technicians have absorbed responsibility for other photovoltaic systems installed in the regions where it has a presence. Despite such efforts in the community, the small size of the operation, and the dispersed nature of the systems have made it difficult to implement an effective organisational structure for maintenance.
Figure 8-9: Grupo Fénix Technician Testing a System at a Rural Health Clinic
293 Chapter 8. Capability Building at Grupo Fénix
8.4. Analysis of the Case Study Using the Framework
The preceding sections of this chapter have described the development of capabilities by Grupo Fénix since Suni Solar began producing PV modules in 1998. The framework developed in chapter 5 will now be used to identify the factors which have supported or constrained different types of learning and to analyse the success of the strategies that have been used to build capabilities and overcome deficiencies in the technological system.
8.4.1. The Nicaraguan Technological System for Small Scale PV Manufacture Figure 8-10 shows the technological system (a) prior to the World Bank / GEF Nicaragua project (abbreviated to WB/GEF project for this discussion), and (b) after the project had commenced in 2003. In each case, functions where the technological system was effective in supporting enterprise-level learning are represented by red highlighting. Blue is used to indicate areas where the technological system has partially supported learning, while functions left un-highlighted have not been fulfilled by the system. In the following sections, each of the functions of networks is discussed, and the ways that the institutional environment has influenced the operation of networks are identified.
Investment Opportunities The market for PV systems was very small in Nicaragua prior to the WB/GEF intervention, as a result of lack of credit, low willingness to pay compared to the high cost of systems and low levels of confidence in the technology. Prior to the WB/GEF project, the government did not provide subsidies, and the institutions that regulate the operation of markets did not protect the rights of, or increase the confidence of consumers via warranty enforcement or standards. Investment opportunities are therefore not highlighted in Figure 8-10a. The WB/GEF project has attempted to create a sustainable market via the provision of customer credit. Increased market density may reduce transaction costs in Nicaragua and therefore improve maintenance performance and the reputation of the technology. However, the project favours imported products and has overlooked the existence of a local manufacturing industry in Nicaragua while the largest of the system retailers has been awarded a contract to supply some of the systems which include Spanish made modules. There is also a risk that the project, at the end of its 5 year life, will not create a self-sustaining market, and will instead leave only the customers who are least able to pay and have therefore not acquired a system through the project even more unwilling to pay the unsubsidised price. Despite the increased
294 Chapter 8. Capability Building at Grupo Fénix
market opportunities via the WB/GEF project, investment opportunities are therefore marked blue in Figure 8-10b.
Figure 8-10: Nicaraguan Technological System for Small Scale PV Manufacture (a) prior to the World Bank / GEF project, and (b) after the project had commenced in 2003.
TECHNOLOGICAL TRAJECTORY
influence perceptions of the provide direction for technological paradigm search
Nicaraguan Small Scale PV Technological System before WB/GEF Project Networks
OTHER ACTORS influence policy and institutions
MARKET & NON-MARKET INTERACTIONS influence operation of markets influence connectivity INSTITUTIONS
knowledge creation & exchange resources for production and innovation investment opportunities provide incentives to invest and improve alter allocation of resources direction & incentives
ENTERPRISE
(a)
TECHNOLOGICAL TRAJECTORY
influence perceptions of the provide direction for technological paradigm search
Nicaraguan Small Scale PV Technological System after WB/GEF Project Networks
OTHER ACTORS influence policy and institutions
MARKET & NON-MARKET INTERACTIONS influence operation of markets influence connectivity INSTITUTIONS
knowledge creation & exchange resources for production and innovation investment opportunities alter incentives to invest and improve direction & incentives alter allocation of resources
ENTERPRISE
(b) 295 Chapter 8. Capability Building at Grupo Fénix
Resources for Production and Innovation The Nicaraguan technological system has been unable to provide virtually all of the resources required for commercialisation of PV technology, both before and after the WB/GEF project. There are very few people with the skills for installation and maintenance of PV systems. The government has not supported the industry through sales tax or import tariff exemptions, or by providing appropriate training, finance or technology development to assist the commercialisation of the technology. There have been no public technology institutes, local consultants or testing facilities to assist manufacturers in developing quality products. Broader small enterprise support such as business development services have also not been available. The materials and equipment for PV manufacture are usually not available in Nicaragua and the poor regulation and transparency of customs facilities and high levels of corruption have constrained the importation of materials and equipment from international suppliers by small enterprises. The poor transport and communications infrastructure has constrained interactions with both suppliers and customers. Through the WB/GEF project, business development services and training have been provided to participating retailers, but the limited time frame (5 years) of the interventions requires that the technological system become self sufficient in the supply of resources within this time. The selection of eligible enterprises, however, has not supported the development of local manufacture, or even the smallest retailers, but has favoured the largest enterprises with existing international links. The resources for production and innovation therefore remain un- highlighted in Figure 8-10b.
Knowledge Creation and Exchange A small number of actors in the Nicaraguan PV supply chain (including suppliers, manufacturers, retailers, installers, maintenance technicians, financiers and technology developers) has reduced the opportunities for knowledge creation and exchange. Even where past projects have attempted to implement maintenance, user education or training schemes, connectivity between actors has been stifled by long distances and poor infrastructure which has increased the cost of transactions and reduced information flows within the supply chain, including for the delivery of after sales service. Knowledge creation and exchange is therefore unshaded in Figure 8-10a. The WB/GEF project has sought to increase the number of and improve connectivity between retailers, installers, maintainers and customers and has provided financial and training resources to improve knowledge creation and exchange. However, the project has not supported knowledge creation and exchange for small scale enterprises, who are disconnected from the actors involved in the project. Because there has been no effort to support local manufacture, there has been no attention paid to the establishment of local suppliers, technology development
296 Chapter 8. Capability Building at Grupo Fénix
or the connectivity of manufacturers with retailers, financiers or customers. There is therefore little increase in knowledge creation and exchange in relation to manufacturing better products for local markets or providing better institutional arrangements for the distribution of local products. Knowledge creation and exchange is shaded blue in recognition of the increased knowledge in the downstream parts of the value chain.
Incentives and Direction for Search The small size of the Nicaraguan PV market has limited incentives to invest in production or innovation. The introduction of standards in the WB/GEF project has provided the incentive to improve in order to comply with standards. Incentives and direction for search are therefore shaded red in Figure 8-10b. However, the process of certification has selectively supported only the largest suppliers in achieving the quality required.
8.4.2. Learning at Grupo Fénix Grupo Fénix has developed a method for the cottage industry production of PV modules, However, it is doubtful that Grupo Fénix can produce modules that would (a) qualify for certification, or (b) maintain their performance under harsh outdoor conditions for 20 years. They are unable to produce modules significantly more cheaply than imported products, which are often better quality. Figure 8-11 shows the part of the framework concerned with learning at Grupo Fénix (a) when production began at Suni Solar in 1998, and (b) when the case study was conducted in 2003. The capabilities and factors that influenced learning in each case are highlighted red where they strongly supported learning, blue where they were partially functioning and not highlighted where they were insufficient. The following sections analyse the importance of, progress in relation to and factors influencing each type of learning.
297 Chapter 8. Capability Building at Grupo Fénix
Figure 8-11: Learning at Grupo Fénix (a) in 1998, and (b) in 2003
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION incentives & resources PRODUCTION direction INNOVATIVE incentives & CAPABILITIES for search CAPABILITIES resources ROUTINES Learning by Doing Learning by Searching new production techniques DOING R&D Improvement Capabilities
RECONFIGURATION
(a)
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION incentives & resources PRODUCTION direction INNOVATIVE incentives & CAPABILITIES for search CAPABILITIES resources ROUTINES Learning by Doing Learning by Searching new production techniques DOING R&D Improvement Capabilities
RECONFIGURATION
(b)
8.4.2.1. Learning by Doing Grupo Fénix began with no knowledge about the manufacture of PV BOS or module assembly. Production capabilities are therefore left unshaded in Figure 8-12a. Learning by doing at Grupo Fénix has resulted in improvements to manufacturing techniques and the development of skills, such as those related to circuit assembly, cell soldering, the construction of frames and the use of RTV silicone. Production capabilities are therefore shaded blue in the
298 Chapter 8. Capability Building at Grupo Fénix
diagram for the case of 2003. However, learning by doing has been limited by inadequate quality control capabilities and routines, lack of availability of materials at a suitable cost and insufficient access to investment opportunities through markets, all of which are left unshaded in 2003.
Figure 8-12: Learning by Doing at Grupo Fénix (a) in 1998 and (b) in 2003
production resources production resources investment opportunities investment opportunities
PRODUCTION PRODUCTION CAPABILITIES CAPABILITIES incentives & incentives & resources ROUTINES resources ROUTINES Learning by Doing Learning by Doing
DOING DOING
(a) (b)
Routines Although there is some quality control checking of modules carried out failure to take cell or module IV curves has prevented Grupo Fénix from learning about the effects of cell sorting, soldering or cell cutting techniques.
Investment Opportunities As a result of limited investment opportunities, the Grupo Fénix enterprises have not been able to operate at full capacity or expand, limiting opportunities for learning by doing.
Access to Resources Grupo Fénix is unable to reliably source the materials it requires for module production at a cost that would enable it to manufacture modules competitively. While there are personnel available in Nicaragua with tertiary education sufficient for PV BOS and module assembly, there are no people with experience of relevant module assembly techniques and few with manufacturing experience more generally. Additionally, infrastructure for importation of goods is inadequate and techniques for small scale module assembly are not available domestically or internationally.
8.4.2.2. Learning by Searching at Grupo Fénix Grupo Fénix began in 1998 with no capabilities for R&D, indicated in Figure 8-13a. Both Suni Solar and Taller Sabana Grande have consciously searched, both internally and through collaborations, to find better or more easily available materials for module manufacture, develop new production equipment and adapt processes to achieve quality improvements and cost reductions. However Grupo Fénix still does not have sufficient innovative capabilities to design BOS or module encapsulation processes that will result in products at sufficient quality
299 Chapter 8. Capability Building at Grupo Fénix
and appropriate cost. Because the systems are dispersed and there is little attention paid to the formalised provision of after-sales service and finance, Grupo Fénix has not gained much experience or innovated in institutional arrangements for the delivery of PV systems and the provision of after sales service that are locally appropriate. Innovative capabilities are therefore left unshaded in both 1998 and 2003.
Figure 8-13: Learning by Searching at Grupo Fénix (a) in 1998 and (b) in 2003
innovation resources innovation resources investment opportunities investment opportunities
COORDINATION & COORDINATION & INTEGRATION INTEGRATION
incentives, incentives, direction INNOVATIVE resouces direction INNOVATIVE resouces for search CAPABILITIES & direction for search CAPABILITIES & direction for search for search
Learning by Searching Learning by Searching
R&D R&D
new production new production RECONFIGURATION RECONFIGURATION technique technique
(a) (b)
Innovation Resources and Knowledge Since good designs from BOS are freely available in the literature and guidelines exist on the correct operation of these circuits, it would be appropriate for Grupo Fénix to use these designs, although it should be noted that the enterprise would require innovative capabilities to adapt them to local requirements and to understand them sufficiently to successfully repair them. However, these designs are not available in Spanish, and Grupo Fénix does not have the ability to find this information. There are no consultants in Nicaragua who would be able to work with Grupo Fénix to find and adapt a suitable circuit and Grupo Fénix cannot access or afford to pay international consultants. While Grupo Fénix has been able to access volunteers for technical assistance, they have not had good enough linkages with these volunteers, as discussed in the following section on learning by interacting. On the other hand, technology for small scale module manufacturing technology does not exist and Grupo Fénix has needed to develop a new technique. Grupo Fénix is constrained in its R&D activities by lack of funds for materials and equipment, small numbers of staff and inadequate education and training. While product certification can provide an incentive to improve, as discussed in the following section, the high cost of these accreditation procedures, and lack of access to technical assistance for compliance are major barriers.
Reconfiguration Both Suni Solar and Taller Sabana Grande are able to reconfigure rapidly because they only have small investments in equipment and a small number of staff that would need to be adjusted to accommodate new production technology. The reconfiguration function is therefore shaded red in both Figure 8-13a and b.
300 Chapter 8. Capability Building at Grupo Fénix
Incentive and Direction for Search The main priority for Suni Solar is commercial survival, that is, selling sufficient volumes of products and keeping production costs down so that reasonable prices and margins can be achieved on the products it produces. Prior to the WB/GEF project, there was limited incentive from the market for it to ensure that its modules comply with standard rating procedures, since customers are not able to know whether the module performs to its rating or not. It is therefore not surprising that the measurement of the IV curves of modules and of cells prior to encapsulation suggested by UNSW was not seen as a priority. The protection of its reputation however became an issue when some of Suni’s panels performed so far below their rating that there was an obvious problem. Suni Solar immediately began testing module IV curves. Incentives are therefore shaded blue in Figure 8-13a. The standards introduced by the WB/GEF Nicaragua project have acted as a strong incentive for Suni Solar to improve sufficiently to receive certification. Incentives and direction for search are therefore shaded red in Figure 8-13b. Unfortunately, the need for accreditation and the process is costly and therefore inaccessible and cannot easily provide direction for improvements.
Coordination and Integration Innovative efforts such as experimentation with LED lights, the production of smaller panels and the improvement of modules have been coordinated with field performance and production experience. However, the knowledge of Suni and Taller Sabana Grande is not shared effectively and Grupo Fénix has a limited knowledge of the market. Coordination and integration is therefore shaded blue for the case of 2003.
8.4.2.3. Learning by Interacting at Grupo Fénix As indicated in the previous discussion, access to resources for production and innovation and to investment opportunities have been limited by Grupo Fénix’s inability to interact with suppliers of materials and technology, and with customers. While the technological system has provided more investment opportunities and resources for production and innovation through the WB/GEF project, Grupo Fénix has not had sufficient linkage and investment capabilities to take advantage of these opportunities.
301 Chapter 8. Capability Building at Grupo Fénix
Figure 8-14: Learning by Interacting at Grupo Fénix (a) in 1998 and (b) in 2003
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
(a)
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
(b)
Investment and Linkage Capabilities Grupo Fénix have had limited linkages with other firms. Grupo Fénix is constrained by its ability to connect with domestic and international suppliers, due to the small size of its purchases, poor communications infrastructure and communications capabilities. It therefore does not learn through interactions with suppliers. Additionally, neither Suni nor Taller Sabana Grande has been able to access finance to invest in production or credit to expand market access. The enterprises are unable to interact with financiers because they not have a proven business model. Access to potential markets has been limited by small scale and lack of connectivity, caused by poor communications infrastructure and Grupo Fénix’s limited presence in rural communities. More significantly, access and influence over the institutional design of project- based markets has been hampered by lack of certification, small size and lack of reputation or influential international linkages. Grupo Fénix was not accredited by the WB/GEF project and its market may in fact be decreased as potential customers purchase subsidised systems through the project.
302 Chapter 8. Capability Building at Grupo Fénix
The ability of Grupo Fénix to link with international universities and individuals has been critical to learning by searching. Linkages between Grupo Fénix and overseas organisations have brought some experts in photovoltaics and business skills, but most of the linkages have been too sporadic, short term and supply driven.
Coordination and Integration Although what is known about local markets is being incorporated into innovative and production efforts, Grupo Fénix has limited information about the nature of its markets to inform its investment decisions. This has been particularly important with respect to the ability to access new markets. Despite their links through Grupo Fénix, interactions between Suni Solar and Teller Sabana Grande have been limited, reducing the sharing of knowledge and resulting in different manufacturing processes between Taller Sabana Grande and Suni Solar. More interaction between the two groups could lead to a better diffusion of knowledge within the organisation as a whole.
303 Chapter 8. Capability Building at Grupo Fénix
8.5. Conclusion
In this chapter, the framework developed in chapter 5 of this thesis has been used to assess factors in the Nicaraguan technological system that have supported or constrained capability building at Grupo Fénix, and to assess the suitability of the capability building strategies adopted. Technology assistance and the study of capability building for small scale enterprises has traditionally been limited to technology transfers and the provision of business development services. The use of the framework in this case study has drawn attention to some factors which impact capability building that are often overlooked. Lack of interactions with other actors, interactions and institutions in the technological system have been found to reduce opportunities for interactive learning, while internal innovative capabilities, which are usually assumed to be unimportant for small scale enterprises, have proven to be critical in the case of Grupo Fénix. The discussion in chapter 10 of this thesis will refer to the analysis of this case study, the case study of the Barefoot College in the following chapter and the pre-existing literature on small scale PV manufacture to suggest the extent to which the findings may apply more generally and therefore to suggest appropriate policies and interventions to support small scale manufacture.
304 Chapter 8. Capability Building at Grupo Fénix
References
Comisión Nacional de Electricidad (2000), Accessed from: www.cne.gob.ni, on: 2002. Coordinadora Regional de Investigaciones Económicas y Sociales (2001), Highlights of Current Labor Market Conditions in Nicaragua, Global Policy Network, Managua. Dross, F., Labat, A., Perez Lopez, M.A., Perez Lopez, M.A., Raudez, R., Bruce, A., Kinne, S. and Komp, R. Vacuum-free, cost-effective, developing-country-material-available solar cell encapsulation, Solar Energy Materials and Solar Cells, In Press, Corrected Proof. Dross, F., Perez Lopez, M.A., Perez Lopez, M.A., Smith, A., Labat, A., Raudez, R., Bruce, A., Kinne, S. and Komp, R. (2006), Capillarity Solar Cell Encapsulation: A new vacuum- free, cost-effective encapsulation technique compatible with very thin string ribbons, 21st European PV Solar Energy Conference, Dresden, Germany, 4-8 September 2006. Environmental Resources Trust (2007), Monitoring, Reporting and Verification Protocol for the Tecnosol PV Project In Nicaragua E+Co Energy through Enterprise, Washington DC, USA. Ernberg, J. and Arce, M.E. (2002), Information and Communication Technology (ICT) in Nicaragua, 7Cs WorldBridge AB, Stockholm, Sweden. Foster, R. and Gómez, M. (2005), Light Emitting Diodes for Photovoltaic Off-Grid Homes, Sandia National Laboratories, U.S. Department of Energy. Frantzis, L., Jones, E., Lee, C., Wood, M. and Wormser, P. (2000), Opportunities for Cost Reductions in Photovoltaic Modules, 16th European Photovoltaic Solar Energy Conference, Glasgow, U.K., 1-5 May, 2000. Government of Nicaragua (2001), A Strengthened Growth and Poverty Reduction Strategy. Grupo Fénix (2007), Web Site, Accessed from: http://www.grupofenix.org, on: June 2007. Inter-American Development Bank (2001), IDB Approves $6,790,000 to Support Technological Innovation in Nicaragua Press Release, Accessed from: http://www.iadb.org/NEWS, on: July 2005. Jochem, F. (2005), El mercado de Energías Renovables en Nicaragua, GTZ. Kinne, S. and Komp, R. (2001), A catalyst for growth : The work of the Grupo Fenix in Nicaragua, Refocus, 2 (1), p 33. Komp, R. (1995), Practical Photovoltaics: Electricity from Solar Cells, Aatec Publications, Ann Arbor, MI, USA. Mathieu, P., De Gouvello, C., Torres, J.E., Sterzinger, G., Barnes, D.F. and INEC (2001), Nicaragua : Sustainable Off-Grid Electricity Service Delivery Mechanisms, A study funded by the Public-Private Infrastructure Assistance Facility (PPIAF), Energy Cluster, Latin America and Caribbean Region, The World Bank. Suni Solar (2003), Business Plan. Velho, L. (2002), Research Capacity Building in Nicaragua: From Partnership with Sweden to Ownership and Social Accountability, Discussion Papers 09, United Nations University, Institute for New Technologies. Vervaart, M.R. and Nieuwenhout, F.D.J. (2000), Solar Home Systems: Manual for the Design and Modification of Solar Home System Components Quality Program for Photovoltaics (QuaP-PV), ECN. World Bank (2003), Project Appraisal Document on a Proposed IDA Credit to the Government of Nicaragua for an Offgrid Rural Electrification (Perza) Project, World Bank, Finance, Private Sector and Infrastructure Department Central America Country Management Unit Latin America and Caribbean Region. World Bank/UNDP (2002), Project Brief - NICARAGUA: Offgrid Rural Electrification for Development.
305 306 CChhaapptteerr 99.. CCaappaabbiilliittyy BBuuiillddiinngg aatt tthhee BBaarreeffoooott CCoolllleeggee
The purpose of this chapter is to analyse capability building in the case of the Social Work and Research Centre (SWRC) in India, known as the Barefoot College, in order to discover typical constraints for small scale PV manufacturers, how capability building occurred in this case, and the ways that the technological system supported or constrained learning. The Barefoot College approach involves the local manufacture and dissemination of photovoltaic solar home systems and solar lanterns by rural community groups via community-based management and training. The case study is of particular interest because the PV systems installed through Barefoot projects have performed relatively well (Jacobson, 1999; Maithel, 1998; Nieuwenhout et al., 2004; Sharma, 2000) and the approach has been replicated widely within India and in a few other countries. The case study fieldwork was carried out during a three week visit, based at the Barefoot College Tilonia campus in Rajasthan, India, during February 2006. Data was collected via participant observation, semi-structured interviews with key stakeholders and through collection of documents and existing literature. The author accompanied and observed Barefoot technicians in their daily routine in remote field centres and through participation in the Barefoot college classroom and field training, including assembling and testing charge controllers and solar lanterns under instruction from a master trainer. The author was in attendance at field centre meetings with solar section staff from Tilonia. Semi-structured interviews and conversations were conducted with Barefoot technicians, system users and field centre personnel in remote villages; and with trainee Barefoot technicians, instructing senior technicians, the head of the solar section and the directors of the college at the Tilonia campus. Notes were recorded with pen and paper during the classroom and field immersion, and during interviews. Documents collected included: Specifications of hardware & prices. Circuit diagrams documenting improvements made to controller, lamp & lantern circuits. Testing results & certification documents for lantern & fixed system. Course notes given to BSEs in training. BSE maintenance manual. BSE installation manual. Practical and theory test procedure for BSEs.
307 Chapter 9. Capability Building at the Barefoot College
Equipment for BSE and rural electronics workshops. Information about projects carried out – numbers of households, systems, costs, procedures.
The status of rural electrification and the use of PV in India are described in section 9.1 of this chapter. Section 9.2 provides background on the history, philosophy and organisation of the Barefoot College and its solar energy projects and section 9.3 details the capabilities of the enterprise. In section 9.4, the framework developed in chapter 5 is then used to establish the roles of different types of learning and the factors in the technological system that have influenced learning in this case.
308 Chapter 9. Capability Building at the Barefoot College
9.1. Background on India
This section reviews the status of rural electrification and the use of PV, the provision of rural infrastructure, support for small enterprises and the role of NGOs in rural service provision in India, prior to the introduction of the case study in the following section.
9.1.1. Rural Electrification in India There is a consensus that the grid will not reach many of the rural poor in India within the foreseeable future. Even in villages that are ‘electrified’, most people cannot afford the connection fees or steep tariffs, which average 110-150 Rs (US$2.80-3.80) per month for a rural household (Chaurey et al., 2004). Houses that are located in hamlets (clusters of houses away from the main village) have to pay extra charges on the basis of the cost of extending the line from the village to the hamlet (Chaurey et al., 2004). Although 80% of villages are considered to be electrified, only 37% of houses actually have electricity (Malhotra & Bhandari, 2002). Where electricity services exist, they are far from satisfactory, frequently characterized by blackouts and brownouts. According to consumers, power cuts of 16–20 hours for 20 days in a month are common (Chaurey et al., 2004). In Rajasthan (the primary area where the Barefoot College operates), only 34% of the rural population have access to electricity, while in four districts (Jaisalmer, Barmer, Dholpur and Jalore), only 20% have access (UNDP, 1999). The expertise within the Indian Ministry of Power is primarily associated with grid extensions, rather than decentralised supply (IEA, 2003). Indian electricity restructuring and privatisation in 1991 and 1993 have failed to make provision for decentralised renewable energy provision to remote communities (Radulovic, 2005). The utilities are financially unviable and under pressure to improve efficiency and electrification of remote regions is an unattractive prospect, due to both the technical difficulty and the high cost of grid extension. A national electricity bill was introduced in 2003 to ensure universal electrification in India (Chaurey et al., 2004). The bill makes explicit provision for the use of distributed energy in rural areas and supports the involvement of local stakeholders such as NGOs and Panchyats (local government organisations) in the management of local electricity provision. The bill also allows generation and distribution in rural areas without a license. There has been a specific Ministry of Non-conventional Energy Sources (MNES) and support for its use in rural electrification in India for many years. Indian renewable energy policy is driven by goals of less dependence on energy imports (65% coal, oil and gas are currently imported), indigenous design, development and manufacture to lead to higher deployment rates and rural electrification for poverty reduction (MNES, 2005b, 2006). There
309 Chapter 9. Capability Building at the Barefoot College
are, however, few integrated policy measures between the Ministry of Power and the MNES (TERI, 2003) or between the MNES and rural development initiatives (Radulovic, 2005).
9.1.2. Support for PV Rural Electrification in India MNES has provided 50% subsidies and interest rate buy-downs for PV systems via the Indian Renewable Energy Development Agency (IREDA). The systems must comply with one of five typologies explicitly sized and specified by MNES in order to be eligible for the subsidies (MNES, 2006). The government also provides customs duty reductions of between 4- 20% on materials and components used by the PV industry. The duty on complete PV systems is higher than for components. Sales tax exemptions also apply on PV equipment in a number of states (Sastry, 2002). The following is a summary of the financial and fiscal support provided for PV implementation:
Financial Measures (MNES, 2006) Direct subsidy – 50% of system and 90% of system in “special areas” o Applies to SHS, solar street lighting and stand alone power plants, Direct subsidies for the preparation of grid connected SPV project applications and to cover project implementation cost and operation and maintenance, Investment interest rate buy down – from market rate of 12% to 5% - MNES covers the difference.
Fiscal Incentives (MNES, 2004) Customs duty reductions for PV equipment and materials, Customs duty exemptions for R&D projects, Customs duty exemptions for goods imported for UN or other international organisations projects, Reduced or exempt from excise duty for RE devices and systems, Depreciation allowed on RE equipment, Tax holidays and deductions for energy generators.
Apart from the central subsidy, there are also a number of other schemes at the state level, but inadequate funding limits the number of systems that can be subsidised. In addition to the subsidy schemes, the government of India have executed and supported a variety of targeted projects and programmes, some of which are listed in Box 9-1. By 2002, about 400,000 solar lanterns, 187,000 home systems, 47,000 street lighting systems, 4,600 solar pumps and a number of 1.25kW mini grids had been installed under the MNES programmes (Sastry, 2002). The Village Electrification programme involved panchayats (village councils), NGOs, and (usually large) retailers, and provided finance and assistance for energy surveys, training and
310 Chapter 9. Capability Building at the Barefoot College
awareness-raising (Sastry, 2003). The Renewable Energy for Rural Livelihoods project, also implemented by NGOs, provided similar capacity building and was focused on the provision of PV for productive uses (UNDP India, 2003a, b).
Box 9-1: Government of India PV Projects and Programmes
UNDP/GOI Renewable Energy for Rural Livelihoods (UNDP India, 2003a, b) GOI (MNES) executing NGOs implementing at community level UNDP funding Ramakrishna Mission (NREL/MNES/TERI) (Stone & Ullal, 1998; Stone et al., 2000) Ramakrishna Mission implementing TERI pre and post implementation studies MNES and NREL 50/50 funding GOI Water pumping programme (MNES, 2005a) financial incentives development of marketing infrastructure GOI Remote Village Electrification Programme (MNES, 2005a) 1kW/household/day target 200 test projects underway- mainly power plants, mini-grids (especially biomass gasifiers) financial support from UNDP (see RE for rural livelihoods project)
There has also been World Bank, Global Environment Facility and Asian Development Bank funding for PV projects in India, including the provision of technical expertise and funding for capacity building. The World Bank PVMTI (PV Market Transformation Initiative) funding is supporting the Sri Shatki project, through which 300 retailers of PV systems will be assisted; and the SREI (Shell Renewable Energy India) project, which aims to develop a retail and service network and support consumer financing in West Bengal (Chaurey, 2001). There have been numerous other NGO projects, notably the SELCO effort, which is a business model for PV sales that includes assurances on system performance and after sales service (SELCO, 2007).
PV Standards and Testing Facilities MNES has developed standards for SHS that must be complied with in order to receive funding either through a direct subsidy or an interest rate subsidy (MNES, 2006). NREL has worked with the Solar Energy Centre of India’s MNES to develop test procedures for PV modules and systems (Stone et al., 1998). The standards are less stringent than standards for many internationally funded projects and the PV-GAP standards. For instance, DC lamp ballasts which produce lower quality waveforms are permitted. While these ballasts reduce the life of the CFL tube, they require simpler cheaper circuits, which are less prone to failure. Testing facilities have been established to assess products against the standards, provide manufacturers with performance guidelines and advise them on improving their products if they are not able to
311 Chapter 9. Capability Building at the Barefoot College
comply with standards. The Solar Energy Centre in Gurgaon, Haryana is the lead training and testing centre, and there are three other test centres (Kumar & Sastry, 1998).
Government Support for PV Rural Electrification in Rajasthan The Rajasthan Renewable Energy Corporation (RREC) executes all the MNES schemes in Rajasthan (the state where the main campus of the Barefoot College is located), including the central subsidy scheme. RREC’s activities are limited by financial resources from the MNES, and by inadequate field-level infrastructure. The distribution of subsidised PV systems by RREC therefore makes no financial provision for maintenance, replacement or capacity building to carry out these functions (Chaurey et al., 2004). Chaurey et al. (2004) believe that unless the local communities and NGOs take responsibility for rural electrification, most of the people in Rajasthan will remain without electricity, since RRED does not have the capacity to service most of the state.
9.1.3. A Critique of the Indian PV Programmes
Lack of Inter-Agency and Micro-Macro Coordination The Indian renewable energy efforts have been criticised for failing to coordinate the various ministries and programmes which require and fund energy, such as the water, agriculture, power and rural development ministries. Within most rural development programmes, appraisal of the role of energy is not institutionalised, and these programmes usually do not have the capacity to carry out rural energy assessments (Rehman, 2002). There is also a lack of participation and micro-macro coordination between rural energy stakeholders such as the rural poor, district level and local government development agencies, public works, contractors and ministries (Malhotra & Bhandari, 2002). As a result, the different circumstances and needs of different communities are not recognised, the poorest are often overlooked, local people are not empowered and inappropriate technologies are used.
Target Driven and Inappropriate Programmes The national programmes have also been criticised for being overly target-driven, and failing to ensure the quality of hardware or the appropriate institutional arrangements for the marketing and maintenance of technologies (Malhotra & Bhandari, 2002). As a result, there are usually insufficient locally available trained personnel to support technologies (Government of India, 1997). There has also been little invested in market surveys or R&D to ensure that appropriate technologies are developed for rural people. For example, of the expenditure on biogas technologies, only 2% has gone to R&D (Rehman, 2002), and there has also been little follow up based on the numerous evaluations of renewable energy programmes (Malhotra & Bhandari, 2002). While the rigid specification of systems that qualify for the PV subsidies is intended to reduce the costs of achieving targets and increase quality, it has not facilitated the
312 Chapter 9. Capability Building at the Barefoot College
development of the most appropriate technologies. For example, the 10Wp solar lantern is a number of times brighter than a traditional kerosene lantern (Rehman, 2002), while a smaller lantern could be affordable to more of India’s poor.
Poor Access to Credit IREDA (Indian Renewable Energy Development Agency) has promoted the provision of credit for remote PV systems through rural cooperatives, state nodal agencies, commercial banks, non-governmental organizations, and renewable energy technology manufacturers (Radulovic, 2005), but has avoided providing loans directly in rural markets due to problems in recovering loans and high transaction costs from working with multiple stakeholders (Chaurey, 2001). Government support for micro-credit is otherwise focused on agriculture, and there has been little effort to tap into these funds for rural energy (Malhotra & Bhandari, 2002). Where there is no locally acting agency supported by IREDA, the rural poor have therefore had difficulty accessing credit.
Faulty Application of Subsidies The application of energy subsidies in India has in many cases failed to reach the very poor and acted against the dissemination of renewable energy. For instance, the Indian government has, since the Green Revolution that began in the middle of the 20th century, subsidised diesel and supplied free electricity to farmers (Bhalla & Reddy, 1994). After the Revolution ended, the subsidies remained, and many still expect free electricity (Chaurey et al., 2004) and do not want to pay for electricity from distributed sources. Kerosene is also subsidised at massive cost to the government and extensive infrastructure for it’s distribution has been established. Rehman (2002) calculated that the money invested in this subsidy could instead be used to subsidise the purchase of solar lanterns by 50%, with the remaining cost covered over three years by the purchaser. In this case, all unelectrified households could purchase a solar lantern within five years. The government subsidies for PV have been changed frequently, creating uncertainty in the market. At the time of the fieldwork, the 50% national subsidy was not available because it is allocated on a first come first served basis and the funding limit had been reached. Rumours indicated that the subsidy had been discontinued and that the new management of MNES favours biomass for rural electrification. The uncertainty about the subsidies causes people to wait, often in vain, for the lowest price, thus making it difficult to establish viable businesses.
9.1.4. Rural Infrastructure and the Small Scale PV Industry Most of the population of rural India is afflicted by extreme poverty, low agricultural productivity and incomes, limited technological capabilities, and poor access to services such as
313 Chapter 9. Capability Building at the Barefoot College
health, education, clean water and energy (Bhalla & Reddy, 1994), which hampers the growth of non-farm activities. Infrastructure in much of rural India is very poor. The use of electricity in rural areas and the capital-intensity of enterprises are very low. Rural roads cannot be accessed by 52 per cent of the rural population. The average rural teledensity is only 1.278, compared to an overall teledensity of 66.13 for the US, 14.26 for the world, and 2.2 for the whole of India (Bery et al., 2003). India has policies that explicitly support rural small scale industries (Saxena, 2003), including fiscal incentives and tax exemption, the reservation of some products for production in small scale industries, common production programmes, exemptions from some labour and licensing regulations, government procurement and price preference for rural industry products, infrastructural support through rural industrial estates and district industry centres and training programmes (Bhalla & Reddy, 1994). However PV is not a reserved industry, and only benefits from government procurement and subsidies as previously mentioned. Finance for PV systems, even with government subsidies, is insufficient for most of the poor. Formal vocational training for PV technicians or manufacturers is not available outside of specific programmes, such as those mentioned in the previous section. Most of the manufacturers of PV BOS and suppliers of PV systems in India are small enterprises, without the capacity to develop marketing networks or carry out R&D (Rehman, 2002). In the absence of financial assistance, these enterprises can usually only interact with large buyers, such as projects and programmes. As a consequence, there is inadequate linkage with or provision of after-sales service to rural users via local actors such as retailers and technicians.
9.1.5. NGOs and the Panchayat System
NGOs The NGO sector in India has played a major role in rural development, by mobilising communities towards social change and through direct implementation of interventions concerned with specific development issues. NGOs are recognised as well placed to implement projects at community level (Malhotra & Bhandari, 2002). However, the scale at which NGOs operate is too small to have a significant impact on the large number of rural poor in India. NGO projects are therefore often designed as demonstration projects. Rehman (2002) suggests that they should be better coordinated with government programmes in order to harness their strong local knowledge and links.
Decentralisation through the Panchayat System There has been a move towards decentralisation of government authority through the Panchayat system, advocated by Ghandi and implemented in the 1960s. National constitutional
314 Chapter 9. Capability Building at the Barefoot College
changes were made in 1992 to accommodate the system, which was adopted in Rajasthan in 1996. Most Indian states have now adopted the Panchyat system, which involves three levels: village, block and district. At the village level, members are elected and decisions are taken by vote. Representatives are elected by the Panchayat area for five-year terms. Seats are reserved for Scheduled Castes, Scheduled Tribes (both proportional to the population) and one third of seats are reserved for women (Vania & Taneja, 2005). There is some funding for training of elected members available through the Indian Ministry of Rural Development. The system is intended to speed decision making and increase democracy and transparency (Johnson, 2003). The responsibilities of the three tiers of district administration include planning and implementation of development and social justice activities and collection of taxes and other fees. Matters set aside in the Constitution for administration at the local level include rural electrification and maintenance of community assets, while khadi, village, cottage, small scale, and cooperative industries are commonly the responsibility of the block level (Panchayat and Rural Development Department, 2007). The village Panchayats are also often made responsible for the local administration of government funded development and poverty alleviation schemes and the identification of beneficiaries under such schemes. The devolution of power and of funds to Panchayats implies that communities will now have access to public financial resources for development programmes of their choice. There is, in principle, a supportive environment within government where community led development efforts, such as the Barefoot College, can be taken seriously.
Failures of the Panchayat System However, it is reported (Administrative Reforms Commission, 2001) that State Governments have frequently not transferred power over the full range of areas of responsibility listed in the Constitution, or only transferred a fraction of the functions with respect to the area. State Rural Energy Departments have in some cases invited the Panchayats to transfer their programs to them. Almost all government sponsored poverty alleviation and development programs are still controlled by government ministries separate from the Panchayat system (Administrative Reforms Commission, 2001). Panchayats are functioning simply as agencies of the Government and not as institutions of self governance. There has also been very little effort to provide the necessary administrative support to the Panchayats, which lack the administrative, technical and financial resources to carry out many of the roles assigned to them (Malhotra & Bhandari, 2002). The system also suffers from lack of coordination. There are often a large number of committees and user groups at village and block levels that are exclusive of the panchayat system and have overlapping functions. These groups are often short-lived, overlapping in function and the resources could be consolidated if there was coordination between these groups (Malhotra & Bhandari, 2002).
315 Chapter 9. Capability Building at the Barefoot College
The system has nevertheless been accepted and in many respects effectively implemented in most states of India, although political and caste issues have prevented universal adoption (Human Rights Watch, 1999). India still suffers widespread corruption, with power structures and caste systems that disenfranchise women and the poor. “Institutional weakness and governance issues exacerbate the lack of funds” (World Bank, 2000 exec. sum). Kothari (1993) states: “Laws have been enacted but rarely implanted. Policies have remained on paper, as collections of pious intentions without workable action plans. Few programs that have been implemented have rarely reached the intended beneficiaries, especially in the manner required. Reservations, representations and various fiscal benefits have either been fraudulently diverted to ineligible individuals or have been restricted to very narrow elites of the economically weaker and minority communities” (Kothari 1993, cited by Jauhari & Kondo, 2003).
This is a difficult context in which to achieve community organisation. However, there have been a number of examples of experimentation with community-run service provision in rural India since the 1970s, including night schools, hand pump maintenance and health centres.
316 Chapter 9. Capability Building at the Barefoot College
9.2. Background on the Barefoot College
In this section, the history, philosophy and organisation of the Barefoot College are described. An overview of the operation of the solar programme of the Barefoot College is then given.
9.2.1. History & Philosophy of the Barefoot College The Barefoot College, officially known as the Social Work and Research Centre (SWRC) is a community development organization that was founded in a small rural town called Tilonia, near Ajmer in Silora Block, Rajasthan in 1972 with the mission of sustainable development through self-sufficiency. The central idea was that urban professionals and rural people could form a partnership that would enable rural people to take advantage of the expertise of educated urban people.
Origin and Start Up When urban social workers first formed SWRC, they aimed to work on the issue of most importance to the local people. Water was almost unanimously the greatest need, despite the existence of a nationwide government hand pump program, through which thousands of hand pumps were installed throughout India in the 1970s. A three tier system had been designed by UNICEF for the maintenance of the pumps, but the combination of bureaucratic inefficiencies, corruption and logistical problems in covering long distances with urban-based technicians over difficult terrain resulted in thousands of pumps lying disused. In the three tier system, a caretaker from the village monitored the pumps for free and there was a mechanic in each block (a collection of villages) who depended on mobile maintenance units. The caretaker had a limited function and was unpaid, while all the responsibility and high salaries went to people who were not from the rural area. Local people wanted new solutions to water supply, since an entire government department and a battery of technicians and administrative personnel were not effectively maintaining the hand pumps. Semi-literate and landless youth were therefore trained by SWRC to perform repairs on the hand pumps and given responsibility for their maintenance. The youth did not have formal engineering or technician training, but they had practical skills gained through working and improvising in the village. Additionally, they had a stake in the efficient working of the scheme, since they and their families suffered lack of water and they were also at risk of losing the respect of the people in the village if they did not maintain the pump effectively. The hand pump mechanic scheme was a success and the approach and Barefoot training was adopted by the state government.
317 Chapter 9. Capability Building at the Barefoot College
Through experience, SWRC realised that development is a long term process, and that people, rather than institutions effect change. The need to empower communities to handle their problems with their own resources and effect change, including political change, was recognised. Within the Barefoot College, therefore, changes were implemented such that local people were to take control of the organisation, occupy positions of authority and become experts, in order to construct an organisation that would fulfil their requirements. Many of the urban professionals left the village when collective decision making processes led to all members of the group receiving much the same salary and to a flat organisational structure with more local people occupying positions of authority. The Barefoot College is now characterized by a decentralized and horizontal power structure, with much of the decision making occurring at the village level. Field centres were established to directly collaborate with the poor in various villages. Around 95% of the staff of the Barefoot College are from the rural villages within which they work. The Barefoot College works with villages on multiple areas of development at one time. Typically, there will be a night school, Barefoot hand pump mechanics, women’s committees and education committees in a community as well as the structures necessary for the operation of the PV programs operating in one village.
Philosophy of the Barefoot College Five non-negotiable principles of the Barefoot College have been identified and quoted by various authors (Chasi & Mussukuya, 1998; O’Brien, 2000): Equality: The Barefoot College upholds that every person in society is important and must be respected. Anyone, regardless of caste, class or gender is eligible for any staff position. Collective Decision-Making: The Barefoot College has attempted to remove both formal and informal hierarchies and encourages all the staff to participate in decision- making processes. Decentralisation: The Barefoot College is committed to grassroots level planning. Rural communities can identify their needs and solve their own problems when they have access to information and education. Self-Reliance: When people develop self-confidence and join together to collectively solve their problems they learn that they can depend on themselves. The Barefoot College aims to build self-esteem and self-reliance through their programs. Austerity: The staff lead a simple life working for the collective aspirations of rural communities rather than striving for individual material goals.
318 Chapter 9. Capability Building at the Barefoot College
While these non-negotiable principles probably still hold, the organisation consciously embraces change, and the conflict that inevitably accompanies change. In general, the organisation is accessible, approachable and flexible, giving it the ability to reach the poor, disabled and disadvantaged. The need for upgrading of skills and ability to fight corruption and seek justice is recognised. Political and social change is seen as central to the development process.
9.2.2. Organisation of the Barefoot College Since the decision making in the Barefoot College is decentralised, it is depicted by the organisation with the village level committees at the top, as pictured in Figure 9-1.
Figure 9-1: Organisational Structure of the Barefoot College
VILLAGE LEVEL COMMITTEES
Energy and Children's Education Piped Water Women's Groups Environment Etc. Parliament Committees Committee Committee
FIELD CENTRES
Chhota Narena Kankalwada Singla Brijpura Kadampura Tikawada
SECTIONS
Education Pre-School Education Health Water Rural Handicraft Communication Solar Women's Development Technology Dissemination
Administrative Development Finance
DIRECTOR | GROUP OF IN-CHARGES FOR SECTIONS & FIELD CENTRE COORDINATORS
BOARD OF GOVERNORS
319 Chapter 9. Capability Building at the Barefoot College
Village Level Committees The Barefoot College programmes are run through village level committees. The village committee supervises the implementation and the financial aspects of a programme in the village, including supervising, monitoring and paying workers. There may be a number of committees relating to different areas of work in a village. For example, the solar programme would be supervised by the village energy and environment committee, while the night school programme would be supervised by the village education committee and the hand pump mechanic programme would be supervised by the village water committee. These committees may work together when their interests overlap. For example, the education committee may report to the energy and environment committee about the performance of the solar technician who maintains the solar lanterns in the night schools.
Field Centres Field centres are set up in some villages where Barefoot activities are well established. These field centres serve as outposts to groups of between 10-35 villages, providing feedback and assistance to village committees and as communication conduits between the villages and the sections at Tilonia when necessary and through regular meetings. They also distribute resources to the villages and can serve as a training ground for workers from the villages. Some field centres have become or began independently from the Barefoot College in Tilonia, run their own projects and have their own sources of funding and governing body. Many of them are officially registered as independent NGOs. More than 25 NGOs in 14 states are affiliated to the Barefoot College through the Sampada1 network (O’Brien, 2000). The majority of these organisations were founded by former SWRC graduates and staff members, often beginning as field centres. The first independent field centre was in Jawaja in 1974 (EZE, 1996). These field centres depend on Tilonia only for certain overarching policy and administrative needs. Tilonia provides all of them (EZE, 1996) Some start up financial support, Training for the chief functionary, Training to staff at various levels, Support at critical periods both internally and with external interactions.
There are a variety of approaches that have been adopted in different Sampada members with different management styles and different philosophies. EZE (1996), however, identified five common principles within the network: The choice of a village base, Integrated approach,
1 Sampada is a Sanskrit word meaning ‘natural wealth’ or ‘heritage’. 320 Chapter 9. Capability Building at the Barefoot College
Professionalising volunteerism, Non ideological position, Investing in people, rather than in projects.
Sections At the Tilonia campus, sections administer funding, monitor programmes, and carry out research, dissemination and training related to different programmes. There are eight sections: Water, Education, Community health, Solar energy, Rural industries (handicraft), Forestry agriculture and animal husbandry, Women’s development, and Communications.
Members of the sections regularly travel to field centres to provide advice and assist with technical or administrative problems.
Central Administration Within the Barefoot College, the decision making process has and is still evolving. Decisions made by the village committees are generally regarded as non-negotiable, within the bounds of the non-negotiable principles previously outlined. The governing body (Tilonia or an independent field centre) acts as a rubber stamp, a central fund raising facility, and a place for policy formation. Village committees and field centres rely on the central governing body for insights and advice based on extensive experience. Two day meetings each month are held with the governing body: a committee made up of the director at Tilonia (Bunker Roy), all heads-of sections, all field centre coordinators and all heads-of-administration. Planning for the next month and allocation of funding is carried out. Proposals are presented and discussed and need to be endorsed by all sections. There is still a strong democratic process. Geospecific problems in the area of field centres in relation to issues such as migration, field mapping and project proposals are raised by the field centre coordinators. Programme specific issues are raised by section heads. Supporting the operation of the sections and the central decision making process are finance, administration and development committees.
321 Chapter 9. Capability Building at the Barefoot College
Financial Barefoot College programmes are on average 30% financed internally. The Barefoot College generates income from professional services and the sale of goods including handicrafts. The Government of India contributes another 30%, towards eight programmes where NGOs such as SWRC are seen as better placed to deliver services than government agencies. 40% comes from aid agencies such as UNDP (photovoltaics), UNESCO (night schools and filmmaking), EC (photovoltaics), World Bank (filmmaking and rainwater harvesting), HIVOS (integrated activities), Plan International (night schools), USAID and Save the Children (Chasi & Mussukuya, 1998; O’Brien, 2000). The Barefoot College PV programme has also been supported by the Council for Advancement of People’s Action and Rural Technology (CAPART), which was set up in September 1986, as a supporting and funding agency for the non-governmental organizations (NGOs) which propagate appropriate rural technologies (CAPART, 2007; Ministry of Rural Development, 2002). The funding is generally project-specific. Infrastructure for various programmes is often provided through funds specifically allocated by the donor to projects. Most of the ongoing costs of running programmes are paid for by the communities, who pay for the services they receive. Infrastructure and wages for the training and research facilities at Tilonia are paid for internally. 91% of the Barefoot College funds go towards programmes, and 9% is absorbed in administration.
9.2.3. Solar Energy Projects The Barefoot approach to rural electrification (Roy & Joshi, 2005) involves creating village-level capacity in manufacture, installation and maintenance of solar PV systems through a network of Rural Electronic Workshops (REWs) and trained Barefoot Solar Engineers (BSEs), and involves the creation of participatory training and maintenance networks, as well as financial and decision making structures at the village level. The Barefoot College first worked with photovoltaics in 1989. Since 2002, about 260 home lighting systems, 500 solar lanterns and two 2.5kW systems have been installed within India and 1000 home lighting systems have been installed outside of India through SWRC programs and with BSEs trained. While the Barefoot College receives many requests to implement solar programmes, smaller, more inaccessible villages are taken on in line with the Barefoot College’s non-negotiable principles. The Barefoot College’s work with a community usually starts with a written proposal from the community. More requests are received than SWRC has the capacity to implement in Silora Block, so the neediest are taken on.
322 Chapter 9. Capability Building at the Barefoot College
9.2.3.1. The VEEC A Village Energy and Environment Committee (VEEC) is formed within a village prior to any decisions being made about a solar programme. The VEEC comprises about 20 members elected by the community. The members will be users of the systems, and must comprise half men and half women, as well as a representative from the field centre. The VEEC administers the solar programme in the village, collects the service fees and operates their own bank account. The VEEC undergoes training in community development and the Barefoot approach (see Appendix 7: SWRC/EU/UNDP Training Modules 2000). The training sessions are highly participatory and involve the committee: Identifying the needs of the village, the dynamics of power in the village, the causes of and responsibilities for inequities and the relationship between the village and the government service providers. Examining the role of the community, the government in development and the potential role of SWRC and village level community organisations. Learning about the Barefoot approach, the particular project that they are part of and the roles and responsibilities of agencies within the project. Preparing for resource mapping. Coming up with an action plan for initiating organisation in the village, resource mapping and surveys and logistics of choosing the BSEs, establishing the REW and coordination with other agencies.
Through the training, the VEECs become aware of the need to discourage the economically or politically powerful from hijacking the project or the functioning of the BSEs or REWs. When dealing with communities for the first time, even the non-negotiable principles such as participation of women or considering the interests of the poor have to be negotiated in order to have them agreed upon and followed systematically. At a village meeting, the VEEC selects the poorest semi-literate and untrained who have roots in the village to be solar engineers. Those selected are usually youth, women or disabled, as they belong to the most economically vulnerable among their communities. These people have been found to be more likely to stay in the village after their training since they come from, are loyal to and understand the community, and have few other opportunities. In some cases they may be night school attendees or women who have been identified as disadvantaged through the women’s committee. The community also selects the village and the families to whom solar will be provided. The community decides on the appliances that they need, and therefore the configuration required to supply the loads. The VEEC also determines how much each family in the
323 Chapter 9. Capability Building at the Barefoot College
community is willing and able to pay upfront and on a monthly basis for the units, and decides the wages of the BSE. The government minimum wage is 73Rs/day = 2200 Rs/month. BSEs are usually paid between 2200 and 3000Rs/month. A bank account is opened, with three signatories from the VEEC. The upfront payments and the monthly payments are invested in interest earning accounts. From the bank account, the BSEs wages and the cost of parts and components for repair and replacement are drawn. The initial cost of the hardware is almost always covered through a donor and does not need to be recovered. The financial arrangements are further detailed in section 9.2.3.5. The VEEC and the elected members of the student parliament and other interested committees are charged with ensuring that the systems are performing as expected and the BSE is doing her/his job. The VEEC also ensures that the BSEs are able to collect the monthly instalments from the users. In instances where the VEEC system has not been followed, the likelihood of appropriate decision making has been reduced. For example, Ethiopian trainees who were sponsored by their government were commonly related to government officials and were not strongly motivated to complete the training or carry out the role of BSE. Although there is always potential for the VEEC to take inappropriate decisions, it seems likely that, given that Tilonia or the affiliated sub-centre has control of the funding and the community depends on the training provided by them, they would be able to strongly influence the committee to follow the Barefoot approach in selection of personnel and planning, at least during the planning phase of the project.
9.2.3.2. The Barefoot Solar Engineers Barefoot Solar Engineers (BSEs) are trained over a period of six months, usually via live-in training at the Tilonia campus, but sometimes at rural electronics workshops. During the training, they learn to assemble the electronic charge controllers, lamp circuits and solar lanterns; and to install and maintain the photovoltaic systems and solar lanterns at a local level. The training is described in more detail in section 9.2.3.3. Table 9-1 lists statistics about some of the BSEs that were in attendance at the Tilonia campus during the author’s visit. Many of the trainees were women and Muslim tribal people, some of whom had no formal education at all, two of whom were disabled physically and one mentally. While most trainees complete the training over a period of six months, two of the trainees at Tilonia when the author visited were expected to take longer to train, because one of them had no formal education, and one was mentally disabled. Both of them are also basically illiterate.
324 Chapter 9. Capability Building at the Barefoot College
Table 9-1: A Sample of Barefoot Solar Engineers Name Age Role Starting Sex Education Village District Religion Year Dashrath 33 MT 1993 M 10 Chotanarana Ajmer Hindu Chagni 44 MT 2000 F 8 Champura Naboor Hindu Jugal 20 T 2005 M 6 Sabalpoor Naboor Hindu Kalu 20 MT 2003 M 0 Nohoria Ajmer Hindu Firoz 22 MT 2000 M 8 Tilonia Ajmer Hindu Gordon 22 T 2005 M 11 Mundoti Ajmer Muslim Kanta 25 T 2005 F 0 Mundoti Ajmer Hindu Najima 44 MT 2003 F 0 Hermarda Ajmer Muslim Sunita 31 T 2005 F 0 Hermarda Ajmer Hindu Suresh 43 T 2006 F 12 Jaliya II Ajmer Hindu
The responsibility given to the engineer to install and maintain the systems is a source of pride and validation. They earn the respect of the communities they serve because of the service they are able to provide.
Box 9-2: Profiles of Solar Engineers
Dewa Ram Prior to becoming a solar engineer, 19 year old Dewa Ram was very poor. His father died when he was 3- 4 months old. He had no income source, but has strong community contacts and is very committed to his work. Dewa Ram is the solar engineer at Solata. He comes 7km by bicycle from a nearby village each day. He maintains 136 solar lanterns (100 in houses and 36 in night schools) and 200 fixed systems, as well as the 2.5kW array on the field centre building. The fixed systems are privately owned systems that were purchased from SWRC. Kewa-Ram gets paid for repairs carried out on fixed systems and for making new circuits.
Dewa Ram was trained in Tilonia for 1 year. He is now able to solve many solar problems. He started working as a Barefoot solar engineer 1 ½ years ago. He visits each solar lantern twice per month. He works on solar activities for 8 hours per day.
Pama Ram Pama Ram is responsible for 50 solar lanterns and 80 fixed systems. He works for 2 hours/day on average and is paid a stipend of 2200 Rs/month as decided by the Kotri VEEC. Pama Ram can make new circuits in the REW at Kotri, and he gets the raw materials from Tilonia. He is paid 10Rs for a lamp circuit and 15Rs for a lantern circuit, which he sends to Tilonia when complete. Pama Ram has trained the users to clean the modules and top up the batteries.
Pama Ram also has other jobs, such as working at the night school. He takes this opportunity to check on the night school lanterns and systems nearby. Otherwise, a message may be sent that a system needs maintenance, and he will attend to it. If he can’t fix the problem in the field he brings it back to the REW and fixes it.
Hanuman Hanuman worked in tailoring, grocery shop and agriculture before training to be a BSE. After 6 months of training, he has been working for 8 months as a BSE along with another BSE in Guda. Hanuman finds the identification of components for making a circuit the most difficult part of his job, and enjoys best fault finding and repairing lamps and chargers, since this is the easiest part for him.
Once the training is complete, the BSEs participate in the installation of the systems they will be responsible for, and in the setting up of REWs in their area. The engineers are then well placed to continue their role of installing SHS and solar lanterns and repairing and maintaining them. Each BSE services between 20 and 50 night schools with solar lanterns, and
325 Chapter 9. Capability Building at the Barefoot College
up to 200 fixed PV systems in homes or on community buildings. Some BSEs, who are located at Rural Electronics Workshops, also continue to fabricate charge controllers, lamp circuits and lanterns, for which they are paid on a per unit basis. Trainees may emerge as trainers after 4-5 years of experience, capable of participating in the training of further BSEs.
Figure 9-2: Barefoot Solar Engineer Hanuman diagnosing a fault in a Solar Lantern in Guda Village.
Every month, the BSE meets and discusses log book problems with members of the VEEC. Members of the solar section from Tilonia or other field centres are sometimes in attendance. Refresher courses may be recommended for the BSEs if problems emerge with the maintenance of the systems.
9.2.3.3. Training of the BSEs The aim of the BSE training is to make the technology easy to understand, which the Barefoot College calls ‘demystification’ of technology. BSEs participate in a six month live-in training course, which covers system installation and maintenance, fabrication and testing of lamps, charge controllers and solar lanterns in the conditions of a REW, management of a REW, fault diagnosis and repair/replacement of system components. The contents of the training are detailed in Table 9-2.
326 Chapter 9. Capability Building at the Barefoot College
Table 9-2: Training of a BSE
Correct module and battery placement, lamp and controller fixing System installation: Wiring of all components Understanding of series and parallel connections Learning how to fabricate transformers to specifications Fabrication and testing of inverters for Being able to identify resistors, capacitors, diodes, CFLs, charge controllers and solar transistors, ICs, lanterns in the conditions of a rural workshop: Understanding the function of each electronic component Assembling and soldering the circuit, wire cutting and stripping Stock management and storage of components and parts Establishment and running of a REW: Operation and care of the equipment in the workshop Documentation of work carried out Being able to use a power source Fault diagnosis and repair/replacement of defective components using portable Being able to use a multimeter to measure current, equipment and the equipment in the voltage and resistance REWs: Knowing the probable cause of different malfunctions i.e. the components most likely to cause a particular type of fault The conditions under which the which the charge controller should start and stop charging the battery Understanding the correct functioning of The conditions under which the load should be the charge controller, lamp inverter and disconnected solar lantern: The correct brightness expected from the lamp The voltage expected across the battery The capacity of the system
The training for Barefoot engineers is similar to a technical college education, whereby the BSEs learn to fabricate, operate and maintain the circuits, but are not able to mathematically design, model or predict the behaviour of the circuit. Throughout the training, trainees handle the electronic components, and the completed charge controllers, inverters and lanterns as well as batteries, modules and complete PV systems. During the training, BSEs make regular field visits, accompanied by master trainers. Maintenance checks and repairs are carried out on systems, constituting valuable practical on the job training in repair, maintenance and testing of SHS and solar lanterns. Leadership and trainer training, in preparation to teach users to maintain their systems and future trainees to become BSEs, also covers subjects such as: Understanding responsibilities towards the community. The Barefoot College approach to community development. An understanding of issues relating to community development, such as health, women’s rights, education as well as energy and water.
327 Chapter 9. Capability Building at the Barefoot College
Development of communication skills for the dissemination of information such as puppet handling, street theatre and songs, writing scripts for puppet plays and street theatre.
Figure 9-3: Training at the Barefoot College
A test including both theoretical and practical components is taken at the conclusion of training. Approximately two thirds of trainees pass the test the first time. Those who don’t pass continue to study until they reach a pass standard. The test paper, a translation of which is made available in Appendix 8, provides an insight into the extent and detail of the competency required of a BSE.
9.2.3.4. Rural Electronics Workshops (REWs) Rural Electronics Workshops (REWs) are established by the BSE for a group of villages such that the REW is easily accessible by the BSE from each village within hours of a complaint. The REWs are always established in accommodation provided by the community, either purpose built, or in some cases a room in a family home offered by the family. The VEECs and the field centres agree to the location of the REW. Table 9-3 contains a list of REWs established between 1989 and 2001. More recently, REWs have been established within Rajasthan in Kalia, Guda, and Kamala. A REW is equipped such that the BSEs are able to fabricate and repair electronic charge controllers and lamp ballasts for fixed solar systems and solar lanterns. All the components required for repair and maintenance are also available at the REW, which is located within easy access of the PV systems and lanterns. The cost of establishing a REW in remote Ladakh in 2000 was calculated at Rs 204,815 (US$4300), including a 300W PV system, DC power supply,
328 Chapter 9. Capability Building at the Barefoot College
digital and analogue multimeters, a transformer winding machine, PCB assembly card trays and jig and soldering irons, along with many other hand tools and a solar panel for testing. The full details of the equipment at the REW in Ladakh are given in Appendix 9. The average REW has a smaller PV systems and is less well equipped, so costs less than half as much.
Table 9-3: Rural Electronics Workshops established between 1989 and 2001 At REW - BSEs fabricated and No of produced BSE’s No. Location/Org. Established District State Fixed trained at Solar Solar the REW Lanterns Units (No.) (No.) 1. SWRC, Ajmer 1996 Ajmer Rajasthan 3 67 531 2. Bikaner Bikaner Rajasthan - 15 - 3. Alwer Alwer Rajasthan - 15 - 4. Nagour Nagour Rajasthan - 20 15 5. Tonk Tonk Rajasthan - - 2 6. SWRC, Kargil 1990 Ladakh J &K 15 467 4 7. SWRC, Leh 1991 Ladakh J & K 25 1215 112 8. SWRC, Sadam 1996 South Sikkim Sikkim 20 239 300 9. SWRC, 1997 Pithoragarh Uttranchal 8 204 160 Tripuradevi 10. SWRC, Tingret 1997 Lahaul & Himachal 5 100 - Spiti Pradesh SHS 11. Sankalp, 1998 Baran Rajasthan 7 226 100 Mamoni 12. Prayatna, 1999 Jaipur Rajasthan 2 2 Solavta 13. Manthan, Kotri 1999 Ajmer Rajasthan 2 21 34 14. SWRC, Jawaja 1999 Ajmer Rajasthan 2 - - 15. ALOK 2000 West Bihar 4 - - Champaran 16. SWRC, 2000 Kokrajhar Assam 2 - 200 Runikheta 17. SAMPARC, 2000 Jhabua Madhya 2 200 381 Raipuriya Pradesh 18. SWRC, Dhanau 2000 Barmer Rajasthan 2 - - 19. SWRC, Zanskar 2001 Ladakh J & K - 3 - 20. Sarthi - Panchmahal Gujrat - 20 - 21. Koiambtoor - Koiambtoor Orissa - 115 - 22. Ranchi - Ranchi Bihar - 90 - 23. Kalahandi - Kalahandi Orissa - 115 - 24. Keonjar - Keonjar Orissa - 10 - 25. Daporijo - Daporijo Arunachal - 10 10 Pradesh Total - - 2775 1741
The maintenance and upkeep of the REW is the responsibility of the BSE, who ensures that the stock of spare parts is sufficient. The VEEC directly monitors the BSE carrying out his duties with regards to the REW, with the assistance of the field centre. Staff from the solar section in Tilonia may also intervene or make suggestions if necessary on occasions when they visit the villages. Batteries are not stored at the REW, since they should not be allowed to discharge, but can be obtained from Tilonia when required. The REWs can become in the future a training ground for further BSEs selected by the VEEC to service the area.
329 Chapter 9. Capability Building at the Barefoot College
9.2.3.5. Quality Control Quality control checks are carried out after a circuit is completed at a REW. A battery at various voltages is simulated using a DC power supply and multimeters are used to check that the charge controllers operate with the correct disconnect and reconnect levels. The operation of lamp circuits is similarly checked.
Figure 9-4: Quality Control at the Barefoot College
If the circuit fails to operate as expected, a checklist for fault-finding is used to identify the circuit component that is most likely to be causing the problem, which may have been caused either by damage to the component during manufacture or a bad solder joint causing a poor connection. Even without referring to the checklist, the experienced BSEs know how the circuits should behave, down to the voltages to expect across various points in the circuit when it is operating in a particular mode. In the event of a circuit assembly mistake or a fault they are often able to predict which components are most likely to cause the fault, and use a multimeter to determine if it is operating correctly or not. Quality control could be more rigorously applied at different steps throughout the manufacture of the circuits, rather than only at the finish of production, in order to save time and prevent wastage of components.
330 Chapter 9. Capability Building at the Barefoot College
9.2.3.6. Financing Most of the systems installed by the Barefoot College are done so within a project sponsored by an external donor. SWRC has received MNES subsidies and loans and the support of NGOs including the EU, UNDP and PLAN International for solar programmes. The donor usually covers the cost of all of the PV systems and the establishment of Rural Electronics Workshops. In some cases, they also cover the cost of the training of the BSEs and the VEECs. Where the donor does not cover all these costs, the Barefoot College may contribute to covering the initial expenses. Once the equipment and training has been provided and the institutional arrangements are in place, however, the ongoing costs of maintenance and the long term replacement of batteries and modules is covered by the users. Users are trained by the BSEs to understand the need for regular upkeep of the systems, battery replacement and minor repairs and have agreed to meet the cost. In consultation with the users and consideration of what they could afford, the VEECs decide upon an appropriate upfront payment. On average, this amount is 400Rs/per lantern - approximately 10% of the price of the upfront cost of supplying the lantern to the user. The upfront payment is deposited into the committee’s bank account. A further amount per month (on average 25Rs for a lantern) is also paid into the account by each household with a lantern. The solar account is used to cover the cost of new batteries and modules and spare parts required to keep the lanterns running. The account is also used to pay a stipend to the BSEs. The amount of the stipend is decided by the VEECs according to what the villagers are willing and able to pay, and is almost always equal to the minimum legal wage of 2200 Rs per month. This stipend has proved, for the local, disadvantaged BSEs to be sufficient to retain almost all of them in their jobs. The cost of repairs of lanterns is estimated by the head of the solar section in Tilonia to be about 50Rs/year on average for spares, excluding batteries. Lanterns require a new battery on average after 2 years, which costs about 500Rs in 2006. Fixed systems cost about 100Rs/year in spare parts (excluding batteries) and also need distilled water and fuses regularly. The battery in a fixed system usually needs to be replaced after 5-6 years (the company gives a 2 year warranty) and the price varies according to the system size. The cost of maintenance for a system is compared with the amount typically contributed by the users in Table 9-4.
331 Chapter 9. Capability Building at the Barefoot College
Table 9-4: Costs and Contributions of Solar Lantern Maintenance in a Village with 100 Lanterns and 2 Barefoot Solar Engineers. Details of costs/contributions Amount (Rs) Stipend for 2 BSEs@ 2200 Rs for 2 years 8 800 Parts & repairs for 100 lanterns @ 50 Rs for 2 years 10 000 Battery replacement for 100 lanterns @ 500 Rs 50 000 Total costs: 68 800 Rs Upfront contribution of 100 households @ 400 Rs 40 000 Monthly contribution of 100 households of 25 Rs for 24 months 60 000 Total contributions: 100 000 Rs
Detailed records kept from a SWRC project in Leh indicate that the cost of spare parts replaced and repairs and maintenance carried out on 532 units in 25 villages in 1993-94 was approximately 100 000 Rs (US$3125), 189 Rs per system or 94 Rs/year. The cost included the early failure of six batteries (21 000Rs) and the breakage of four PV modules (30 000Rs). The systems had been installed between 1989 and 1993, with an average age of 1.5 years (calculated from table 1 in Roy, 1996, p 47). The repairs included the replacement of the six batteries and four PV modules as well as inverters, controllers and lamps, and also included a fee to cover travel expenses. Appendix 10 details the repairs carried out on 230 of the systems installed in Leh. The stipend for the 12 BSEs for 24 months was 105 600 Rs, making the total outgoing costs approximately 200 000 Rs. The projected contribution from the households of around 20 Rs/month/unit, gives a total of 255 360 Rs over the period. Despite the early battery and module failures and the relative inexperience of the BSEs in this area, if the recovery of fees was 78%, there would still be sufficient money collected to cover all the repairs without dipping into the initial contributions which are generally reserved for replacements. The Barefoot College relies on project-based initiatives to fund the set up of REWs and the training of BSEs. The college is unsustainable without donor support. It is selective in it’s acceptance of donor funds, since the mobilisation of the communities and the full financial and decision making powers of the committees are central to the approach.
9.2.3.7. Maintenance and Monitoring of the Systems Each remote BSE has a stock of spare parts. The BSE has the responsibility to maintain systems that are the property of the field centre or community, such as night school systems that are visited twice a month by the BSE. The VEEC and the elected members of the student parliament are charged with ensuring that the systems are performing as expected and the BSE is doing her/his job. The BSE performs routine maintenance such as cleaning the panels and topping up the batteries with distilled water. If people have a problem between visits, they can send a message to the BSE who comes to repair the system. If a BSE cannot fix a problem, the component can be returned to Tilonia for repair. The BSE will usually replace the component with one from his/her stock. There is a 2 year installation guarantee on parts from SWRC Tilonia.
332 Chapter 9. Capability Building at the Barefoot College
Figure 9-5: BSEs Performing Repairs in the Field
9.2.3.8. Supply of Materials, Equipment and Technology Equipment and materials for BOS manufacture, solar panels and even cast plastic solar lantern bodies are relatively easy to procure in India, with large numbers of manufacturers able to produce these items at low cost. Since the Barefoot College does not have design capabilities for the circuits they manufacture, consultants with experience in electronics for photovoltaics have been engaged to design the circuits and make the changes according to their requirements. Consultants with the capabilities to provide circuit designs are also readily available in India. Simplicity, understand- ability and repair-ability are the primary design criteria. While they do not have the capabilities to design circuits, the senior members of the solar section at the Barefoot Collage have sufficient understanding of the circuit, how it works and the functions of the components to be able to engage with consultants on the improvement of the circuit design. The first checklists for quality control used in manufacturing and fault-finding used in the field were also developed by consultants, but these have since been expanded and improved by the Barefoot College. The Tilonia Campus has a post office through which information can be exchanged with literate members from field centres, a telephone exchange, internet facilties and two 4WD vehicles which are used to make visits to field centres. It is within 5 minutes walk from a main road and 15 minutes walk from a train station on the line to Jaipur (the capital city of Rajasthan). The staff are therefore able to access the information they need to negotiate over the purchase of materials and over technology development.
333 Chapter 9. Capability Building at the Barefoot College
9.2.4. Impacts, Sustainability and Replicability Since 2002, 260 SHSs, 500 solar lantern systems and two 2.5 kW community PV systems have been installed within India using the Barefoot approach. The approach has resulted in well maintained PV systems, delivered via low cost, sustainable financial and organizational arrangements. The manufacturing costs are comparable with commercially manufactured systems, but the project administration and training costs are much lower in this community-run approach, as the costs of urban consultants are largely avoided.
Livelihood Impacts The Barefoot College programmes have provided electricity to remote villages, including lighting for night schools and clinics, and reduced expenditure on alternative light sources. Table 9-5 provides an idea of the scale of the impact of the Barefoot College solar programme. Of course, the figures would be much larger now.
Table 9-5: Barefoot College Electrification Statistics 1989-2002 Total no of fixed Solar Photovoltaic (SPV) units installed 4947 Total number of Solar Lanterns Fabricated and distributed 3726 Total number of Night Schools Solar Electrified 521 Total number of States/ Districts where fixed SPV units & Solar Lanterns are 29 Districts in 16 installed States Total number of Barefoot Solar Engineers trained 209 Total number of villages covered 531 Total number of houses solar electrified 8003 Total number of people benefited by solar electrification (approx. figures) 79,590 Total kilowatts of Solar Energy capacity 273 KWp Source: (Roy & Joshi, 2005) Beyond the benefits of electricity, employment is created through the Barefoot approach, and local skills, confidence and empowerment are developed, including community organizational capacity and the ability to influence authorities and access institutions and opportunities. In particular, employment has been provided for women and the disadvantaged. Through some of the programmes, people have been trained in income generating activities that are made possible by the provision of electricity via the community PV system and solar lanterns. In particular, extension of the daylight hours provides opportunities for women to earn income through the production of handicrafts as well as traditional activities such as spinning of wool. The provision of mechanised shearing equipment and electricity for churning milk has provided opportunities for increased agricultural productivity. Technical and organisational skills as well as social capital and networks have been built through the programmes. Individuals and communities have been empowered in the use of technology and become self-reliant in the management of the technology. The social capital built through the process has increased the capacity of communities for their own development, and the political influence of communities, and in particular of the most disadvantaged individuals within communities.
334 Chapter 9. Capability Building at the Barefoot College
Political Empowerment The Barefoot College has been able to impact policy at local and macro levels. In the 1995 Panchayat elections for Silora Block, 40% of the elected leadership were either staff of SWRC or committee members within field centres, giving the SWRC influence in community politics and priorities. The Rajasthan State Government now runs various education programmes based on Barefoot principles, including the Lok Jumbish community education and the Siksha Karmi night school programmes (Chasi & Mussukuya, 1998). A community-based water pump maintenance system has replaced the three tier system which was controlled by the District level and SWRC is the Official training centre for water pump mechanics in the state. The Barefoot College has always had a policy of working with government and local bureaucracy in order to protect the rights and entitlements of the rural poor, and has a proven track record as a source of innovative programmes, so is difficult for the authorities to ignore. They also have strong community links and can hence advise authorities and facilitate communication with communities. The network also gives some political security to village organizations and field centres by providing financial and political support at critical moments.
Replicability Many field centres (SWRC outposts that usually host REWs, for example) have become independent NGOs, running their own projects and securing their own sources of funding, but affiliated with the Barefoot College through the Sampada network and can rely on Tilonia for shared resources. 1000 SHSs have also been installed outside of India since 2002, in countries such as Afganistan and Ethiopia, with BSEs trained at the Barefoot College Tilonia campus. The Barefoot College believes its methods will work in any poor rural community anywhere in the world, providing the community is committed to the importance of developing and depending on their own knowledge and skills (UNDP, 2003). The SWRC structure in India benefits from its influential and well-connected founder, Bunker Roy. The importance of such a personality in influencing authorities and accessing funding may be difficult to compensate for in his absence. Other obstacles to replicating the approach include the long time required to build up a stock of knowledge and expertise, trust in a community, the financial, technical and political assets to provide a support network for communities and to overcome resistance and entrenched power structures. There is also a need for significant organisational size to implement an integrated development approach and to access consultants and materials cost- effectively. In communities that have been oppressed, there may also be a need for an outsider to catalyse action.
335 Chapter 9. Capability Building at the Barefoot College
9.3. Capabilities at the Barefoot College
The following sections detail the development of capabilities at the Barefoot College. The remainder of this chapter analyses the learning processes that have enabled the Barefoot College to build these capabilities and the factors that have facilitated or constrained these learning processes. The capabilities for the manufacture and commercialisation of PV system components in the Barefoot College have developed over time, since a PV system was first installed at the medical clinic of the main Tilonia campus of the Barefoot College in 1984. The responsibility for maintenance of the first system was given to Bhagwat Nanden, who is now the head of the solar section. He, however, had no experience of PV systems, nor did anyone else in Tilonia village. When there was a fault with the system, technicians were called in to fix the problem, and the Tilonia staff slowly learnt how to maintain the system. The system was later extended and calling in a technician from outside was becoming such a frequent occurrence that a solar workshop was set up on the campus.
9.3.1. Production Capabilities Beginning with no knowledge of PV systems in 1984, the Barefoot College has learnt to cost effectively manufacture charge controllers and lamps of appropriate quality for small PV systems through an organisation made up of BSEs at remote REWs.
Cost Table 9-6 lists the costs of materials and labour and the price for the Barefoot College products. The total costs of the materials are given, and the breakdown of these costs is detailed in Appendix 11. The BSEs in a field centre are paid 15 Rs for their labour input to a lantern circuit, and 10 Rs for a charge controller or lamp circuit. The funds for setting up REWs are provided through donors. The data given in Table 9-6 indicates that there is sufficient margin for transport and REW operating costs to ensure the financial sustainability of the manufacture and training facilities.
Table 9-6: Costs of Materials and Prices of Barefoot College Products (Rs2) Product Cost of Labour Transport & Price Materials Operating Costs Solar lantern with 10Wp PV 3735 15 450 4200 (US$107) module and 12V 7Ah gel battery. Charge Controller 12V 4 A 286.45 10 153.55 450 (US$11) Lamp 12V, 1Amp 372 10 68 450 (US$11)
2 US$1 = 39 Rs 336 Chapter 9. Capability Building at the Barefoot College
The prices that the Barefoot College charges for its components and systems are in line with those set down by MNES for subsidy qualification. While some commentators believe that the MNES price limits are too low and will result in the use of low quality components (Hirshman, 2006), others believe that there is a need to provide systems of an appropriate cost/quality ratio to poor users. The Barefoot College has been able to comply with these price limits, and monitoring indicates that the systems they provide are of appropriate quality.
Quality It is not possible to directly compare the quality of the balance of systems components manufactured by the Barefoot College BSEs with those manufactured by other enterprises, because they have not been used within projects where components manufactured elsewhere have been subject to the same installation and maintenance routines. The quality of the energy service delivered by the Barefoot College, however, can be compared with that provided by PV systems in other projects and programmes. Nieuwenhout et al.(2004) carried out a survey of solar home system status from 19 sources in the literature, the results of which are displayed in Table 9-7.
Table 9-7: Overview of status of solar home systems from 18 different sources Partially Not Age at Time Working Working Working of Survey Activity Country Well [%] [%] [%] [months] World Bank Indonesia 77 23 0 27 Sukatani demonstration project Indonesia 11 130 Sukatani demonstration project Indonesia 3.2 108 Sukatani demonstration project Indonesia 100 0 0 12 Retail sales Swaziland 73 17 10 31 Retail sales Swaziland 77 36 ESMAP-programme Kenya 77 2 21 18 Test marketing small batteries Kenya 56 33 11 >6 Ramakrishna Mission (survey) India 2 30 SWRC India 71 27.5 1.5 29 Urjagram project MNES India 42 12 ASTEC project India 90 5 5 18 PLAN International Guatemala 53 4 43 36 SEC retail sales Kiribati 10 48 SHS project Camarones Chile 41 59 0 ESKOM programme RSA 46 24 30 36 Pilot dissemination phase Tunesia 34 32 34 84 Intermediate phase Tunesia 6 48 Average 63 23 15 44
Source: (Nieuwenhout et al., 2004)
337 Chapter 9. Capability Building at the Barefoot College
The Barefoot College (SWRC) systems, which were installed in extremely isolated villages in the Ladakh region of India (Jacobson, 1999; Maithel, 1998), far from the Tilonia campus, performed favourably compared with the other systems in the survey. It can be concluded that in the case of the Barefoot College, the production capabilities required to manufacture charge controllers, lamp inverters and solar lanterns of an appropriate quality and cost have been developed in a rural context.
9.3.2. Innovative Capabilities The Barefoot College has built the capability to innovate in the development and adoption of new products, quality control routines and organisational structures.
New Products The circuit and lantern bodies manufactured at the Barefoot College have improved over time in response to feedback from the BSEs and users and, in the case of the latest changes, in order to comply with standards required for eligibility for subsidies through the Government of India. While the BSEs do not have the capabilities for design, they have sufficient understanding of the operation, manufacture, failure modes and maintenance of the circuits to collaborate with consultants. The head of the solar section believes the 1992 certification process was useful and has resulted in greater reliability and ease of repair of the systems, with 5-6 iterations of adjustments to the circuits before certification was achieved. Correction was made to the ballast frequency and a more robust circuit design was achieved. Each time, consultants worked with the Barefoot College to improve the design. New circuits are simpler than the previous ones, improving reliability, fault-finding and repair, as well as being more robust due to better fusing. Table 9-8 lists some of the adaptations to the solar lantern design since the first lanterns were made in 1996. The old lanterns are now slowly being replaced with new ones.
Table 9-8: Adaptations to the Barefoot College Solar Lantern Design Old New Reason for change Steel body Plastic body Plastic is more breakable, but still very strong and lighter and cheaper. Glass lamp shield Clear plastic lamp shield Clear plastic breaks less often. Better grip on handle Ergonomics Charging time 5 hours Charging time 6 hours Longer use of light. If switch was left on after battery If the battery is flat and the switch Improved battery life. ran flat, when lamp was on, the lamp stays off while recharging, light would come on, lantern is charging. and battery would not charge. Short CFL life due to wrong Frequency corrected. Improved CFL life. frequency.
338 Chapter 9. Capability Building at the Barefoot College
Figure 9-6: (a) Old and (b) New Solar Lanterns Produced at the Barefoot College
(a) (b)
The current controller circuit is also simpler than the previous one, and it is therefore easier for an engineer to remember fault finding and repair of the circuit. The old circuit was vulnerable to damaged components if a wrong connection was made, with ICs, transistors and diodes at risk. Now a fuse protects the sensitive components in the circuit. Dasrath, a BSE master trainer who was trained in 1992 says if the new circuit is not misused, it can last for 10 years.
New QC Routines A fault-finding checklist for quality control and the identification of faults during maintenance was initially developed by foreign consultants, and has been improved through the experience of master trainers.
New Organisational Structures There is also constant innovation in the organisational aspects of the delivery of PV systems to the end user. The Barefoot College engages with many donor NGOs who bring new ideas about training, monitoring and financing, with many communities who have their own ways of doing things, and with different sections within the Barefoot College who bring innovative ideas from their experiences. There is flexibility in the delivery of Barefoot programmes. No two are the same and each one builds on the experience and knowledge gained from the previous ones. This flexibility and the constant questioning of the programmes enables organisational innovation.
339 Chapter 9. Capability Building at the Barefoot College
9.3.3. After-Sales Service Capabilities The Barefoot approach has been shown to result in reasonably low rates of system failure. 200 of a total 5000 SHSs installed in Ladakh between 1989 and 1999 were surveyed by Jacobson in 1999 where he found that 71% of systems were fully functional, 27.5% were partially functional and only 1% were non-functional (Maithel, 1998). The systems were 35Wp SHSs with a 75Ah battery protected by a charge controller, and ran two 9-11W fluorescent lamps for three hours per day in winter. 1000 of the earlier system installations were carried out by SWRC, and many of the SWRC systems were still operating with their original batteries in 1999, when some of the systems were ten years old. Jacobson (1999) attributes this performance to the community-led maintenance infrastructure and extensive hands-on training program of the SWRC, the financial accessibility of the scheme for the villagers, and the compliance with the MNES quality standards. The good performance of the SWRC systems created a large demand for PV systems and prompted further installations in the region. However, 3500 of the 5000 systems installations in Ladakh (70%) between 1996-1999 (within three years of the survey), have been executed outside of the SWRC methodology. It has also been reported that government engineers in Ladakh have approached BSEs to have their systems repaired (Mitra et al., 1999).
“the focus is on installing new systems; little money or time is allocated to developing a maintenance infrastructure. This approach will likely result in future problems, as systems that are not maintained will fail prematurely.” (Jacobson, 1999).
Maithel (1998) commented that: “While the quality of installation and reliability of electronic circuits was found to be better in the PDD/JAKEDA systems [a project where a commercial system supplier installs the system], the SWRC systems had the advantage of continued and reliable maintenance support. In the long term, the performance of the SWRC systems will be better because of regular servicing… It was observed that the …PDD/JAKEDA batteries had started corroding due to misuse.” (Maithel, 1998)
The author accompanied BSEs to repair 11 SHS systems, and in every case, the system was restored to full working condition, even if it required replacing the charge controller, which was then taken to a rural electronics workshop (REW) and successfully diagnosed and repaired. Familiarity with the failure modes of the circuits is an indication of the ability of the BSEs to maintain the products. BSE Dewa Ram says that the most common failures are transistors in the circuits or the battery life is finished (after about 2 years for a solar lantern). A short circuit can also cause a transistor to burn. Bagwat Nandan identifies D1 – 4007 and the 2955 transistor in the lamp circuit as a common cause of faults. He says they burn out when there is an incorrect connection made. If the 2955 transistor fails, the lamp still works, but the battery doesn’t charge.
340 Chapter 9. Capability Building at the Barefoot College
Local manufacture of electronic components for PV systems clearly has the potential to enhance technical capabilities in maintenance. Where a community-managed structure is in place, the demystification of the technology resulting from local manufacture and local control over the design and implementation of the structures for its use can result in quality delivery of PV services overall. Therefore the needs of local people can more effectively be met at a price they can afford, while also being more sustainable.
341 Chapter 9. Capability Building at the Barefoot College
9.4. Analysis of the Case Study using the Framework
The preceding sections of this chapter have described the development of capabilities by the Barefoot College since the first PV system was installed at the Tilonia campus in 1984. The framework developed in chapter 5 will now be used to identify the factors which have supported or constrained different types of learning and to analyse the success of the strategies that have been used to build capabilities and overcome deficiencies in the technological system.
9.4.1. The Indian Technological System for Small Scale PV Manufacture Figure 9-7 shows the Indian technological system for small scale PV manufacture. Functions where the technological system has been effective in supporting enterprise-level learning are represented by red highlighting. Blue is used to indicate areas where the technological system has partially supported learning, while functions left un-highlighted have not been fulfilled by the system. In the following sections, each of the functions of networks is discussed, and the ways that the institutional environment has influenced the operation of networks are identified.
Figure 9-7: Indian Technological System for Small Scale PV Manufacture
TECHNOLOGICAL TRAJECTORY
influence perceptions of the provide direction for technological paradigm search
Indian Technological System for Small Scale PV Manufacture Networks
OTHER ACTORS influence policy and institutions
MARKET & NON-MARKET INTERACTIONS influence operation of markets influence connectivity INSTITUTIONS
knowledge creation & exchange resources for production and innovation investment opportunities alter incentives to invest and improve alter allocation of resources direction & incentives
ENTERPRISE
342 Chapter 9. Capability Building at the Barefoot College
Investment Opportunities As described in the background to this case study, there are a large number of unelectrified or partially electrified villages in India which have little hope of gaining access to either grid electricity, or being able to afford a PV system, even under the subsidy schemes offered by MNES. Fossil fuel subsidies have also acted to reduce the uptake of PV for lighting and water pumping. Nevertheless, a large number of new PV SHS are installed in India each year, the majority of which are subsidised by MNES, and installed either through MNES programmes or other NGO projects. Despite the faulty application of diesel and kerosene subsidies and uncertainties in the application of subsidies, programmes and projects have provided a fairly stable long-term domestic SHS market over the past 25 years and a reasonable level of confidence in the technology on the part of users. Small scale PV manufacturers may take advantage of these market opportunities provided that they are given access to these projects and subsidies. Investment opportunities are therefore coloured blue in Figure 9-7.
Resources for Production and Innovation Manufacturing equipment and materials such as soldering irons, transformer winding tools and electronic components; and PV system components such as CFL tubes, batteries, PV modules and BOS; are manufactured and available in the largest urban centres in India (Kumar et al., 2000). The government has also redistributed national resources toward the PV manufacturing industry by preferentially supporting domestically manufactured PV system components over imported ones through import duty and sales tax reductions. The human capital for PV commercialisation is also available in India. There are many people in India with electronics design and manufacture knowledge and skills and there have been many people trained in PV installation and maintenance through various PV projects. However, design and manufacturing skills are concentrated in urban areas. Because there are such a large number of dispersed and isolated unelectrified villages, and there has been insufficient emphasis on capacity building in MNES programmes, there are many villages without trained technicians. In spite of recognition of the importance of small and micro enterprises in India, a vast number of these types of enterprises also do not have any access to assistance such as finance, business development services or technical training. Infrastructure is also lacking in rural areas of India, hampering the purchase of materials for production, sales and marketing activities and after sales service. While many poor villages in India cannot easily access materials, markets, expertise or other kinds of support for production and innovation, the skills, equipment and materials are available in urban areas and resources are therefore shaded blue in Figure 9-7.
343 Chapter 9. Capability Building at the Barefoot College
Knowledge Creation and Exchange Both Indian government and NGO PV programmes have attempted to increase the level of knowledge about PV systems through awareness-raising and training, but these efforts have placed insufficient attention on participation and capacity building, which have the potential to increase knowledge flows. This has been particularly true in the government programmes, which have been primarily target-driven, while the MNES subsidy scheme has not made any provision for awareness-raising, financing or training. The lack of coordination between ministries has prevented knowledge creation and exchange through inter-ministry interactions. The poor linkages between ministries, various levels of government administration, development agencies, contractors and rural users have reduced flows of information about the needs of users and stifled the development of appropriate technology and institutional structures to support it. Poor connectivity is exacerbated by the hierarchical structure of institutions in India, which arouses distrust in the poor, including in the Panchyat system, which is often hijacked by locally powerful people. Indian market-based diffusion of PV systems has also been unable to effectively provide routes for users, installers, technicians and manufacturers to interact. This is the result of the remoteness and low concentration of users, which makes the cost of information, maintenance and regulation high. Failure to act on programme evaluations has also limited the dissemination of knowledge about successful programme design, while small R&D budgets and restrictive specifications for subsidy eligibility have prevented the development of appropriate technology. On a positive note, however, Indian testing facilities for PV system components provide technical support for manufacturers such as guidelines for compliance and design assistance when components fail. The large number of actors in the technological system including manufacturers, system integrators, retailers and users provides opportunities for knowledge creation and exchange, provided that the institutional and infrastructural obstacles can be overcome. On balance, knowledge creation and exchange is not well supported by the technological system and is left unshaded in Figure 9-7.
Incentives and Direction for Search MNES subsidies require the certification of components, providing a strong incentive for manufacturers to comply with the standards set down, while the assistance provided by the testing facilities provides direction for manufacturers to improve in order to reach the standards. The certification procedures do not, however, provide incentive or direction to exceed these standards. There are a limited number of designs approved and there is little incentive in large projects for manufacturers to adapt their designs to suit local requirements.
344 Chapter 9. Capability Building at the Barefoot College
With respect to government programme implementation, there is emphasis on the number of systems installed, and in the case of the Renewable Energy for Rural Livelihoods project, on dollars earned through the generation of electricity. There is little incentive to improve the provision of electricity to the poorest via innovative institutional arrangements. Poor connectivity between manufacturers and users prevents market incentives related to user requirements informing innovative efforts by manufacturers. Subsidies for fossil fuels also reduce the market incentives for the provision of widespread infrastructure for photovoltaic systems. However, in recognition of the strong incentives and direction provided by the certification requirements, this factor is shaded blue.
9.4.2. Learning at the Barefoot College Starting with the installation of a PV system at the Tilonia Campus in 1984, the Barefoot College has progressively built the capabilities to install, maintain and manufacture PV system components and has successfully managed the delivery of thousands of PV systems that have performed well in even the remotest and harshest conditions in rural India. Figure 9-8 shows the part of the framework concerned with learning at the Barefoot College (a) when the solar programme first began in 1984, and (b) when the case study was conducted in 2006. The capabilities and factors that influenced learning in each case are coloured red where they strongly supported learning, blue where they were partially functioning and not highlighted where they were insufficient. The following sections analyse the importance of, progress in relation to, and factors influencing each type of learning.
Figure 9-8: Learning at the Barefoot College (a) in 1984, and (b) in 2006
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION incentives & resources PRODUCTION direction INNOVATIVE incentives & CAPABILITIES for search CAPABILITIES resources ROUTINES Learning by Doing Learning by Searching new production techniques DOING R&D Improvement Capabilities
RECONFIGURATION
(a)
345 Chapter 9. Capability Building at the Barefoot College
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION incentives & resources PRODUCTION direction INNOVATIVE incentives & CAPABILITIES for search CAPABILITIES resources ROUTINES Learning by Doing Learning by Searching new production techniques DOING R&D Improvement Capabilities
RECONFIGURATION
(b)
9.4.2.1. Learning by Doing The Barefoot College began with no capability in PV production, as illustrated in Figure 9-9a, but is now able to produce PV system components that comply with MNES certification requirements and which perform well in the field. This section examines the extent to which learning by doing explains this learning and the role of the technological system for PV in India in supporting or constraining learning by doing in the enterprise.
Figure 9-9: Learning by Doing
production resources production resources investment opportunities investment opportunities
PRODUCTION PRODUCTION CAPABILITIES CAPABILITIES incentives & incentives & resources ROUTINES resources ROUTINES Learning by Doing Learning by Doing
DOING DOING
(a) (b) Learning by doing has occurred as BSEs have accrued experience in the manufacture of circuits, familiarity with components in the circuit, experience in the use of the fault-finding checklist and in the installation and maintenance of systems.
Routines Quality control routines at the Barefoot College ensure that components work correctly in the first instance, but also promote learning, since the testing indicates where mistakes have
346 Chapter 9. Capability Building at the Barefoot College
been made, allowing the BSEs to adjust their work accordingly. The maintenance routines and the regular meetings between BSEs, field centres and section heads ensure that information about failure modes of circuits contribute to learning. Through training routines, the knowledge of the experienced BSEs is passed on to the trainees. The Tilonia workshop has become a training centre and a critical mass of people have learnt about PV technology. While the Barefoot College initially had no established routines for component testing or training, as illustrated by no shading in Figure 9-9a, routines now strongly support learning throughout the organisation, as indicated by red shading in Figure 9-9b.
Barefoot College Investment Opportunities The schemes administered by the Rajasthan state government and many of the other state nodal agencies are poorly resourced and unable to provide a reliable service via PV systems. Because their systems and components have been awarded the certification required for eligibility, the Barefoot College has been able to take advantage of MNES subsidies and been involved in government programmes and those of international NGOs. The Barefoot College therefore receives more requests for implementation of solar programmes in villages than it has the capacity to implement. There is a large potential market available to the Barefoot College, which it can continue to access as long as it can maintain the donor support that it relies on. While the Barefoot College had good access to potential markets for PV through their work in communities prior to the initiation of the solar programme, as illustrated by blue shading in Figure 9-9a, as their reputation has grown, these opportunities have become vast, as indicated by red shading in Figure 9-9b. Successful projects provide a positive feedback loop for donors and clients.
Barefoot College Access to Resources While resources for production are difficult to access in remote Indian villages, due to their small scale, inadequate infrastructure, information and connectivity, the Barefoot College has used the influence and connectivity of its network to access donor finance, materials and equipment for production at bulk purchasing rates and to engage consultants to provide designs and quality control and maintenance checklists. The Barefoot approach has also been able to provide training for the villages within which they work. The provision of resources, however, relies on accessing new donor funds in order to reach new villages. While the Barefoot College did not have good connectivity with PV suppliers, donors or technology developers in 1984 (Figure 9-9a), improved linkage capabilities now enable them to access these resources, as indicated by blue shading in Figure 9-9b.
347 Chapter 9. Capability Building at the Barefoot College
9.4.2.2. Learning by Searching The Barefoot College does not carry out the formal R&D required to introduce new products, but engages consultants to carry out this work. While it does not have the capability for design, the Barefoot College carries out the research to supply locally appropriate specifications and field feedback to consultants which are the basis for new circuits and quality control routines.
Figure 9-10: Learning by Searching
innovation resources innovation resources investment opportunities investment opportunities
COORDINATION & COORDINATION & INTEGRATION INTEGRATION
incentives, incentives, direction INNOVATIVE resouces direction INNOVATIVE resouces for search CAPABILITIES & direction for search CAPABILITIES & direction for search for search
Learning by Searching Learning by Searching
R&D R&D
new production new production RECONFIGURATION RECONFIGURATION technique technique
(a) (b)
Innovation Resources Consultants that can design the circuits produced by the Barefoot College are available in India. State testing facilities in India have also supplied the resources to support the Barefoot College in improving the quality of their products. Because the Barefoot College was not initially able to connect with these organisations, innovative resources are left unshaded in Figure 9-10a, but they now have good access, indicated by red in Figure 9-10b.
Coordination and Integration Because production, installation and maintenance of the balance of systems components are carried out by each Barefoot Solar Engineer, knowledge gained through testing of circuits and maintenance and repairs of systems has been able to be used to improve the production techniques. This knowledge has also been employed to inform investment in R&D, since it is passed on to decision makers through frequent interactions between BSEs, meetings of field centres and solar section staff. Ongoing interactions between BSE, users, village energy committees and the college ensure that the learning of the group in relation to the manufacture and use of the technology and organisational arrangements is shared and the implementation can be tailored to local requirements. Because the programmes such as PV and health are usually integrated, each of the committees and participants in each programme may also learn from each other about the technologies and equally importantly about organisational approaches. The solar programme is also therefore more likely to be incorporated into important community decision making processes concerning education and health. Since each project is negotiated separately, the
348 Chapter 9. Capability Building at the Barefoot College
Barefoot College also has opportunities to learn by interacting with the many donor NGOs they work with. Learning by searching through external interactions has relied on the research and production capabilities of the Barefoot College and the ability to link with technical consultants. Because the Barefoot College relies on consultants with whom they only have infrequent contact to suggest solutions, it is likely that the technology may not be improved on the basis of field feedback very often. Through it’s other activities, the Barefoot College had good (blue) capabilities for coordination and integration when it began the solar programme, but experience has improved these capabilities in relation to the solar programme, indicated by red in Figure 9-10b.
Reconfiguration Although the REWs are equipped such that they could manufacture a variety of BOS circuits, the training given to the BSEs is specific to the charge controllers, lamps and lanterns that they have initially manufactured and installed. Upgrading of the products therefore requires retraining of the BSEs. New designs have instead been implemented by introduction into villages obtaining PV systems for the first time. While reconfiguration in production is limited by the need to upgrade the technology through the whole of the supply chain, it is possible to incrementally upgrade new and existing villages. Because each Barefoot College PV programme involves negotiation with the community, reconfiguration of the project implementation is central to the barefoot approach. Reconfiguration is therefore shaded blue.
Incentive and Direction for Search Incentives to improve the quality of designs have been provided institutionally by the MNES certification requirements for eligibility for subsidies, but these requirements also do not provide incentive to move beyond the minimum standard. However, the organisational structure of the Barefoot programmes provides incentives for improvement to both products and programme implementation. Because the Barefoot technicians come from the village within which they work, the incentive to carry out the job well and earn the respect of their community is high. The village committee, which pays the technician, and can carry complaints to Tilonia or the field centre, also provides incentives to the BSE to carry out the job properly and to improve the performance of the technology. Many of the changes to the circuit design have been initiated through a feedback from the users and the technicians in the field. In the absence of competition, however, the incentives for improvement to the product design may be reduced. The local decision making process also promotes ownership of the process, so the village committees themselves have strong incentive to carry out the administration well. Because they are poor, uneducated and disadvantaged, conflicting incentives to leave their village in search of a better job are reduced.
349 Chapter 9. Capability Building at the Barefoot College
9.4.2.3. Learning by Interacting The locally based technician, decision making, fee collection and monitoring of PV systems within the Barefoot approach can operate within resource and infrastructural constraints primarily because of the support gained through interacting with the Barefoot College network.
Figure 9-11: Learning by Interacting
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
(a)
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
(b) The Barefoot College main campus in Tilonia and the field centres that have become independent take care of all the challenging investment functions for the entire network centrally. Start-up funds for projects are sourced, and the solar section and accounting departments have the capability to plan and manage the financial and organisational aspects of projects, including investments in infrastructure such as REWs and in training the BSEs and communities. From the main centres, product designs and quality control procedures are also provided to REWs. The network is also able to provide support at critical moments, to mitigate the risks taken on by the community and to support the BSE.
Coordination and Integration of Learning Interactions with suppliers and technology developers are coordinated with the needs of the manufacturers as a result of feedback from learning in production, quality control and
350 Chapter 9. Capability Building at the Barefoot College
market feedback. This coordination depends on the institutionalised linkages between actors in the Barefoot College programmes. These capabilities have become well developed since the solar programme began, as indicated by red on the diagram relying, whereas they relied initially on capabilities developed in other programmes (blue).
Linkage Capabilities The local base of the Barefoot technicians facilitates close interactions between users and technicians who install, maintain and repair the systems. The technician may also be the one that manufactures the BOS, or if not, has frequent interactions with the REW which does. The connectivity is therefore good and the transaction costs are reduced since there is little transport required and the trust between community members reduces the risk of the transaction. High connectivity does not arise just from markets, but from trust. The user interactions with the barefoot technician are peer-to-peer, so there is likely to be a good connection and good communication of advice on how to use the system or feedback on the performance of the system. There is likely to be a good understanding of the rights and responsibilities of the users and technicians, since it is the local VEEC that decides on the rules and enforces them. In addition, the Village Committees which are explicitly democratic and must involve women, have a strong connectivity with the community and are based on trust. While the Barefoot College can benefit from these small scale aspects, local aspects of their operations in villages, their linkages with donors and suppliers is primarily the result of the large scale provided by the Barefoot College network of many villages, workshops and technicians. The SWRC network also enables communities to influence the government rural electrification approaches and access large NGO funds. Start-up funds for field centres and other village initiatives, such as night schools and health centres, cover the cost of system hardware and the training of the BSEs, enabling installations to occur at a scale and density so that Barefoot technicians can easily access systems. The Barefoot College began with strong community linkages, but did not have established links to technology developers, suppliers and PV donors, as illustrated by blue shading in Figure 9-11a. Linkages are now well established both upstream and downstream, indicated by red shading in Figure 9-11b.
351 Chapter 9. Capability Building at the Barefoot College
9.5. Conclusion
In this chapter, the factors in the Indian technological system that have supported or constrained capability building at the Barefoot College have been identified, and the success of the capability building strategies employed has been analysed. The use of the framework has in particular drawn attention to the path dependence of learning, the importance of internal innovative capabilities for PV manufacturers and the value of networks in facilitating interactive learning and access to resources and institutions in the technological system. The discussion chapter of this thesis will refer to the analysis of this case study, the previous case study on Grupo Fénix and the pre-existing literature on small scale PV manufacture in order to suggest typical barriers and appropriate policies and interventions to support small scale manufacture.
352 Chapter 9. Capability Building at the Barefoot College
References
Administrative Reforms Commission (2001), Seventh Report on Panchayati Raj in Rajasthan, Government of Rajasthan: Department of Administrative Reforms, Jaipur, India. Bery, S., Gupta, D.B., Krishna, R. and Mitra, S. (2003), The Nature of Rural Infrastructure: Problems and Prospects, Version of the introductory chapter of the India Rural Infrastructure Report, National Council of Applied Economic Research (NCAER). Bhalla, A.S. and Reddy, A.K.N. (1994), The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme, Intermediate Technology Pub., London. CAPART (2007), Council for Advancement of People’s Action and Rural Technology, Accessed from: http://capart.nic.in/, on: June 2007. Chasi, M. and Mussukuya, E. (1998), Success Story Evaluation of the Barefoot College - The Social Work Research Centre, India, The United Nations Environment Programme. Chaurey, A. (2001), The Growing Photovoltaic Market in India, Progress in Photovoltaics: Research and Applications, 9 (3), pp 235-244. Chaurey, A., Ranganathan, M. and Mohanty, P. (2004), Electricity access for geographically disadvantaged rural communities--technology and policy insights, Energy Policy, 32 (15), p 1693. EZE (1996), Evaluation of SWRC Network - Report, Evangelischer Zentralstelle fuer Entwicklungsdienst (EZE). Government of India (1997), Ninth Five-Year Plan 1997-2002. Volume II. Thematic Issues and Sectoral Programmes, Planning Commission. Government of India, New Delhi. Hirshman, W.P. (2006), A Tale of Two PV Cities, Photon International, 5/2006 (May 2006), pp 80-85. Human Rights Watch (1999), Broken People: Caste Violence Against India’s “Untouchables”, Human Rights Watch, New York · Washington · London · Brussels. IEA (2003), PV for Rural Electrification in Developing Countries – A Guide to Capacity Building Requirements, Deployment of Photovoltaic Technologies: Co-operation with Developing Countries, IEA PVPS Task 9. Jacobson, A. (1999), Case Study 14: Rural Electrification in Ladakh, India, in, "Chapter 16, Special Report on Methodological and Technological Issues in Technology Transfer", Intergovernmental Panel on Climate Change (IPCC) Working Group II. Jauhari, V. and Kondo, E.K. (2003), Technology and Poverty – Some Insights from India UNU/IAS Working Paper No 103. Johnson, C. (2003), Decentralisation in India:Poverty, Politics and Panchayati Raj, Working Paper 199, Overseas Development Institute, London, U.K. Kumar, R. and Sastry, O.S. (1998), Performance, Evaluation, and Development of Solar Photovoltaic Lighting Systems in India, 2nd World Conference on Photovoltaic Solar Energy Conversion, Vienna, Austria, 6-10 July 1998. Kumar, S., S.C.Bhattacharya and Leon, M.A. (2000), A Status Survey on PV Systems and Accessories in Asia, World Renewable Energy Congress, Brighton, U.K., July 2000. Maithel, S. (1998), Renewable energy plan for Ladakh region, Report No 1996RE61, sponsored by Ministry of Non-conventional and Energy Sources, Tata Energy Research Institute (TERI), New Delhi, India. Malhotra, P. and Bhandari, P. (2002), Rural and rural energy development in India, A meeting convened jointly by the The Energy and Resources Institute (TERI) and PESD New Delhi, India, 5th-7th November 2002. Ministry of Rural Development (2002), Annual Report, Chapter 7: CAPART, Government of India, p 47. Mitra, C., Lal, M. and Camps, M. (1999), Development and Dissemination of Solar Energy Systems invillages of the Himalayas in India: Proceedings of the First National workshop, SWRC and ASVIN, New Delhi, India.
353 Chapter 9. Capability Building at the Barefoot College
MNES (2004), Renewable Energy in India - business opportunities, Government of India, Ministry of Non-Conventional Energy Sources, New Delhi, India. MNES (2005a), Annual Report 2005-2006, Government of India, Ministry of Non- Conventional Energy Sources, New Delhi, India. MNES (2005b), New and Renewable Energy Policy Statement 2005, Government of India, Ministry of Non-Conventional Energy Sources, New Delhi, India. MNES (2006), Sanction Order for Implementation of Solar Photovoltaic (SPV) Programme(s) during 2005-06, Government of India, Ministry of Non-Conventional Energy Sources, Solar Photovoltaic Group, New Delhi, India. Nieuwenhout, F., de Villers, T., Mate, N. and Aguilera, M.E. (2004), Reliability of PV stand- alone systems for rural electrification, Part 1: Literature Findings, Tackling the Quality in Solar Rural Electrification, TaQSolRE. O’Brien, C. (2000), Profile: An Innovative Development Project in Rajasthan, India, UNESCO Management of Social Transformations (MOST) Programme Panchayat and Rural Development Department (2007), Panchayati Raj Information, Accessed from: http://pnrdassam.org/praj.htm, on. Radulovic, V. (2005), Are new institutional economics enough? Promoting photovoltaics in India's agricultural sector, Energy Policy, 33 (14), p 1883. Rehman, I.H. (2002), Rural energy policy and planning: issues and perspective, A meeting convened jointly by the The Energy and Resources Institute (TERI) and PESD New Delhi, India, 5th-7th November, 2002. Roy, S.B. (1996), De-mystification of SPVs to provide lighting: an example of community- supported on-site initiative in Ladakh in the Indian Himalayas, Energy for Sustainable Development, 2 (5), pp 45-51. Roy, S.B. and Joshi, A. (2005), Solar Electrification of Remote and Inaccessible Villages: The Barefoot Approach, Asian Regional Workshop on Electricity and Development, Bangkok, Thailand, 28-29 April 2005. Sastry, E.V.R. (2002), The photovoltaic program in India, in 29th IEEE Photovoltaic Specialists Conference, New Orleans, Louisiana, USA, p 28. Sastry, E.V.R. (2003), Village electrification programme in India, p 2125. Saxena, N.C. (2003), The Rural Non-Farm Economy in India: Some Policy Issues, Rural Non- Farm Economy and Livelihood Enhancement, DFID-World Bank Collaborative Research Project, Natural Resources Institute. SELCO (2007), Web Site, Accessed from: http://www.selco-india.com/, on: November 2007. Sharma, B.D. (2000), Annex 4. Experiences in India, in Nieuwenhout, F.D.J. (ed), "Monitoring and Evaluation of Solar Home Systems - Experiences with applications of solar PV for households in developing countries", Netherlands Energy Research Foundation ECN and Department of Science, Technology and Society of Utrecht University. Stone, J.L., Tsuo, Y.S., Ullal, H.S., Wallace, W.L., Sastry, E.V.R. and Baoshan, L. (1998), PV Electrification in India and China: The NREL's Experience in International Cooperation, Progress in Photovoltaics: Research and Applications, 6 (5), pp 341-356. Stone, J.L. and Ullal, H.S. (1998), The Ramikrishna Mission Economic PV Development Initiative, 2nd World Conference on Photovoltaic Solar Energy Conversion, Vienna, Austria, 6-10 July 1998. Stone, J.L., Ullal, H.S., Chaurey, A. and Bhatia, P. (2000), Ramakrishna Mission initiative impact study-a rural electrification project in West Bengal, India, in Twenty-Eighth IEEE Photovoltaic Specialists Conference, pp 1571-1574. TERI (2003), Enhancing electricity access in rural areas through distributed generation based on renewable energy, A Policy Discussion Forum Base Paper, Tata Energy Research Institute., New Delhi. UNDP (2003), Current Trends, Perspectives, Events: Windows on the South, Cooperation South, Number 2-2003, pp 93-104. UNDP India (2003a), Renewable Energy for Rural Livelihoods Project Brief, Poverty Eradication & Sustainable Livelihoods Programme, UNDP Project of the Government of India.
354 Chapter 9. Capability Building at the Barefoot College
UNDP India (2003b), Renewable Energy for Rural Livelihoods Sub-Programme Document, Poverty Eradication & Sustainable Livelihoods Programme, UNDP Project of the Government of India. Vania, F. and Taneja, B. (2005), People, Policy, Participation: Making Watershed Management work in India, International Institute for Environment and Development (IIED), London. U.K. World Bank (2000), India: Policies to Reduce Poverty and Accelerate Sustainable Development, World Bank Structural and Social Policy Review, World Bank, Washington D.C., U.S.A.
355 356 CChhaapptteerr 1100.. DDiissccuussssiioonn aanndd CCoonncclluussiioonnss
In this chapter, the findings from the case studies in the previous chapters are analysed with reference to the existing literature and the framework developed in this thesis. The role of each type of learning for PV manufacturers in developing countries is discussed. Typical barriers, successful capability building strategies and suitable institutional arrangements and/or interventions related to each type of learning are therefore suggested. Section 10.1 employs the part of the framework related to the technological system to systematically discuss each of the roles of technological systems in supporting learning. The ability of developing countries to satisfy these requirements is then discussed, based on the experience of the case studies and using data from the literature review carried out in the background chapters to suggest the extent to which the findings can be generalised to other cases. Conclusions are made about which types of countries are better suited to attempt PV manufacture at different scales. Suitable institutional arrangements and interventions to support local manufacture are suggested on the basis of what has proven useful in the case studies, augmented by the capability building literature. Section 10.2 looks at the roles of each of the three types of learning: learning by doing, searching and interacting in different types of PV enterprises in developing countries. Strategies that may be used by enterprises to enhance learning are identified, based on the experiences of the case studies. The PV literature and the literature on capability building in developing countries is again used to suggest how much the findings can be applied to other enterprises. In section 10.3, conclusions are made in relation to the value of the new framework and the simulation software that have been developed in this thesis. Limitations of the study and areas for further research are identified. The key findings from this thesis are summarised, including typical barriers, suitable capability building strategies, enabling environments and interventions to support the manufacture of PV system components in developing countries.
357 Chapter 10. Discussion and Conclusions
10.1. Technological Systems for PV Manufacture in Developing Countries
Market and non-market interactions enable the actors in technological systems to exchange resources and information and to learn by interacting. Because of inadequate institutional arrangements, both markets and non-market interactions in developing countries are subject to high transaction costs and poor information flows. These problems impact the ability of manufacturers to access resources for production and innovation, their ability to find and service markets for their products and their ability to learn from their interactions in these networks. In this section, the framework developed in this thesis, on the basis of the technological systems approach, is used to systematically analyse the functions of networks for PV manufacture in developing countries. The part of the framework dealing with technological systems is depicted in Figure 10-1.
Figure 10-1: Technological System for PV Manufacture in Developing Countries
TECHNOLOGICAL TRAJECTORY
influence perceptions of the provide direction for technological paradigm search
Technological System
Networks
OTHER ACTORS influence policy and institutions
MARKET & NON-MARKET INTERACTIONS influence operation of markets influence connectivity INSTITUTIONS
resources for production and innovation investment opportunities knowledge creation & exchange provide incentives to invest and improve direction & incentives alter allocation of resources
ENTERPRISE
In the following discussion, what is known from the PV and capability building literature is supplemented by insights from the case studies in order to build a picture of the typical operation of PV networks in developing countries, by addressing each of the headings in Figure 10-1 that link market and non-market interactions to the enterprise. Barriers to the
358 Chapter 10. Discussion and Conclusions
successful provision of the resources, opportunities and incentive to enterprises are identified and institutional arrangements or interventions that may improve their success are suggested.
10.1.1. Resources for Production and Innovation One of the roles identified for technological systems in the framework above is to supply actors with the resources required to carry out the functions of technology generation, production and diffusion. In the case of PV, after-sales service also deserves special attention because of the difficulty of delivering, using and maintaining technology in remote markets, which are the most significant markets for PV in developing countries. The resources required include human and financial resources, physical infrastructure and the materials and equipment required for production. It will be demonstrated in the following sections that nearly all of these resources are likely to be in short supply in developing countries, particularly in the small scale sector and in remote areas. Specific policies and interventions to support their provision are therefore suggested.
Production Equipment and Materials Advanced technology, such as PV cell manufacture, relies on the technology embedded in physical capital. The most sophisticated equipment for solar cell manufacture includes equipment that uses lasers, plasmas, vacuums and equipment for screen printing, the quality of which has been found to be of critical importance in the case of Suntech. The review of the PV industry in chapter 3 revealed that the manufacture of equipment and materials for PV cell manufacture is concentrated in a few countries, primarily Germany, the US and Japan, with China emerging as a new producer. Manufacturers in developing countries will therefore have to import most of their equipment and materials from industrialised countries. The quality and purity of materials, such as metal pastes for the screen printing processes and chemicals for the wet chemistry processes and gaseous deposition, including those used in the diffusion and anti-reflection coating processes, is also critical. Although more advanced developing countries may have the capacity to supply low cost chemicals, they are likely to have difficulty achieving the purity required, which was initially found to be the case in China. Where equipment was produced locally in China, it was also found to be poor quality, although benefits in terms of low cost and local availability of spare parts and service were obtained. As a result of the development of the Chinese silicon and equipment manufacture industries, local manufacture of low cost inputs to production has proven to be one of the greatest competitive advantages for Suntech. This industry development was facilitated by increasing markets for suppliers to the rapidly growing PV cell manufacturers, but Suntech has also acted as a prime mover in supporting the R&D efforts and investment in production of these manufacturers. Developing countries aiming to support PV cell manufacturing should
359 Chapter 10. Discussion and Conclusions
consider investing resources into supply industries. This could be achieved via fiscal measures, investment grants and supporting R&D and education in appropriate fields. Module assembly requires specialist materials such as glass and encapsulating materials that have been developed specifically for the industry. Commercial techniques for module assembly also rely on specialised lamination equipment. These materials and equipment are usually not manufactured in developing countries and must be imported. The Suntech case study revealed that, as a result of strong industry growth and the support of manufacturers such as Suntech, EVA encapsulant of a suitable quality is now being produced in China, but that the specialised glass has taken longer to master. Module manufacturers can benefit from low cost equipment and materials in the same way that cell manufacturers can, and support for the development of supply industries would therefore be of benefit. However, the case of China has shown that the manufacture of these materials is extremely challenging. There are many developing countries that may have the capacity for module assembly that would not be able to manufacture materials and equipment of sufficient quality, so would need to rely on importing. BOS manufacture requires electronic components, PCBs, lamp mountings, casings for charge controllers and equipment such as power supplies, soldering irons and multimeters. These materials and equipment are available at low cost and sufficient quality in urban areas of countries with electronics industries, such as India and China, which are therefore more favourable environments for small scale PV manufacture, whereas in other developing countries, such as in Nicaragua, and in many of the other less industrialised countries reviewed in chapter 4, including Bolivia (Aguilera & Lorenzo, 1996) and Sri Lanka (Huacuz & Gunarante, 2003), they are not available locally or are of unacceptable quality. Hence enterprises will need the capabilities to source and import inputs, which is likely to be particularly challenging when small quantities of materials are required.
Financial Capital
Through the literature review in chapter 2, it was established that 1GWp cell manufacturing plants are expected by 2010, at a cost of around US$1 million per MWp (Lüdemann, 2005; Solarbuzz, 2007). In order to operate at a competitive scale, new manufacturers will therefore need to raise significant finance to invest in such a facility. The availability of financial capital is determined by investment attractiveness, which is influenced by the levels of business regulation, and certainty in macroeconomic variables and policy (Katz, 1984); as well as the availability of appropriate resources for production, markets and opportunities for interactions. Once a country has established its ability to support PV production, or possibly related industries, raising capital for new plant in that country will be straightforward, but there are only a minority of countries that would be viewed as appropriate
360 Chapter 10. Discussion and Conclusions
by investors. The case of China has demonstrated that state-owned enterprises may not have sufficient access to finance for improvements. Although the plant for small scale PV manufacture is fairly basic, requiring minimal financial investment, there is a need for finance for startup, working capital and for expansion and innovation. The available micro-credit schemes have been found to reach only a small proportion of businesses (Allal, 1999). For instance, although they were available in Nicaragua, they were not accessible to Grupo Fénix. The poor often can not take advantage of such schemes where they are available, since they do not have sufficient information, are risk adverse and often lacking in confidence (Saxena, 2003). Most small enterprises therefore use their own private savings for establishing a business, and profits for expansion. Since poor entrepreneurs are usually not able to take entrepreneurial risks with the assets they depend on, lack of access to finance will constrain investment in small scale manufacturing. Financial capital is also crucial in user financing, since PV is very expensive relative to the income of rural users, who are also generally not able to access funding at a reasonable cost through financial institutions (Romijn, 2001). IEA’s (2003) report on capacity building recommends awareness raising on implementation models for PV, risk mitigation and factors affecting risk in PV financing, and training in providing services to rural populations be provided to financial institutions. The findings of this thesis would support these recommendations.
Infrastructure Infrastructure is a strong determinant of transaction costs and risks. Good infrastructure can increase information flows and market access, can save time and increase productivity as well as making previously inaccessible goods, services or opportunities available. In the case of modern sector manufacture, infrastructure requirements include energy supply, transportation, telecommunications and ports. PV cell production has particularly high electricity requirements, especially for silicon feedstock production. There is therefore a need for a secure supply of electricity at a reasonable cost. A higher degree of industrialisation generally will have positive synergies in terms of infrastructure provision, as a government will have more justification for investing in infrastructure where industry already exists. Countries which have relatively industrialised districts are therefore more favourable for PV cell manufacture. While module assembly has much more modest electricity requirements, module assembly using commercial technology can be constrained by inadequate electricity supply, as occurred in Nepal (Katic, 2002). Small scale enterprises, which are usually located either in underdeveloped urban areas or in rural areas, generally have less access to infrastructure services such as energy supply, modern communications systems or transport than modern sector enterprises. Energy, for
361 Chapter 10. Discussion and Conclusions
example, is often supplied preferentially to industrial zones and urban areas over poor rural areas. The infrastructure available to small PV manufacturing enterprises will influence their connectivity to, and the cost of their transactions with users, suppliers and with the organisations that design and implement institutional arrangements. Infrastructure is required for delivery of systems and installation and maintenance services. Remote areas are particularly poorly served by infrastructure, because of the dispersed population, poor ability to pay and the high unit costs of provision. Telecommunications and internet services usually do not reach the poor, reducing access to information about technology and markets. Despite being in an urban area, poor access to telecommunications was a barrier for Suni Solar in relation to the importation of goods and connecting with customers. While the economic benefits of good infrastructure are recognised, most developing countries cannot afford to provide it, especially in remote areas. Small scale manufacturers may therefore need to employ strategies such as forming networks to overcome lack of access to infrastructure. These strategies are discussed in section 10.2.
Human Resources Sharif (1992) observed that although good quality production equipment may be imported, the skills to use the equipment, the documentation and the organisational frameworks must be in place to use them effectively. People with the right technical skills are often not available in developing countries as a result of shortcomings of education systems, exacerbated by ‘brain drain’. Hiring experienced staff is likely to be particularly problematic in the PV industry in the short term, because there is a shortage of personnel with industry specific technical expertise as the industry rapidly expands. Due to the lack of codification of much of the knowledge relating to commercial solar cell production processes, the tacit knowledge embodied in experienced engineers will be important in achieving good results in the manufacturing environment. The availability of skilled personnel is largely dependent on national science, technology and education policies. In order to build PV expertise, countries may invest in PV specific tertiary education and promote postgraduate exchanges with overseas universities. There are, however, few educational opportunities specifically related to photovoltaics manufacture worldwide. Training people in manufacture is challenging due to commercial considerations and access to facilities and therefore occurs mainly on the job. The Virtual Production Line software, described in chapter 6, or similar software simulation tools may facilitate education related to PV manufacture in developing countries, and has been useful in the case of Suntech. An existing PV industry or complementary industries and PV research organisations is also likely to build a stock of appropriate personnel. Developing countries with
362 Chapter 10. Discussion and Conclusions
infant or rapidly growing PV industries, however, will need to recruit internationally, particularly to fill key roles. These countries should facilitate the hiring of international experts. While lack of availability of highly skilled personnel is likely to be a barrier to PV manufacturers in developing countries, cheap labour is an advantage in terms of manufacturing costs. In the case of Suntech, lower cost and higher yield production was achieved through low labour costs, which enabled the use of manual handling and more emphasis on quality control and local maintenance. Rural labour is likely to be very cheap, which may offer cost advantages to manufacturers in these areas. However, most small enterprise owners and employees in developing countries have not had access to formal vocational training. Formal training often requires literacy, is not generally designed to suit the skill requirements of the poor, and is usually not designed to train people in enterprise development (Bennell, 1999). Technical, business and organisational skills are therefore often in short supply in rural areas and amongst the poor and learning on the job is the most common way that small scale workers acquire skills. Given the complexity of PV technology, this type of skills development is likely to be of inadequate scope and quality. Small PV businesses could access generalised small business support, such as business development services where they are available, but PV specific training is also required. PV specific training is rare, and training related to small scale BOS or PV module manufacture would only be available through a handful of programmes such as to those described in the review of small scale PV manufacture in chapter 4. The importance of the small scale sector needs to be recognised at a policy level for the investments to be made in these services. The success of a small scale technological system also depends on the knowledge and skills external to small scale enterprises. If the commercialisation of PV technology is to be a success, the technological system must supply people with the capabilities to carry out all the functions in the supply chain: to install and maintain systems, manufacture or supply quality components, and manage the administration and financial aspects of implementation schemes. Developing countries have a scarcity of trained personnel, and are particularly lacking in technical and administrative skills (Gow & Morss, 1988). There is therefore a need for training specifically for PV technicians or manufacturers and for well designed business development assistance for small enterprises and manufacturers. Capacity building within project administrations and government ministries is also required.
Resources for Innovation In addition to the human, financial and physical capital required to carry out innovative activities in relation to PV manufacture, governments may fund R&D, since the investment
363 Chapter 10. Discussion and Conclusions
costs and risks are often too significant for private enterprise to undertake and because much of the benefit accrues to other actors in the technological system. While PV R&D in China is now approaching the levels of the leading industrialised nations, PV R&D expenditure in developing countries is usually insignificant, and the facilities and technology are often outdated, as was the case in China prior to the last decade. PV enterprises in developing countries therefore face barriers to learning by searching and may therefore be at a disadvantage if they cannot access technological collaborations with research organisations internationally to overcome the domestic shortcomings. Flexible IP arrangements have enabled collaborations at Suntech, and have been important in the US PV industry (Norberg-Bohm, 2000). National strategies that promote this approach are recommended. Much of the manufacture of PV BOS specifically for use in developing countries is carried out in small, inadequately resourced enterprises which do not have the capabilities to invest significantly in R&D. Additionally, most of the R&D directed at the development of BOS carried out in industrialised countries is directed toward the large grid-connected markets in these countries. R&D carried out locally in developing countries is more likely to result in BOS technology that is an appropriate price and quality and well suited to the requirements of local users. There is, however, little funding for this type technology development, and the government and NGO organisations that work with appropriate technology in developing countries are likely to have inadequate PV specific capabilities. Consultants are alternative sources of innovation, but only countries with existing electronics industries, such as India, will have this type of expertise available and many small manufacturers will be unable to pay for consultants. Increased attention from international organisations on PV technology development assistance can play a role to assist small PV manufacturers. Local standards and testing facilities may also be a useful resource to support innovation, since they can provide guidance and assistance to local enterprises. Chinese testing facilities have assisted local module manufacturers in reaching standards suitable for export markets, while Indian, Nepalese (Katic, 2002), Chinese (ter Horst & Zhang, 2005) and Zimbabwean (Tani, 2003) facilities have assisted small manufacturers in improving BOS designs. Most developing countries will require technical and financial assistance to develop standards and testing facilities and to build the capabilities to provide compliance assistance to manufacturers.
10.1.2. Opportunities and Incentives for Investment in Production Markets are the primary source of opportunities and incentives for enterprises to invest in production. While most developing countries have a significant market for PV for rural electrification, these markets are small compared to European markets and usually insufficient
364 Chapter 10. Discussion and Conclusions
to allow PV cell manufacturers to obtain economies of scale sufficient to achieve competitiveness. Because the largest markets for PV are in industrialised countries such as Germany and Japan, the opportunity to access these markets will influence investment attractiveness. Export incentives can increase the competitiveness of manufacturers. The existence of a local market, particularly if it is strongly supported by the government, may also influence investment decisions. Local markets, where quality may be less important than in export markets, are likely to provide an opportunity for new manufacturers to iron out production difficulties, while market support by government signals a commitment to the industry and will give investors confidence. Incentives for investment may be altered by fiscal policies and allocation of resources. Investment incentives provided internationally for enterprises in the PV industry include sales tax relief, exemption from import duties, soft loans, government investment grants and credit and subsidies to increase local market size. Other factors in technological systems also influence the competitive advantage firms will find in different countries and therefore the incentives to invest in that country. The availability of resources, such as an appropriately skilled labour force, low cost inputs to production; and the opportunities for interactions, such as the existence of industry clusters and opportunities for R&D partnerships, are likely to influence a firm’s decision to locate manufacturing in a particular country. Developing countries offer opportunities for low-cost production, due to low cost labour, land, services and potentially the development of an industry supplying low cost equipment and materials for production. As previously discussed, however, skilled personnel, materials and equipment for production and innovation resources are unlikely to be available. Additionally, the business environment in developing countries often involves high levels of administrative complexity and cost in complying with business licensing, tax regulations and labour laws (Tybout, 2000; World Bank, 2006), which discourages investment. The regulation of foreign trade and high taxes have also discouraged foreign investment (Tybout, 2000), which can be an important source of international expertise. Macroeconomic problems, such as shortages of foreign currency (Lall, 2000; Tybout, 2000), inflation, high interest rates and balance of payments problems (Hobday, 1995) also need to be controlled to allow firms to invest confidently. A stable, high growth environment will also encourage firms to invest more in capability building. While these factors are challenging to control, foreign investment can be explicitly encouraged via fiscal measures and investment grants. While PV cell manufacturers in developing countries will most likely need to look to export markets, small scale PV BOS and module manufacturers can produce for local markets, which are almost exclusively rural. These markets are dispersed and remote, and the willingness to pay is usually too low to support a market of sufficient size or density to allow manufacturers
365 Chapter 10. Discussion and Conclusions
to achieve economies of scale in production, to justify investment in innovation, or to allow the effective provision of after-sales service. Where electrification is viewed as a basic need, governments and NGOs may intervene to provide these services. Political incentives and existing expertise, however, often favour grid connection over distributed generation. The promise of the grid can reduce markets for PV, which has been the case in Nicaragua. The case of India revealed that the commercialisation of PV for rural electrification also faces barriers resulting from the entrenched fossil fuel technologies, for which vast infrastructure and institutional arrangements exist. This is the case in a large number of countries. Rural electrification policies which promote fair competition between energy sources and recognise the role of distributed generation are needed. This is especially important during privatisation, when private operators may neglect the expensive and non-profitable remote population, which occurred in Nicaragua. In India, a lack of inter-agency and inter-ministry dialogue and integrated planning over rural energy needs and energy provision has also been recognised as a barrier to success of PV implementation. Where financial assistance, such as subsidies and credit, is available through NGO and government programmes and projects, manufacturers have a better chance of accessing markets of sufficient size. Customer densification through awareness campaigns or good demonstration of technology may also reduce the cost of connecting to rural users. However, because these demonstrations tend to focus on hardware quality, internationally recognised brands are usually favoured in these projects. In the case of national aid programmes, donations are often tied to the use of products from that country, although Australia does not usually follow this practice. Small manufacturers will have a better chance of gaining access to these markets if they are able to achieve recognised certification for their products. Grupo Fénix was not able to access markets created through the large World Bank / GEF Nicaragua project without product certification, while the Barefoot College, which complied with national standards, was able to attract government subsidies and donor support. Programmes or projects that require preferential use of local goods can help to overcome some of these market barriers and provide incentives for investment in local manufacturing. Investment in small scale manufacture will also be constrained by the absence of or lack of access to the resources required, such as materials, equipment, infrastructure and expertise embodied in people and organisations. While interventions such as sales tax and import duty exemptions on PV systems and components may reduce system costs, they may act to disadvantage small manufacturers unless imported inputs for local manufacture are also eligible for these concessions. Of particular importance to the sustainability of the rural PV markets upon which small scale manufacturers directly depend are the resources required to install and maintain systems, carry out market assessments, provide credit and collect fees. Capability building to support small scale PV manufacture should therefore also address these issues.
366 Chapter 10. Discussion and Conclusions
10.1.3. Knowledge Creation and Exchange Both the capability building and institutional economics literature suggest that isolation from the international suppliers and markets are likely to be barriers for manufacturers in developing countries. This has proven to be the case for modern sector PV manufacturers in China where transactions with foreign suppliers and investors have historically been restricted and manufacturers did not therefore have access to the latest technology in processing or equipment and did not interact with export markets which would have demanded quality and cost improvements. While China is a special case as a result of its past closed door policies, high transaction costs and poor information flows are commonly associated with developing countries, which are likely to be exacerbated by geographical isolation from the handful of countries, such as Germany, Japan and the US, where the majority of PV value chains reside. High transaction costs are caused by lack of clarity over property rights, especially in relation to IP, administrative barriers, and fiscal, trade and investment policies that deter trade and investment. A good regulatory environment can keep transaction costs down and improve information flows. On the other hand, domestic interactions are likely to be more intensive and involve more face-to-face exchanges than international interactions. They are therefore often more fruitful sources of learning. The presence of a critical mass of local manufacturers in China has allowed the emergence of local suppliers to the industry. Learning through interactions with local suppliers of capital goods and materials has enabled Suntech to reduce costs and implement innovations, which have been an important source of their competitive advantage. It is expected that innovation may also be stimulated when a number of local suppliers compete with each other. The development of domestic industry clusters can improve the connectivity and information flows between manufacturers and suppliers, and therefore increase knowledge creation and exchange. Technology parks are often assumed to enhance learning by interacting through close proximity of actors, but there is little evidence that learning by interacting has been enhanced by being located in a technology park in the case of Suntech. It has been established in the literature review in chapter 5 that small enterprises in developing countries, particularly those located in rural areas, are usually extremely dislocated by geographical isolation, poor communications infrastructure and by inappropriate regulation, such as cumbersome administrative procedures in relation to accessing markets and importation, and lack of protection to ensure the rights of participants in markets. This poor connectivity results in inadequate information flows and high transaction costs. Suppliers of materials and equipment to small scale PV enterprises will be located in urban areas or possibly abroad and are often not responsive to the needs of these enterprises. They will therefore find it difficult to access information about and may find it difficult to enforce guarantees from urban or international suppliers. They therefore face high costs and 367 Chapter 10. Discussion and Conclusions
risks in their transactions with suppliers. There is little that could easily be done at an institutional level to give small enterprises better access to urban imported goods, but the formation of networks may help them overcome these difficulties as discussed in section 10.2. The markets for PV in developing countries, in which small and medium PV enterprises will participate, are small, dispersed and unstable, so connectivity is even worse than it is in input markets. Many of the exchanges of information and resources in the delivery of PV systems to rural communities take place through non-market interactions, such as projects and programmes. Small retailers and manufacturers are likely to find it difficult to access information about these projects or programmes. Prohibitive administrative procedures, fiscal procedures and tendering processes can be barriers to small enterprises. Poor connectivity between poor, uneducated users and urban based suppliers or authorities is often the result of inappropriate institutional arrangements or of cultural factors, and leads to poor understanding of and enforcement of the rights and responsibilities of different actors. Despite good cultural relations with rural users, the institutional arrangements for system installation, maintenance and management in the case of Grupo Fénix did not support good connectivity, leading to poor system maintenance and reduced market size. The Barefoot College case study demonstrated that knowledge creation and exchange may be supported by networks and groups of users or manufacturers, which can offer better opportunities for interactive learning and the exchange of information about products and markets. The Barefoot College network also facilitated small producers influencing and accessing institutions that determine the nature of their markets and access to infrastructure, including standards and testing facilities. If users, technicians, manufacturers and suppliers are well connected, field experiences can feed back to improve designs and the selection of components, which has occurred at the Barefoot College. The characteristics of networks and linkages between manufacturers and rural users that best facilitate commercialisation of PV technology are discussed in section 10.2.
10.1.4. Incentives for Innovation The technological system must provide incentives for both technological and institutional improvements in all phases of technology generation, production, diffusion and after-sales service if enterprises are to become and remain competitive. According to conventional economic theory, these incentives are provided by the market. Porter (1998) believes that as a driver of technological change, the size of domestic demand is less important than its character, since buyers that are sophisticated will pressure firms into innovation. The incentives for innovation were indeed particularly weak for the Chinese state-owned PV manufacturers, which had preferential access to local markets. Lall (1992) therefore sees international competition as the most important incentive to skill and technology upgrading, but
368 Chapter 10. Discussion and Conclusions
points out that stiff competition can also stifle innovation when firms struggle to survive. Export markets, particularly those in Europe, have offered strong inducements for Suntech in China to innovate, because of the competitive environment and high quality requirements. While infant PV cell and possibly module manufacturing industries may need some protection from international competition, more innovation and therefore lower cost and higher quality products will result from exposure to competition. Where capabilities are initially low, this competition could be gradually introduced as local enterprises build their capabilities, eventually enabling them to compete in export markets. The introduction of standards through the China Renewable Energy Development Project has also provided incentives for module manufacturers to increase their quality, and they have been assisted in achieving the standards, allowing them to access export markets and increasing the acceptance of Chinese products generally. However, quality certifications have been faked by some Chinese manufacturers. There is a need to develop the institutions to enforce quality standards. Although export markets are a strong inducement to innovate, in general, the market does not provide sufficient incentives for the private sector to make the large investments required for technology development in high technology industries. Most countries, including many developing countries, therefore select certain industries within which to support R&D, either by funding it directly in research institutes or by encouraging innovative efforts in enterprises. As discussed in section 10.1.1, the most advanced PV manufacturing nations have large expenditure on PV R&D, often between 50 and 100 million US dollars per year. Because most developing countries will not be able to match these investments, they will need to come up with innovative ways of encouraging international research collaborations in order to reach and stay at the technological forefront. Market failures in relation to technology development are particularly severe in the case of technology for the poor, because poor users do not have the economic resources or sufficient connectivity with technology suppliers to send strong signals via markets. There are therefore insufficient incentives for investment in improved hardware, such as designs and techniques for small scale production, or for improvements in the organisation of value chains, especially in relation to the maintenance and repair of products for rural use. Rural PV markets in developing countries can be broadly grouped into cash-based, and project-based markets, which may include donor, credit or fee-for service models. On the basis of the literature review in Appendix 2 and observations from the case studies, Table 10-1 lists the advantages and disadvantages of different incentive mechanisms in both cash and project based markets for (a) quality in the design and production of hardware and (b) quality in the delivery of service.
369 Chapter 10. Discussion and Conclusions
Table 10-1: Incentive Mechanisms for Improvement in the Design, Production and Delivery of Quality PV Systems in Rural Areas of Developing Countries
Design and Production of Delivery and After-Sales Service Hardware Cash Market User Preferences Maintenance Agreements Users can express their preference Systems are often sold with no maintenance through the market. Price is a higher agreement. Where they do exist, small retailers priority than quality for poor users may go out of business, leaving agreements un- who may therefore provide more serviced. Poor users may not be able to enforce incentive for low price than high their rights. Groups of users are better able to quality and may not use standards, enforce contracts. Codes of conduct may be warranties or brands to make agreed to by retailers who wish to build decisions. confidence in the technology.
Standards and Certifications Reputation Certifications can be faked and poor Reputation may be an incentive for local retailers users usually do not recognise or who wish to preserve their market to provide understand them. Standards may good service. However, where customers are encourage less diversity in products. dispersed, information flows to new customers may be inadequate. Warranties Manufacturers have incentive to reduce the risk of a claim under warranty, but warranties are difficult to enforce by poor customers and small retailers in cash markets, especially warranties from international suppliers.
Brand Reputation Dispersed customers may not be well informed about which products work well. It is usually assumed that imported products are superior. Project or User Preferences Project Objectives Programme Users may not be consulted, so their Project objectives may be geared toward preferences do not act as incentives, achieving targets rather than supplying energy to resulting in inappropriate system users or preserving the reputation of the hardware. Project design may reduce technology. Project timescales may provide the variety of systems available to incentives for short term participation of actors suit the needs of customers. only while subsidies are available. Technical support may therefore cease, while unstable Standards and Certifications markets make it difficult for small enterprises to Can be enforced by project survive. implementers who can ensure validity of certifications. Standards may not Maintenance Agreements ensure long term performance in the Maintenance agreements usually exist. They may field. Standards may prevent small not be implemented well at the local level due to: manufacturers from participating Poor linkages from central administration to without financial and technical local level, assistance. Inappropriate fee collection arrangements, Inadequate enforcement of users rights and the Warranties responsibilities of all actors. Users may not be aware of their rights and they may not be enforced. Fee for service arrangements tie payments to Need good organisational structure at service, increasing incentives for improved local level, contract enforcement and maintenance. links to manufacturers to enforce warranties.
Brand Reputation Favouritism for known brands or influential suppliers may be a barrier to small scale manufacturers.
370 Chapter 10. Discussion and Conclusions
Warranties have been an ineffective incentive to improve quality in many countries, such as Kenya, because the institutional arrangements are not able to enforce the warranties. In response, a code of conduct has been adopted by an association of PV retailers in Kenya to improve the reputation of the technology (de Villers, 2005). Certification of installers and maintenance technicians could also act as a quality improvement incentive for those aspects of PV delivery, if there is provision for the certification to be revoked as a result of poor performance. It has been found that where systems are donated to communities, which has been the case in many instances in Nicaragua, there is less incentive to keep the system going or to enforce warranties. Organisational structures which make individuals accountable for the system work better, especially when the community makes a commitment to the technology via payments. Fee for service arrangements would be expected to encourage better hardware and maintenance on the part of the service provider, but have not been investigated within the scope of this thesis. However the link between fee collection and maintenance was present in the case of the Barefoot College, and appeared to be an effective incentive mechanism. Reputation-based incentives and warranty enforcement, however, may work well both cash and project-based markets if there are strong links and accountability between manufacturers, installers, users and the administration of the scheme. Protection of its brand reputation encouraged Grupo Fénix to begin testing IV curves of modules after an installation with low output was discovered. The incentive was enhanced by strong links between Grupo Fénix and its customers. While the Barefoot College PV programme is donor-based, there are strong and permanent local links between actors because the manufacturing, installation, maintenance and administration functions are all carried out in the villages by local people. There is therefore a reputation based incentive for the technicians to provide quality service, and the committee structure enforces the rights of users and ensures that payments for the service are made. Local groups of users may also be able to enforce warranties and verify quality certifications more effectively than they could individually. Standards required for acceptance into projects and for eligibility for subsides have been strong incentives for both the Barefoot College in India and Grupo Fénix in Nicaragua, although high costs and lack of support prevented Grupo Fénix from achieving compliance.
In summary, arrangements that provide better incentives for good quality energy supply from PV include: Ensuring that user preferences influence hardware choice and institutional arrangements, Enforcing standards, warranties and user obligations, for example by tying service, maintenance agreements and warranties to payments, through a regulatory body that
371 Chapter 10. Discussion and Conclusions
can monitor long term performance, or by forming user groups or codes of conduct among suppliers. Ensuring good links between users and service providers and contracts with clear roles and responsibilities, usually easier where there is a local administrative presence, Avoiding short-term projects and instead implementing long term programmes and smart subsidies that can be carefully targeted and gradually removed.
Disadvantage to local manufacturers can be avoided by: The establishment of local testing facilities and supporting manufacturers in meeting standards, Local procurement rules.
There are also often insufficient incentives to develop appropriate technologies for the poor. For instance, new encapsulation materials that are currently under development promise vacuum-free module assembly, which may enable the use of simpler, lower cost encapsulants, equipment and techniques and create opportunities for small scale manufacture. There is, however, no incentive for materials and equipment to be developed to suit small scale module manufacture, although the author has been informed that products suitable may already exist and would certainly be possible to develop (Lenges, 2006). Incentives may also be used to encourage private sector involvement in technology for the poor, such as through subsidies or public procurement programmes. However, these interventions have sometimes acted as disincentives to improvement. For example, where small business assistance has included an assured public market for the product, there has been no competitive pressure to optimize production (Romijn, 2001). Governments or NGOs may also directly allocate resources to R&D for appropriate technologies and in doing so, influence the direction of technology development.
10.1.5. Interactions Between Networks, Policies and Technological Trajectories Policies and institutions are likely to favour existing technologies, because of the path dependence of institutional change and the influence of vested interests in those entrenched technologies. Political acceptance or legitimisation of PV as a suitable or significant technology may be necessary to achieve suitable institutional arrangements to facilitate the growth of markets and to work to ensure the availability of resources required by the industry. The actors in the technological system can influence the institutional arrangements in their interest, as illustrated in Figure 10-2, which shows part of the framework related to the interactions between networks, institutions and the technological trajectory.
372 Chapter 10. Discussion and Conclusions
Figure 10-2: Interactions between Networks, Institutions and Technological Trajectories
TECHNOLOGICAL TRAJECTORY
influence perceptions of the provide direction for technological paradigm search
influence policy and institutions
Networks INSTITUTIONS
Institutional barriers to PV rural electrification related to entrenched technologies can include the subsidisation of fossil fuels, the expectation of access to the grid on the part of users and political attachment to grid extension. Institutional arrangements in favour of PV will only arise if the technology can be legitimised, which involves acceptance by those involved in the design and implementation of institutional arrangements for the delivery of the technology, but also requires acceptance by users. The standards set within projects or at a national level inform the technological trajectory. Better performance through the implementation of standards and capacity building have been used in China to increase market size and improve the perception of the technology. While access to the international technological paradigm has been required to reach the forefront of technology, Chinese manufacturers have now begun to influence the international technological trajectory through low cost manufacturing. Technology for the poor faces particularly high barriers to acceptance by users, due to the poor’s low ability to take risk and slow acceptance of change. The promotion of technology from external sources is problematic because poor users are more likely to accept the opinion of their peers in relation to the suitability of technology. Past experiments with PV technology in many countries have resulted in an inferior reputation due to poor performance, high cost, the generation of small amounts of power that do not service productive uses and failure to target the poorest. As more users adopt a technology, however, favourable institutional arrangements may emerge and the technology may be legitimised in the eyes of policy makers. PV technological systems need to provide quality service to the users and fulfil the goals of projects in order to legitimise the technology. The provision of standards and quality testing and codes of conduct can also inform the technological trajectory in the case of small scale technology. However, it should be recognised that the technological trajectory for PV system components in developing countries is often determined by decision makers from industrialised countries, in particular, those who are involved in large donor or unilateral aid programmes. These agencies carry out demonstrations of technology, credit and repayment
373 Chapter 10. Discussion and Conclusions
schemes, and administration. International research organisations have also developed standards such as the Thermie Universal Standard for installation and maintenance (THERMIE, 1998), standards for testing facilities, design and manufacture of PV system components from the Quality Programme for Photovolatics (Atmaram & Roland, 2001; Varadi et al., 2003; Vervaart & Nieuwenhout, 2000) and the Sandia manuals (Architectural Energy Coorporation, 1991; Dunlop, 1997). Through the dissemination of case studies, successful PV project implementations can also influence future decision making. Groups are better able to influence the institutional arrangements in favour of the technology than enterprises. Politically influential groups in Germany, such as renewable energy advocacy groups, environmentalists and wind industry actors have increased the support for PV through lobbying and engagement in the political process. The establishment of renewable energy targets in China has led to the establishment of institutional arrangements that support PV markets and the development of a manufacturing industry. It appears that they result in part from the policy of universal electrification and energy security concern, but the scope of the case study in China was not sufficient to confirm the influences that determined the policy decisions. The small scale PV sector, however, does not have many powerful lobbyists, and may therefore be at a disadvantage in effecting institutional changes. Actors with better connectivity to regulatory bodies will be able to influence procedures and standards in their interest. The feedback from monitoring may also be used to provide direction for the improvement of institutional arrangements.
374 Chapter 10. Discussion and Conclusions
10.2. Capability Building Strategies in PV Manufacturing Enterprises in Developing Countries
The literature review in chapters 2-4 revealed that the capabilities of PV manufacturing enterprises in developing countries are usually inadequate. The majority of cell manufacture has been historically located in industrialised countries, and is otherwise limited to a few developing countries which have capacity in other high-tech manufacturing industries. Despite lower manufacturing costs, most PV cell manufacturers in developing countries have until recently been unable to produce products of sufficient quality to access export markets, which are the largest and fastest growing markets for PV. They have therefore been unable to achieve economies of scale in production. Prior to the last decade, most of the cells and modules exported from developing countries have been manufactured under foreign ownership or joint venture arrangements. In these cases, technological independence is not achieved. The case study on Suntech in China has demonstrated that the capabilities can, however, be built rapidly and independently. It has been proposed that PV cell manufacture may be a suitable technology for advanced developing countries because of its technological proximity to the growing electronics industry, its potential to provide electricity for remote areas and contribute to energy security. The strategies that have allowed Suntech to overcome barriers in the technological system and build capabilities in PV cell manufacture are therefore of interest and this section discussed the extent to which they are likely to be useful for other enterprises. The literature review in chapter 4 found that the majority of the manufacture of small charge controllers and lamps for use in developing countries is carried out domestically and usually at a small scale. Local manufacture has the potential to provide components which are better suited to local markets at lower cost and improve the maintenance and use of the technology, and can also promote the creation of local livelihoods, the reduction of leakages from the local economy in the form of payments for energy services and the development of local capabilities. However, while some small scale manufacturers have been able to produce good quality products, the quality of locally produced components has been found to be inadequate in many cases. The ways that small scale manufacturers have been able to achieve the capabilities to successfully manufacture and commercialise the technology are therefore also of interest. The observations from the case studies will be used to suggest strategies that are likely to enable small scale PV enterprises more generally to manufacturers to overcome the constraints of initially low levels of capabilities and poorly functioning technological systems.
375 Chapter 10. Discussion and Conclusions
10.2.1. Learning By Doing, Production Capabilities and Routines Learning by doing includes the increased understanding and improvements that come about through experience in production. Figure 10-3 shows the part of the framework developed in this thesis concerned with learning by doing.
Figure 10-3: Learning by Doing
production resources investment opportunities
PRODUCTION CAPABILITIES incentives & resources ROUTINES Learning by Doing
DOING
Because of the path-dependence of technological learning, beginning with mature technology is likely to be a good strategy for latecomer enterprises in developing countries, giving them an opportunity to build capabilities and markets before attempting to introduce new product designs. The review of the PV cell manufacturing industry in chapter 3, indicated that while a plant may be operated with basic process capabilities, manufacturers need to make continuous incremental improvements to manufacturing cost, cell efficiency and yield in order to maintain competitiveness using the mature silicon wafer-based technology. Many of these incremental improvements do not involve innovative processes or device designs, but are the result of learning by doing. In the case of Suntech, learning from experience in the fine-tuning of processes and equipment operation resulted in a competitive 14.5% efficiency within a year of commencing production. Although they do not use complex equipment or processes and the technology does not experience a fast learning rate, quality is important for small scale BOS manufacturers to access and maintain the health of markets. Manufacturers must master techniques such as soldering, component identification and transformer winding and carry out quality control procedures. While information on how to carry out these processes can be acquired externally, experience in production is required to develop good technique and a good understanding of the correct operation of the circuit.
376 Chapter 10. Discussion and Conclusions
In small scale module manufacture, robust products that can withstand harsh outdoor conditions for at least 20 years are required, while expensive materials must be used carefully to avoid waste. Experience is required to master the use of materials and manufacturing techniques. In the case of Grupo Fénix, incremental improvements in the use of materials, in particular the encapsulant, have arisen through experience. Cost reductions could also be expected through streamlining production processes via learning by doing, but this was not observed in the case of Grupo Fénix because it was unable to operate at full capacity. Clearly, learning by doing is important for all PV manufacturers across the spectrum, from small, cottage scale industries to large modern sector manufacturers. Small volumes of production, however, will reduce the extent to which learning by doing can accrue to small scale manufacturing enterprises and therefore the overall rate of improvement via automatic learning by doing.
Monitoring, Quality Control and Training Routines While learning by doing is traditionally viewed as occurring automatically, the case studies reviewed in this thesis have confirmed the view put forward in the learning literature that automatic learning by doing is limited to improvement of routine tasks, and mainly accrues to individuals. Much of the learning related to complex concepts is dependent on the effort expended by enterprises on the implementation of production monitoring, adjustment and maintenance routines that identify areas for improvement, while the learning of enterprises relies on training routines that ensure knowledge is shared amongst individuals. Inadequate improvement routines have been identified in the literature review as one of the weaknesses of manufacturers in developing countries. In the Suntech case, however, improvements were realised rapidly, including improvements in the operation and maintenance of complex equipment and in the fine-tuning of manufacturing processes that in some cases have complex interactions, are not well understood and have not been characterised widely by equations or algorithms. A strong focus on production monitoring through quality control routines has enabled fine-tuning and problem solving, while up time for equipment and therefore capacity usage has been high as a result of maintenance routines. Although the information in the case study was not sufficient to confirm it, routines may also be expected from experiences in other high technology industries to facilitate more efficient plant operation in relation to line balancing and inbound and outbound logistics. While the key foreign-trained staff possessed production and quality control capabilities when Suntech began production, training has been of particular importance in ensuring that these capabilities have been passed on to other staff. The use of the Virtual Production Line (VPL) has enabled Suntech to overcome some of the cost, language, time barriers and capability constraints to training, and has allowed users to gain virtual experience with the interactions
377 Chapter 10. Discussion and Conclusions
between processes, the use of quality control tests and important manufacturing concepts such as yield. There is potential for the VPL to be used by other manufacturers in developing countries to overcome lack of expertise and training in the technological system, and to reduce the costs on the job training. Translation of the software would make the experience embodied in the software and the accompanying help files available to manufacturers from more countries. The literature review in chapter 5 indicated that for small scale enterprises, quality control and management functions relating to production organisation such as inventory accounting are particularly likely to be inadequate. This has been borne out by the case of Grupo Fénix, which did not have the quality control capabilities to ensure that products met specifications. Learning by doing was therefore constrained, as the enterprise did not recognise the importance of methods such as matching cells prior to interconnection to improve performance. Small scale manufacturers would benefit from seeking out quality control procedures, possibly by engaging consultants, as the Barefoot College has done. The staff of the Barefoot College have been able to improve their skills as the testing of circuits has drawn attention to areas for improvement. Life-cycle product testing could also assist them in identifying areas for improvement, especially in the case of modules, which are required to withstand harsh outdoor conditions. Training has also been effective in facilitating learning by doing at the Barefoot College. Through training, the knowledge from experienced BSEs is passed to new BSEs, increasing the overall capabilities of the enterprise. Training was also used to enable communities in administering programmes. The Barefoot communities are able to plan and implement their project and do so effectively because they are responsible for reviewing their own performance and making adjustments. Experience can therefore feed back to more suitable maintenance, financing and fee collection arrangements. The literature review revealed that many training interventions for small scale enterprises have been ineffective. The success of the Barefoot College training can be attributed to the long timeframes which allow trainees the time to obtain practical experience which can be directly applied, and the peer-to-peer approach, which enables new ideas to be framed in a way that can be understood by the trainees. In conclusion, while experience enables enterprises to improve, in the use of complex techniques or the implementation of complex organisational arrangements, monitoring and quality control are essential to ensuring that characteristics of the production or organisation which cannot easily be perceived are observed. In addition, training is critical to ensuring that knowledge is diffused throughout staff, especially where there is little local expertise and foreign trained staff or consultants are used.
378 Chapter 10. Discussion and Conclusions
Linkage and Investment Capabilities and Access to Resources The framework developed in this thesis has proposed that doing based learning relies on the availability of resources for production. While the technological systems in different countries have different abilities to supply these resources, as discussed in section 10.1.1, the ability of enterprises to link effectively to markets and suppliers, and their ability to make the right investment decisions will determine how well they can overcome supply deficiencies on the part of the technological system. Where there are skills shortages, modern sector manufacturers may be able to recruit internationally, which was a strategy successfully used by Suntech, which relied on the PV specific tacit knowledge and experience embodied in key staff that were primarily Chinese nationals who had been trained overseas as part of a Chinese policy of gaining international technological expertise. Many countries will not have the same advantage and manufacturers are likely to have difficulty attracting high quality international experts. Their ability to do so will be influenced by the remuneration they offer and the reputation and potential of the enterprise, as well as the attractiveness of the location, lifestyle and available facilities for expatriates. Where production equipment and materials of appropriate cost and quality cannot be obtained locally, enterprises may need to import them. Modern sector manufactures should be able to obtain the information they require to make these purchase through industry reports and journals, but detailed knowledge of the most appropriate equipment and materials to use and the suppliers that can make the best quality equipment will come only from industry experience. New enterprises may be able to access training on the operation of equipment and some process knowledge by purchasing a turnkey plant. Enterprises with good linkage and investment capabilities, such as Suntech, however, may be able to procure their own equipment, including custom machinery, enabling them to implement novel processes and higher degrees of automation. Equipment suppliers in the PV industry generally lock customers in to long term maintenance contracts, but more experienced firms can negotiate maintenance and training contracts with equipment suppliers, thus building maintenance capabilities and reducing costs. Small manufacturers, on the other hand, are generally poorly connected with suppliers, particularly if they are located in remote areas or are not well served by communications infrastructure. Suppliers are also unlikely to make the effort to supply small quantities on an intermittent basis. It may be difficult for small scale producers to establish contact with international suppliers, due to the small size of order, and the administrative requirements of importation procedures. This was a particular problem for Grupo Fénix. The small scale of production at Grupo Fénix also precluded investment in commercial manufacturing equipment, through which economies of scale could be obtained.
379 Chapter 10. Discussion and Conclusions
In contrast, the components required by the Barefoot College were available in urban areas in India, and the network of small enterprises enabled the purchase of larger quantities and the investment decisions to be made centrally, where there were better linkages to suppliers and investment capabilities. Small manufacturers also need access to financial institutions and/or donors. The plant for small scale PV BOS manufacture is fairly basic, requiring minimal financial investment, but there is a need for working capital and for ongoing R&D. Small scale manufacturers are likely to have trouble raising capital. Investments may need to be shared with other economic activities, due to the small size and low density of PV markets. Financial institutions are unlikely to want to invest in PV because it is viewed as a risky, fluctuating market, using unproven technology and selling to poor customers. Small enterprises will therefore need to demonstrate a good business plan to be accepted for a loan. A network of small manufacturers, such as the Barefoot College network can help overcome these barriers in accessing donor finance for investment in production facilities.
Linkage and Investment Capabilities and Investment Opportunities Modern sector PV manufacturers need the capabilities to access markets, particularly export markets in the case of cell manufacturers. Obtaining certification and promotional activities in export markets are strategies that can expand market size. There may also be some scope for gaining preferential access to domestic markets. Small scale manufacturers are likely to find it difficult either to identify customers who are able and willing to pay or to be able to judge the size of the market. It may also be easy to rapidly saturate a small local market. They require the capabilities to carry out feasibility studies and market research. Financial capital is also crucial for accessing markets through user financing, since PV is very expensive relative to the income of rural users. A substantial percentage of markets for PV in many developing countries are constituted of government programs and NGO projects. The scale, quality requirements and administrative and tendering processes in projects and programmes are likely to be challenging for small firms. Small enterprises should aim to achieve certification in order to access these markets. They may also be better able to achieve scale and influence infrastructure provision, markets or regulatory arrangements as a group, as illustrated by the Barefoot College case. This may be particularly important in an environment of corruption.
380 Chapter 10. Discussion and Conclusions
10.2.2. Learning by Searching and Innovative Capabilities Learning by searching is the conscious search for and testing of new technological options. The portion of the framework illustrating the process of and requirements for learning by searching is illustrated in Figure 10-4.
Figure 10-4: Learning by Searching
innovation resources investment opportunities
COORDINATION & INTEGRATION
incentive direction INNOVATIVE & direction for search CAPABILITIES for search
Learning by Searching
R&D
new production RECONFIGURATION technique
Because of the high rate of technological change in the PV cell manufacturing industry, new enterprises need not only to catch up, but to keep abreast of changes. They must therefore develop innovative capabilities and put significant effort into search activities. The literature review in chapter 3 revealed that efficiency improvements, which rely on the introduction of new processes and/or device designs, are expected to be one of the major sources of improvements with the mature solar cell technology. The case of Suntech has confirmed that world-class efficiencies have been achieved through the development of new processes and device designs. In addition, the development of innovative testing techniques have enabled better production line monitoring and therefore improved yields and efficiencies, which have been important sources of competitive advantage at Suntech. The relatively early stage of development of PV technology, and hence the expectation of a move to alternative production technologies, implies that cell manufacturers will need to be ready to change to new technology in the longer term. While there are many new technology start-ups emerging in the PV industry, including some with manufacturing facilities in developing countries, such as Sunpower, which has a plant in the Philippines, these are primarily foreign owned and it was not within the scope of this research to observe their innovative activities.
381 Chapter 10. Discussion and Conclusions
While innovation is often assumed to be unimportant to small scale enterprises, the case studies indicate that it is a significant source of improvement for small scale manufacturing of PV system components. In the Grupo Fénix case, it was necessary to develop new processes and equipment for cottage industry module assembly and to adapt locally available materials for the purpose. In the Barefoot Collage case, it was necessary to commission the design of circuits for charge controllers, lanterns and lamps to suit local markets and quality control and maintenance checklists to suit local maintenance capabilities. While the Barefoot College engineers did not design the circuits, they were engaged in the innovation process by providing specifications based on field experience and requirements. Their participation in the innovation process contributed to the learning process of the organisation as new knowledge related to circuit design and components, quality standards, field performance, repairability and user preferences was acquired and/or reframed for use in the innovation process. Small scale manufacturers may also innovate in the delivery of their products, for example, by providing innovative financing or after sales service arrangements, which are often critical in achieving locally appropriate strategies for delivering PV technology.
Coordination and Integration Coordination and integration capabilities are those that allow enterprises to align production, search and investment activities. The results from monitoring and from R&D efforts can guide the search for technology and interactions with actors external to the enterprise. Manufacturers that implement comprehensive quality control will be better able to understand and therefore improve their processes and device designs. Suntech has been able to closely align the research carried out through their interactions with UNSW and their production priorities, which has enabled them to incrementally upgrade production. In the Barefoot College case study, the information gathered through quality control, and through the experience of field technicians and users, have informed the adaptation of products, quality control and maintenance routines to better suit local requirements. The activities of the solar section were also aligned with the other development activities of the Barefoot College. For instance, products such as lanterns, which were most suitable for use in night schools, were produced. Grupo Fénix has also been able to align R&D efforts with user requirements via their local links, resulting in development work on low watt LED lighting.
Reconfiguration Enterprises in fast-changing industries such as PV cell manufacture need to reconfigure their assets in order to take advantage of R&D. While the literature surveyed in chapter 3 predicted that the implementation of increasing levels of automation would lead to cost reductions in PV manufacture in industrialised countries, manual processing has given Suntech
382 Chapter 10. Discussion and Conclusions
the flexibility to upgrade production lines and therefore implement improved technology rapidly. Rapid expansion has also enabled Suntech to implement new technology. Reconfiguration is less important for small scale BOS and module manufacture, but improvements based on field experience rely on flexibility to implement change. When capabilities are based around a particular design, and understanding is primarily know-how, rather than know-why, which is the case at the Barefoot College, reconfiguration is more difficult. The engagement with innovation via the use of consultants, however, has promoted a better understanding and facilitated reconfiguration when it has been necessary, for example to comply with standards.
Linkage & Investment Capabilities, Innovation Resources and Opportunities Linkage and investment capabilities enable enterprises to procure resources for innovative activities and to engage in collaborative research to supplement internal R&D. Good links to leading research institutes and a flexible approach to IP arrangements have facilitated research collaborations in the case of Suntech. International hiring has enabled Suntech to acquire personnel with R&D expertise, which has also relied on industry knowledge and know who, the good reputation of Suntech and the existence of foreign trained Chinese nationals who were willing to return to China. Small enterprises do not generally have the resources or expertise to invest heavily in internal R&D facilities. BOS circuit design and quality control procedures must therefore be assimilated from an external source. Small scale enterprises may need to rely on consultants for innovative capabilities, but still need internal capability since the technology must be well- understood in order to adopt designs effectively, carry out manufacturing and quality control and maintain the technology in the field. Networks have enabled the Barefoot College to access R&D resources in terms of donor funding and in terms of capabilities to coordinate R&D. Working with consultants would not be possible for individual rural workshops.
10.2.3. Learning by Interacting and Linkage Capabilities Learning by interacting is the acquisition of technology through interactions with suppliers and customers in value chains, and through technology purchases and collaborations. The part of the framework concerned with learning by interacting is reproduced in Figure 10-5.
383 Chapter 10. Discussion and Conclusions
Figure 10-5: Learning by Interacting
INTERACTING
Learning by Interacting
INVESTMENT & LINKAGE CAPABILITIES
production resources innovation resources investment opportunities investment opportunities informs investment & linkage
COORDINATION & INTEGRATION
A number of different types of learning by interacting: with suppliers, with customers, through research collaborations and through clusters and networks, and the knowledge that can be generated through each, will now be discussed. The importance of linkage and investment capabilities in accessing investment opportunities and the technology embodied in resources for production and innovation has already been discussed. In this section, strategies to improve linkage and investment capabilities are discussed.
Learning by interacting with suppliers The capability building literature predicted that interactions with equipment suppliers would be important sources of learning. In the semiconductor industry, equipment manufacturers have been a major source of technological change, and are expected to be particularly important to the PV industry as it transitions to mass production. In the case of Suntech, interactions with suppliers of materials and equipment have improved Suntech’s knowledge about equipment, maintenance and the impacts of material characteristics on cell characteristics. As a consequence, Suntech has been able to carry out maintenance independently of the equipment supplier and engage with suppliers on the development of proprietary equipment for innovative processes. Through these interactions, Suntech has also been able to provide feedback to suppliers on the suitability of equipment and materials, helping to improve them and therefore giving Suntech access to the materials and equipment it requires at sufficient quality and at low cost. Of particular importance has been the development of new technology for the production of solar grade silicon, since there is a global shortage of silicon and some manufacturers are currently unable to operate at full capacity. Companies such as Suntech, with secure supply arrangements, can benefit from the growing German market and high prices. Countries with existing chemical, metals, semiconductor, and
384 Chapter 10. Discussion and Conclusions
manufacturing industries are likely to offer more opportunities for cell manufacturers to learn by interacting. Learning by interacting with suppliers of materials is likely to be of less importance to BOS manufacturers, because most of the components used are standardised. Opportunities to interact with suppliers of encapsulant materials could, however, be of great benefit to small scale module manufacturers, since the materials for commercial production techniques are not suitable to small scale manufacture. New or existing materials that are suitable could be identified through supplier interactions.
Learning by Interacting with Customers The evidence from the literature suggests that learning may also occur through interactions with customers or downstream parts of the value chain. Manufacturers in Japan, for example, are focusing on grid-connected applications with a large market potential, in particular, building-integrated (BIPV) applications. There has been an effort to standardise and mass-produce elements of these systems, via interactions with the construction industry (Balaguer & Marinova, 2006). In Germany, architects and project managers have also been involved in the development of BIPV products. Suntech has only just begun to move into building integrated PV markets where learning by interacting with users is expected to be of importance, so the case study cannot confirm the importance of this type of learning. In their linkages with users, small scale manufacturers need to not only secure a market for their products, but also ensure that the technology is acceptable to the user. Learning by interacting with users is therefore of great importance, in relation to the hardware, but also with respect to organisational arrangements. This type of learning only emerges where there is good connectivity between manufacturers and users. In the Barefoot College case, strong links were present and technology has been improved in response to user requirements and technician requirements. It can be concluded that groups and networks with community ownership can improve the connectivity and information flows between users and technicians because these social structures enforce informal and formal contracts. Interactions with other NGOs and integrated development approaches also promote learning from innovation in delivery.
Research collaborations Research collaborations can enable firms to overcome their dependency on licensing of technology and subcontracting arrangements, and can help firms move from R&D aimed at improving manufacturing technology, to basic research aimed at developing new products and processes. Because most of the innovative technology in the PV industry is developed in research institutes, linkages with these organisations is expected to be of prime importance to new manufacturers, particularly when their own R&D capabilities are not well developed. Because
385 Chapter 10. Discussion and Conclusions
developing countries usually do not invest strongly in public R&D for PV, manufacturers will need to collaborate with international research institutes. Suntech has confirmed the importance of research collaborations, which have enabled the introduction of innovative processes, new device designs and innovative quality control techniques. Suntech has had particular success in their collaborations, being able to rapidly implement new technology on the production line. This success can be attributed to collaborating on research which is closely aligned with production priorities, having good connections in industry on the part of key staff and adopting a flexible approach to IP arrangements, rather than insisting on ownership of IP. While small scale enterprises may not have adequate capabilities to successfully carry out independent R&D, collaborations with NGOs or other organisations can be useful sources of technology and can build innovative capabilities. Grupo Fénix has engaged with international volunteers and with the University of NSW to make some improvements to their module assembly process. These collaborations produced variable results and it has been concluded that these research collaborations need to be long term, aligned with the priorities of the small scale enterprise and participatory in order to be effective.
Linkage and Investment Capabilities Linkage and investment capabilities that enable learning by interacting in small scale PV implementations rely on local structures and networks. The benefits of networks in improving linkage and investment capabilities and therefore facilitating access to resources, opportunities and interactive learning have been identified. Here the characteristics of networks that lead to effective linkages are discussed. The ability of small scale enterprises to organise in networks depends strongly on their social capital, while their ability to undertake local advocacy and influence authorities depends on their collective knowledge and power. The Barefoot College case showed that good connectivity can be achieved by building capabilities and social capital by: Establishing trust through long term engagement with communities, Allowing people the time to learn and see solutions for themselves, Overcoming entrenched power structures and achieving transparency in decision making, Establishing peer-to-peer relationships, Employing people within the village, and By empowering communities to manage their own programmes.
Trust reduces the risks of transactions, increases the supply of information and allows people to guarantee each other. Participation ensures that the institutional structure is culturally and socially appropriate.
386 Chapter 10. Discussion and Conclusions
10.3. Conclusion
In conclusion, the findings in relation to effective capability building strategies and enabling environments and interventions are summarised, identifying the contribution that this work has made to the knowledge about PV manufacturing in developing countries. The thesis then concludes by discussing the value of this study and its limitations, including the effectiveness of the framework that has been developed and its usefulness for future industry development efforts.
10.3.1. Tentative Conclusions about Capability Building for PV Manufacturers in Developing Countries
Effective Modern Sector Capability Building Strategies Getting mature technology right first Hiring international experts Rapid expansion to achieve scale economies Research collaborations Engaging in export markets & international competition Interacting with and investing in local suppliers Training and international exchanges The use of software training tools such as VPL
Modern Sector Enabling Environments and Interventions Building a critical mass of industry players required for LBI and expansion: o Large and stable markets (may be export, but local markets build confidence) o Support investment in PV manufacturers (tax holidays, government grants, import duty exemptions) o Support investment in supply industries (favourable fiscal arrangements, investment grants) Accessing international expertise through foreign investment o Incentives for foreign investment Flexible IP arrangements to allow access to technology more cheaply Remove excessive protection for local manufacturers and privatise state-owned enterprises Export incentives Good manufacturing infrastructure, low wages, good infrastructure for manufacturing
387 Chapter 10. Discussion and Conclusions
PV specific education locally Support R&D for PV (long term impact) Reducing import duties or sales taxes on equipment and materials for PV manufacture
Effective Small Scale Capability Building Strategies Good quality control, which improves learning and creates good reputation for product Engaging in networks to access resources, markets and influence Establishing manufacture – maintenance – user links The use of training networks Building business skills – procurement, inventory control, market analysis
Small Scale Enabling Environments and Interventions Long term and smart subsidies to support sufficient market density Projects and programmes that support and do not exclude small scale and local manufacturers and support local management structures The introduction of standards and the enforcement of warranties, but with capability building Removal of taxes and duties on materials required for manufacture if not locally available Finance and credit for local manufacturers Removing subsidies on conventional generation Making provision for distributed systems in energy restructuring Supporting R&D by NGOs and/or governments Hands-on, peer-to-peer, demand driven and participatory training and technical interventions throughout the supply chain Participative and demand driven design of institutions Community-based maintenance technicians who value the opinion of the community
388 Chapter 10. Discussion and Conclusions
10.3.2. The Value of the New Framework, Limitations of the Study and Further Research There are a number of reasons why PV manufacture may be a beneficial industry for industrialising countries to engage in, including its technological proximity to electronics technology, its export potential, its role in providing energy to remote communities and in improving energy security. Research that reveals the factors that influence capability building for PV manufacture in developing countries is therefore of interest. Previous studies on learning in the PV industry have focused on learning rates, the direction of technological change and the national-level policies that can promote the development of a PV industry. These studies have not explicitly linked learning at an enterprise level with institutional factors. They have also paid very little attention to the factors that influence the successful development of PV enterprises and industries in developing countries. It has been suggested in the literature that small scale enterprises have an important role to play in poverty reduction, while small scale PV manufacture could have positive impacts on the capabilities for the installation, maintenance and use of the technology, but there has been little investigation of the circumstances under which small scale manufacturers could build the capabilities required for these positive synergies to emerge. This study has provided a framework that can be used to fill these gaps in the current knowledge and assist decision making in relation to industry development efforts. The framework combines a technological systems approach with an enterprise level view of learning, which allows the identification of factors in technological systems that impact different types of learning, and the relationships between different types of firm-level capabilities and learning processes. Through the application of the framework to the case studies, it has been possible to identify areas where the technological systems in each case have not provided the resources, opportunities and incentives for capability building. As a result general comments have been made about the suitability of different types of countries for hosting PV manufacturing, and the institutional arrangements or interventions that could be used to promote capability building for PV manufacturers in developing countries. The Suntech case study may be of particular interest because this type of study is rare due to the commercial limitations on data collection in the modern sector. The number and scope of case studies carried out for the purposes of this thesis was necessarily limited and too small to make more than tentative conclusions. Additionally, the focus was on the technological extremes of small scale manufacture and modern-sector cell manufacture, while the cases of commercial scale module manufacture and of medium scale urban BOS manufacturers have not been studied. The use of the framework to analyse and
389 Chapter 10. Discussion and Conclusions
compare further cases would enable a better understanding of appropriate ways of supporting manufacture in developing countries. Nevertheless, the framework developed and tested in this thesis may now be used as a tool to systematically analyse the appropriateness of different types of PV manufacture in particular countries, identify the weaknesses in their PV technological systems and therefore being able to suggest where resources should be invested and to suggest appropriate institutional changes. The framework is general enough to be applicable to both the small scale and modern sectors, and could almost certainly be applied to other industries. While there has been a significant amount of research carried out on capability building in modern sector industries, there has been much less work on capability building in the small scale sector. In particular, the technological systems approach has not been used for this type of analysis. The approach used in this thesis may therefore be of interest in relation to studying and supporting small scale manufacturers, particularly when they use relatively complex technologies1. The lessons from the PV case studies may also be applicable to other high technology industries.
1 For example, telecommunications technology is comparable in complexity to PV technology and is now being deployed in remote areas of developing countries. 390 Chapter 10. Discussion and Conclusions
References
Aguilera, J. and Lorenzo, E. (1996), Rural photovoltaic electrification programme on the Bolivian high plateau, Progress in Photovoltaics: Research and Applications, 4 (1), pp 77 - 84. Allal, M. (1999), Working Paper 2: International best practice in micro and small enterprise development, ILO/UNDP Micro and Small Enterprise Development and Poverty Alleviation Project in Thailand, Finnegan, G. (ed), International Labour Organization / United Nations Development Project. Architectural Energy Coorporation (1991), Maintenance and Operation of Stand-Alone Photovoltaic Systems, Sandia National Laboratories. Atmaram, G.H. and Roland, J.D. (2001), Quality Improvement of Photovoltaic Testing Laboratories in Developing Countries, Quality Program for Photovoltaics (QuaP-PV), ASTAE PV GAP. Balaguer, A. and Marinova, D. (2006), Sectoral Transformation in the Photovoltaics Industry in Australia, Germany and Japan: Contrasting the Co-evolution of Actors, Knowledge, Institutions and Markets, Prometheus, 24 (3). Bennell, P. (1999), Learning to change: Skills development among the economically vulnerable and socially excluded in developing countries, Employment and Training Papers: 43, International Labour Organization, Geneva. de Villers, T. (2005), Strategies to increase the confidence in PV and expected impact on the PV Market, Tackling the Quality in Solar Rural Electrification, TaQSolRE. Dunlop, J.P. (1997), Batteries and Charge Control in Stand-Alone Photovoltaic Systems: Fundamentals and Application, Sandia National Laboratories, Albuquerque, NM, USA. Gow, D.D. and Morss, E.R. (1988), The notorious nine: Critical problems in project implementation, World Development, 16 (12), p 1399. Hobday, M. (1995), East Asian latecomer firms: Learning the technology of electronics, World Development, 23 (7), p 1171. Huacuz, J.M. and Gunarante, L. (2003), Photovoltaics and Development, in Luque, A. & Hegedus, S.S. (eds), "Handbook of Photovoltaic Science and Engineering", John Wiley & Sons. IEA (2003), PV for Rural Electrification in Developing Countries – A Guide to Capacity Building Requirements, Deployment of Photovoltaic Technologies: Co-operation with Developing Countries, IEA PVPS Task 9. Katic, I. (2002), Quality Assurance of Solar Home Systems in Nepal, 18th European Photovoltaic Solar Energy Conference, Rome, 7 - 11 October 2002. Katz, J.M. (1984), Technological Innovation, Industrial Organisation and Comparative Advatages of Latin American Metalworking Industries, in Fransman, M. & King, K. (eds), "Technological Capability in the Third World", St. Martin's Press, New York, pp 113-136. Lall, S. (1992), Technological capabilities and industrialization, World Development, 20 (2), p 165. Lall, S. (2000), The Technological Structure and Performance of Developing Country Manufactured Exports, 1985-1998, QEH Working Paper Series – QEHWPS44, Queen Elizabeth House, Oxford University, Oxford, U.K. Lenges, G.M. (2006), Conversation at European PV Conference with, Bruce, A., Dresden. Lüdemann, R. (2005), Experience and Expectation of Silicon Solar Cell Mass Production - Requirements for Next Generation Equipment, 1st International Advanced Photovoltaic Manufacturing Technology Conference, Munich, Germany, April 13th. Norberg-Bohm, V. (2000), Creating Incentives for Environmentally Enhancing Technological Change: Lessons from 30 Years of U.S. Energy Technology Policy, Technological Forecasting and Social Change, 65, pp 125-148.
391 Chapter 10. Discussion and Conclusions
Porter, M.E. (1998), The competitive advantage of nations : with a new introduction, Free Press, New York. Romijn, H. (2001), Technology Support for Small-scale Industry in Developing Countries: A Review of Concepts and Project Practices, Oxford Development Studies, 29 (1). Saxena, N.C. (2003), The Rural Non-Farm Economy in India: Some Policy Issues, Rural Non- Farm Economy and Livelihood Enhancement, DFID-World Bank Collaborative Research Project, Natural Resources Institute. Sharif, N. (1992), Technological dimensions of international cooperation and sustainable development, Technological Forecasting and Social Change, 42 (4), p 367. Solarbuzz (2007), Solar Cell Manufacturing Plants, Accessed from: http://www.solarbuzz.com/Plants.htm, on: Jan 2007. Tani, T. (2003), JICA PV Project - Case of Zimbabwae, in IEA PVPS Task 9 (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA. ter Horst, E. and Zhang, C. (2005), Impacts of Technology Improvement and Quality Assurance in the WB/GEF China Renewable Energy Development Project on PV industry and market development in China, 15th PVSEC. THERMIE (1998), Universal technical standard for solar home systems, Thermie B SUP 995- 96, EC-DGXVII, European Commission Joule-Thermie Programme. Tybout, J.R. (2000), Manufacturing Firms in Developing Countries: How Well Do They Do, and Why?, Journal of Economic Literature, 38 (1), pp 11-44. Varadi, P.F., Domínguez, R. and Schaeffer, L. (2003), Quality Management in Photovoltaics:Quality Control Training Manual for Manufacturers, Quality Program for Photovoltaics (QuaP-PV), ASTAE PV GAP. Vervaart, M.R. and Nieuwenhout, F.D.J. (2000), Solar Home Systems: Manual for the Design and Modification of Solar Home System Components Quality Program for Photovoltaics (QuaP-PV), ECN. World Bank (2006), Doing Business in 2006: Creating Jobs, The World Bank, Washington D.C., U.S.A.
392 Appendix 1: Virtual Production Line CD-ROM
A1-1 Appendix 2: Experiences with the Diffusion of Photovoltaics in Developing Countries
High Cost Although PV is often the least cost alternative form of electricity supply, the cost of PV electricity sometimes exceeds the cost of traditional sources of energy for lighting. Users of Thai photovoltaic battery charging stations reported net additional costs over the alternatives, although they use 40% less dry cell batteries and have reduced their consumption of candles and kerosene (Green, 2004). Clients of Zambian PV energy service companies also pay more for solar electricity than they were previously paying for kerosene, candles and batteries (Gustavsson & Ellegard, 2004). Conversely, there are also many examples of energy cost savings through the use of PV. Users of Brazilian SHS, for example, report that the use of PV for lighting is cheaper than kerosene (Zilles et al., 2000). The very poor cannot afford electricity, whether it is provided via grid extension, photovoltaics or other decentralised means. Their capacity to pay can be increased by subsidised electricity services, or through finance that distributes the cost over the life of the system. The poor, however, usually do not have access to financial services. Finance may be provided through projects or electrification programmes, but even where finance is available, the poorest may not qualify. In the Zimbabwe GEF PV project, for example, since a steady income was required for eligibility for a loan, most of the poor did not qualify (Mulugetta et al., 2000).
Poor Technical Capability
Poor Design In a PV project or programme, a limited number of designs are usually approved by the project developers, who generally have the capabilities to ensure that the system is sized correctly, but limit the choice of the user with respect to system size and capabilities. Cash sale systems may offer more choice to the user, but often include panels that are undersized in relation to the battery (Nieuwenhout et al., 2004). For instance, small vendors of PV systems in Kenya make the majority of system design decisions, but since only 7% of the vendors specialise in PV (Duke et al., 2002), they are not well qualified to design systems. There is often no controller provided, resulting in a battery life of 1-2 years, instead of up to 10 years (Nieuwenhout et al., 2004). Since the battery is the most expensive component in a PV solar home system, battery misuse drastically increases the life cycle cost. It should be noted that an Indonesian study (Retnanestri, 2007) has found that the omission of charge controllers has reduced system costs by $US20 and, when users are willing and able to monitor their demand, has resulted in simpler, less problematic systems.
A2-1 Poor Installation and Maintenance PV systems typically require less maintenance than other decentralised electricity technologies, but system failures caused by poor installation and maintenance have been the one of the major causes of PV program failures. Good maintenance depends on user education, the availability and capabilities of technicians and the availability of spare parts.
Cash Sales In the cash-sales model, the owners may purchase components and install the system themselves. Cash sales usually do not include any maintenance contract, and leave the owner of the system to carry out and pay for any repairs and replacements. Maintenance contracts included with sales can improve the system performance. For instance, in a Sahelian water pumping project, small PV dealers signed a full guarantee maintenance contract, and some of these contracts have been extended by 5-10 years (Shanker, 2003). 5-10 years after installation, 95% of the systems still provide water, and the mean time between failures is 6 years. Even where there is a contract to provide maintenance, problems may arise if businesses fail in the fluctuating PV market created by project ebb and flow. The Zimbabwe GEF project involved small retail businesses and centralised procurement. The installing retailer was required to visit systems at least 3 times in the first 2 years. When the retailers went out of business after the market died down post project, there was no one to maintain the systems (Mulugetta et al., 2000).
Project or Programme Early projects assumed that PV systems would require little maintenance, and often failed to provide financially or institutionally for maintenance. Based on early experiences, projects and programmes now generally include some provision for maintenance, but some projects still neglect this important area. In a recent Thai water pumping project, villagers were not trained to manage or maintain systems and fees were not collected for maintenance and spares. As a result 40% of systems had already failed within 4 years (Kaunmuang et al., 2001). Effective maintenance has proven particularly difficult to achieve in centralised projects, because of the remoteness of the systems and therefore the difficulty in responding quickly to system failures. In a South African EU sponsored schools project, field workers had limited knowledge and were not able to repair systems. They could only report problems to the centralised project management, so that response times were slow (Klunne et al., 2001). Poor communication and transportation as well as lack of funding have constrained the availability of maintenance and spare parts in a Tongan project (Tukunga et al., 2002). Experience has shown that effective maintenance is best provided through locally based technicians who are sufficiently trained and have a stock of spare parts. Spare parts must be available in the field to avoid weeks or months of delay, particularly if the parts must come
A2-2 from overseas. Villagers who obtained systems through the BANPRES-LTSMD projects in Indonesia were unable to replace batteries with high quality ones, since they didn’t know where to buy them (Fitriana et al., 1998). In South African schools, where appliances were not delivered as promised, users lacked interest in the system. In clinics, however, appliances and spare parts were made available through the hospital network (Klunne et al., 2001). The systems in clinics were much better maintained and were much less susceptible to theft. The training of local personnel can improve the maintenance and performance of systems. A local technical assistance scheme whereby technicians were trained via practical classes and installation demonstrations in Brazil has been effective. Technical problems have been solved locally, including replacements of failed parts and systems being moved as required (Zilles et al., 2000). Once trained, however, rural technicians may be better able to get a well paid job in a nearby town. In Tonga, there is a lack of technicians, partly because some of the them have moved to the main island (Tukunga et al., 2002).
Fee for Service In a fee-for service model, the maintenance is carried out by the service provider, since their contract is to provide the electricity, and the user is likely to cease payments if the service is not provided. Zambian energy service companies (ESCOs) have ensured locally available service and care of the batteries (Gustavsson & Ellegard, 2004) by using subsidiaries of existing companies that had rural outposts for other purposes. Wade (2003), after reviewing Pacific island experiences, concluded that full time technicians with no less than 50 systems in their service area are best placed to perform maintenance effectively and efficiently. The technicians must therefore be community-based, rather than nationally based, and a critical mass of demand is required to allow maintenance to be provided in a sustainable manner. Village technicians backed up by networks of rural qualified technicians (in larger towns, no more than 50km from the users) and a central management to train and oversee the technicians have been suggested by de Villers (2005), following a comprehensive review of PV system performance. Field technicians require regular training. In Kiribati, frequent training of local technicians has led to much better maintenance performance than in neighbouring Tuvalu, where training was only provided at the beginning of the programme (Fitriana et al., 1998; Wade, 2003).
Poor Capability to Use System Because PV is an unfamiliar and complex technology, its introduction to rural users is challenging. A connection must be established between the technology and the user. It is well documented that the introduction of technology to poor users takes time, and requires participation (Bhalla & Reddy, 1994; Khosla, 1994; Smillie, 2000; Stewart, 1983).
A2-3 Users need to be trained to use a PV system in accordance with its limitations. In a Zambian fee for service implementation, users were not involved in any service or maintenance and were to report failures to the ESCO. 'Blackouts' in 55% of systems indicate that completely untrained users didn't understand the capacity of system (Gustavsson & Ellegard, 2004). Within four years of 2000 PV radio receivers being distributed to Ethiopian schools for distance education, virtually all the batteries had failed and the schools were not using the equipment, which was otherwise in good working order, because they did not know that the failure was due to the battery or where to obtain new batteries (ESD, 2003). The provision of information and training to users requires careful consideration of cultural factors, especially in remote tribal areas. In remote Thai villages, government programmes involved minimal training of users, consisting only of advice from the installation team to those who were around at the time of the installation. There were infrequent follow up visits to check on the systems, while the technicians and installers didn’t speak the local languages and many villagers didn't speak Thai. Additionally, female users didn’t interact with male installers. As a result the users received virtually no information about the technology (Green, 2004; Sriuthaisiriwong & Kumar, 2001).
Inappropriate Organisational Structures in Projects and Programmes Many projects and programmes have failed due to inappropriate organisational structures. A demonstration mindset has led to unsustainable fee-recovery and business models (Radulovic, 2005). Two main types of project organisation are discussed below: Large-scale, centralised structures, Local structures. Problems have resulted in both of these types of structures. Local structures have encountered problems relating to entrenched power structures and consequent problems with fee collection. In the case of centralised projects, poor micro-macro linkages often result in a lack of connection and consultation with the users in relation to the planning of projects and programmes, which has led to: Project-based timeframes, The use of technology that was not culturally appropriate, Inappropriate fee collection arrangements.
Large-scale, Centralised Structures Because of the dispersed nature of rural PV markets, the provision of after sales service and fee collection is time-consuming and expensive. Projects and programmes may attempt to carry out some of these functions with more centrally based personnel, but problems have been
A2-4 encountered with poor information flows, high costs and poor understanding of local cultural factors in centralised arrangements. In a Senegalese privately operated ESCO, for example, the fee collector lived far away from the users and the expense of fee collection made the programme financially unsustainable (Seck, 2002). The distance from the implementing NGO office to service areas also made management difficult during the JICA ESCO pilot (Tani, 2003a). In Thai battery charging stations, 72% of 50 charge controllers surveyed were not working when inspected between 6 and 10 years later, often due to a blown fuse. Repairs were often not carried out due to lack of funds or maintenance staff, although spare parts are available in nearby towns (Sriuthaisiriwong & Kumar, 2001). Many users in Indonesian government–sponsored PV projects were found to be unaware of their rights and obligations (Nieuwenhout et al., 2004). Some users sold their systems while they were still in the lease phase of a lease-purchase agreement. Others were not aware of their rights to obtain replacement parts under warranty.
Local Structures Fitriana et al. (1998), after a review of the Indonesian BANPRES project, concluded that it is beneficial to integrate a programme into already existing structures in the village, e.g. the village co-operatives. In Iguape-Cananéia in Brazil, traditional support structures were revitalised to manage associations and supportive relationships (Zilles et al., 2000). The associations set up to maintain systems and provide information have a long project timescale and are flexible. The user associations may chose the mode of implementation, and the user is engaged in building battery boxes and support structures. However, local structures have been problematic in some instances, often due to entrenched power structures. In Kiribati, village elders sometimes demanded a share of the monthly fee in the first couple of years, while some users lost the use of their system. It was concluded that management structures need to be locally appropriate and technicians should have sufficient status in the community (Tani, 2003b). In China, it is felt that it is virtually impossible to implement a PV project without working through the local village leaders (Ling et al., 2002). Attempts to implement more local solutions have also encountered difficulties because there is a minimum number of systems that must be installed within a certain area in order to feasibly employ a full time local technician. In Kiribati, a successful ESCO model with a few hundred systems relies on donations of equipment, and cannot be privatised without increasing in size by 5-10 fold (Gillett & Wilkins, 1999).
A2-5 Micro-Macro Linkages in Projects and Programmes Projects and programmes often have poor linkages with user or local groups. Poor infrastructure and remoteness have increased transaction costs and reduced information flows. In Zambia, for example, a government agency has inspected the quality of installations and achieved a high quality, but the high costs of this centrally-based inspection precludes them from inspecting all of the growing number of installations (Ellegard et al., 2004). Poor linkages have been observed between users, installers, maintenance technicians and equipment suppliers. In Zimbabwe, problems have arisen with links between installers and maintainers where these roles are separate. Because contractors carried out the installations, there was no direct relationship between suppliers and the implementing service providers, restricting their opportunities to learn from each other. The contractors who performed the installations did not like to interact with or take advice from better trained service provider technicians who observed (Ellegard et al., 2004). In many cases, users have not been aware of their rights and obligations. In some cases, such as in the Zimbabwe GEF project, the decision making has not involved any public or rural community consultation, but has been confined to government ministries and selected actors (Mulugetta et al., 2000). Lack of transparency and inappropriate project design has resulted. Similarly, in Tonga, the equipment was not designed with user consultation (Tukunga et al., 2002), resulting in conflicts over the level of service provided, such as the number of lights, and the fee charged. Users have been unwilling to pay monthly fees due to a 5% government commission. Only 35% of the fees have been collected. The agendas of different agencies, especially those providing the funds or the government of the country concerned may not be aligned with the interests of the local actors. For example, in the Zambian ESCO project, the focus was to achieve a sustainable project, which was seen to require highest quality components, but through this process some small businesses, which could have been part of the long term PV industry were destroyed (Gustavsson & Ellegard, 2004).
Project-Based Timeframes Projects are inherently short term, so may require a large number of trained personnel for the duration of the project. There is likely to be a shortage of such people. The bureaucracy of projects can also cause lengthy administrative delays. Competitive tender processes caused delays, lost trained staff and customers in a Zambian fee for service implementation process (Ellegard et al., 2004). The Zimbabwean GEF project involved centralised procurement, and approved small enterprises to do installations. There was a three month lag between the application for funding and the installation, which was problematic for small companies (Mulugetta et al., 2000). In China, some PV companies were growing in anticipation of a GEF
A2-6 subsidy. The project was postponed for several years due to national coordination issues, which had a negative impact on the companies (Li, 2003).
Acceptability of Technology User choice can lead to a more appropriate system in terms of price, size and quality. For example, the development of small and inexpensive amorphous silicon (a-Si) modules over the last 15 years for developing country markets has played a crucial role in the growth of solar markets in Kenya, Morocco, and other African countries (Acker & Kammen, 1996; Duke et al., 2002; Kammen, 1996). Small orientation lights have been found to be culturally appropriate and to save half of the investment cost of the system (Nieuwenhout et al., 2004). Low level lights (2W incandescent car dashboard lights) were used in Brazil as night time security and low-level light for evening chats. The light resembles more closely the accustomed level of lighting and does not attract mosquitoes (Zilles et al., 2000). Sustainable markets for PV will ensure the industry viability and maintenance infrastructure. These markets will only emerge if the technology delivers to consumers what they want.
Fee Collection in Projects and Programmes In the Indonesian BANPRES project, the non-payment of users is related not only to the low technical performance of the SHS but primarily to the attitude towards the Presidential Aid Programme. Neither the users nor the KUD staff is convinced that people should have to pay for the presidential aid.(Fitriana et al., 1998) Fee collection within the community can lead to the money being spent on other things, since the systems operate well for the first few years and other needs are more pressing (Wade, 2003). The cooperative decision making model used in Tuvalu meant that users set their own fees and didn’t allow enough for long term replacements and maintenance (Gillett & Wilkins, 1999). In a water pumping project in the African Sahelian countries, fee collection success varied widely and was found to be dependent on local management organisation, not on the capacity to pay. Recovery of maintenance fees have been decreasing, since there is a long time delay before the funds are needed (Shanker, 2003). Fitriana (Fitriana et al., 1998) recommends:
Clarify that paying money is necessary, Make a social commitment (e.g. by forming user groups) to make sure that people intend to pay the money, Make money collection easy and clear, Control the pay-back with clear authorities, Give incentives for proper pay-back,
A2-7 If the money is not paid, find solutions and take action if necessary (important to have an organisational structure to apply the pressure).
Poor Operation of Markets and Institutions When PV systems are sold directly to users for cash, in addition to the problems associated with the common lack of maintenance contracts; inadequate market institutions often result in poor flows of information, contract enforcement and disincentives for the use of PV. Large project-based subsidies can destroy the market for either fee for service or sales based PV provision and hence impede the development of a commercially viable private PV industry (Green, 2004). Projects have often been short term and unsustainable. In Zimbabwe, for example, a large number of systems were installed as part of the GEF project in order to promote the development of a solar market. After the project funding ended, most of the new companies formed to serve the project market did not survive, and some existing businesses were destroyed during the project (Mulugetta et al., 2000). Users are usually unable to judge the quality of the system they are purchasing, and therefore cannot make an informed decision about the price/quality ratio. The cost of the transaction is therefore very high to the user in terms of risk. Duke et al. (2002) note that even Kenyan PV vendors and technicians are usually not able to measure module output with enough accuracy to determine if a module meets its rating. This risk may be mitigated by warranties, brand recognition or standards.
Quality of the Hardware While users have very little basis upon which to judge the quality of hardware, it is much easier to ensure high quality hardware is used in projects or programmes, primarily through standards or the purchase of equipment from known international suppliers. Nevertheless, long term performance of these products has not always been good, as discussed in the section on standards below, and borne out by the comparison between performance of imported and locally produced components in section Error! Reference source not found..
Warranties Warranties for solar home systems may improve the confidence of the customer, but often do not work well because users may not know their rights and some vendors and importers may not honour their obligations (Duke et al., 2002).
Brand Recognition Poor people value appearance and brand status, as well as low cost. For instance, while most people purchasing modules in Xining, Northwest China do not know the difference between single crystalline and multicrystalline modules, they often ask whether the cells and
A2-8 modules are imported or domestically manufactured (Ling et al., 2002). In Mauritania, rural people reacted negatively to PV modules labelled ‘made in China’ (de Villers, 2005). In a well functioning market, users would be able to judge the quality of the product from the brand reputation. Brands that performed well would be selected by users and poor quality products would fail. However, brands can be faked and the dispersed rural market usually does not have sufficient contact with other users to know which products have performed well in the field.
Standards The introduction of minimum standards has the potential to ensure that PV system components are good quality. The effectiveness of standards in reducing risk and improving information flows to users is undermined because poor users do not know what quality certifications and labelling means and these certifications can be faked (Real, 2006). Nieuwenhout et al. notes that standards can curb local market development, which is often the long term aim of interventions. Standards can be very expensive: US$20 000 for module certification (Atmaram & Roland, 2001). This can increase electrification costs and also reduce the will of projects to insist on certification. In general, standards are likely to disadvantage the local industry compared with imports. If the standards are too strict, the number of manufacturers and competition will be reduced. De Villers (2005) recommends that standards should be introduced with consultation and with capacity building and assistance for manufacturers. High equipment standards do not necessarily mean quality of service delivery. Producers that have passed standards have sometimes been found to be unable to sustain production quality (Nieuwenhout et al., 2004), produce products that later fail in the field, or to incorporate design features that are unnecessary or inappropriate to local conditions. There is some debate over whether simple, sturdy low-cost electronic devices for battery charging and lamps; or high-tech, more complex devices that result in longer battery and tube life, and a lower life-cycle cost for the system overall are more appropriate (Huacuz & Gunarante, 2003). The development of standards is aligned with the priorities of technical committees from industrialised countries. There is a lack of field feedback to these committee members, who are far from the field. Nieuwenhout et al. (2004) suggests that standards should be able to be verified at a local level via local testing facilities. These facilities, however, if they are to comply with proposed standards from the global PV GAP program, for example, are also extremely costly to set up. There are only 5 such laboratories accredited worldwide. Local standards that do not need to be as rigorous may be more appropriate and may be incrementally increased over time in line with the capabilities of local manufacturers. International suppliers, however, prefer international standards, so they don’t have to undergo costly compliance in
A2-9 many countries. Projects and interventions are often under pressure from international suppliers to include international standards.
References Acker, R.H. and Kammen, D.M. (1996), The quiet (energy) revolution : Analysing the dissemination of photovoltaic power systems in Kenya, Energy Policy, 24 (1), p 81. Atmaram, G.H. and Roland, J.D. (2001), Quality Improvement of Photovoltaic Testing Laboratories in Developing Countries, Quality Program for Photovoltaics (QuaP-PV), ASTAE PV GAP. Bhalla, A.S. and Reddy, A.K.N. (1994), The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme, Intermediate Technology Pub., London. de Villers, T. (2005), Strategies to increase the confidence in PV and expected impact on the PV Market, Tackling the Quality in Solar Rural Electrification, TaQSolRE. Duke, R.D., Jacobson, A. and Kammen, D.M. (2002), Photovoltaic module quality in the Kenyan solar home systems market, Energy Policy, 30 (6), p 477. Ellegard, A., Arvidson, A., Nordstrom, M., Kalumiana, O.S. and Mwanza, C. (2004), Rural people pay for solar: experiences from the Zambia PV-ESCO project, Renewable Energy, 29 (8), p 1251. ESD (2003), PV Market Chains in East Africa, World Bank Draft Study, Energy for Sustainable Development. Fitriana, I., Kantosa, E., Sudradjat, A., Kuhmann, J., Preiser, K. and Schweizer-Ries, P. (1998), On the Way from Sukatani to the 50 MW Programme - A Socio-Technical Analysis of Solar Home Systems in Indonesia, IEEE 2nd World Conference on Photovoltaic Solar Energy Conversion, Vienna, Austria, 1998. Gillett, B. and Wilkins, G. (1999), Evaluation of the PREP Component : PV Systems for Rural Electrification in Kiribati & Tuvalu, Final report for the European Commission DGVIII Development, AEA Technology Environment (ETSU). Green, D. (2004), Thailand's solar white elephants: an analysis of 15 yr of solar battery charging programmes in northern Thailand, Energy Policy, 32 (6), p 747. Gustavsson, M. and Ellegard, A. (2004), The impact of solar home systems on rural livelihoods. Experiences from the Nyimba Energy Service Company in Zambia, Renewable Energy, 29 (7), p 1059. Huacuz, J.M. and Gunarante, L. (2003), Photovoltaics and Development, in Luque, A. & Hegedus, S.S. (eds), "Handbook of Photovoltaic Science and Engineering", John Wiley & Sons. Kammen, D.M. (1996), Household Power in a New Light: Policy Lessons, and Questions, for Photovoltaic Technology in Africa, Tiempo: Global Warming and the Third World (20), pp 1-8. Kaunmuang, P., Kirtikara, K., Songprakorb, R., Thepa, S. and Suwannakum, T. (2001), Assessment of photovoltaic pumping systems in Thailand - one decade experience, Solar Energy Materials and Solar Cells, 67 (1-4), p 529. Khosla, A. (1994), Effective Design and Diffusion of Rural Technologies, in Bhalla, A.S. & Reddy, A.K.N. (eds), "The technological transformation of rural India : a study prepared for the International Labour Office within the framework of the World Employment Programme", Intermediate Technology Pub., London, p 44. Klunne, W.E., Westra, C.A., Cox, R.L.M. and Gys, L.J.C. (2001), Improving the Avialabilty of PV-Systems at South African Schools and Clincs, 17th European Photovoltaic Solar Energy Conference, Munich, Germany, October 22-26, 2001. Li, Z. (2003), Global Environment Facility (GEF) / World Bank Assisted China Renewable Energy Development Project PV Component, in IEA (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA PVPS Task 9 Deployment of Photovoltaic Technologies: Co-operation with Developing Countries.
A2-10 Ling, S., Twidell, J. and Boardman, B. (2002), Household photovoltaic market in Xining, Qinghai province, China: the role of local PV business, Solar Energy, 73 (4), p 227. Mulugetta, Y., Nhete, T. and Jackson, T. (2000), Photovoltaics in Zimbabwe: lessons from the GEF Solar project, Energy Policy, 28 (14), p 1069. Nieuwenhout, F., de Villers, T., Mate, N. and Aguilera, M.E. (2004), Reliability of PV stand- alone systems for rural electrification, Part 1: Literature Findings, Tackling the Quality in Solar Rural Electrification, TaQSolRE. Radulovic, V. (2005), Are new institutional economics enough? Promoting photovoltaics in India's agricultural sector, Energy Policy, 33 (14), p 1883. Real, M. (2006), Presentation to Workshop: PV for Development - Ensuring highest quality for sustainability and scale-up, 21st European Photovoltaic Solar Energy Conference, Dresden, Germany, 4-8 September, 2006. Retnanestri, M. (2007), Enhancing the Sustainability of Off-Grid Photovoltaic Energy Service Delivery in Indonesia, PhD Thesis, School of Electrical Engineering, The University of New South Wales, Sydney. Seck, L. (2002), Charging up: Senegal, UNDP Special Unit for SSC pp 89-95. Shanker, A. (2003), Lessons from the Regional Solar Programme (RSP) in the Sahelian countries, in IEA (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA PVPS Task 9 Deployment of Photovoltaic Technologies: Co-operation with Developing Countries. Smillie, I. (2000), Mastering the machine revisited : poverty, aid and technology, ITDG Publishing, London. Sriuthaisiriwong, Y. and Kumar, S. (2001), Rural electrification using photovoltaic battery charging stations: a performance study in northern Thailand, Progress in Photovoltaics: Research and Applications, 9 (3), pp 223 - 234. Stewart, F. (1983), Macro-policies for appropriate technology: An introductory classification, International Labour Review, 122 (3), p 279. Tani, T. (2003a), JICA PV Project - Case of Zimbabwae, in IEA PVPS Task 9 (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA. Tani, T. (2003b), PV Rural Electrification in Kiribati, in IEA PVPS Task 9 (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA. Tukunga, T., Healy, S. and Outhred, H. (2002), Experience with PV Lighting Systems in Tonga, Solar 2002 - Australian and New Zealand Solar Energy Society, 2002. Wade, H. (2003), Summary of PV Rural Electrification Experiences in the Pacific Islands, in IEA PVPS Task 9 (ed), "16 Case Studies on the Deployment of Photovoltaic Technologies in Developing Countries", IEA. Zilles, R., Lorenzo, E. and Serpa, P. (2000), From candles to PV electricity: a four-year experience at Iguape-Cananéia, Brazil, Progress in Photovoltaics: Research and Applications, 8 (4), pp 421-434.
A2-11 Appendix 3: An Introduction to Solar Cells
Silicon Material Most solar cells are made from semiconductors (most commonly silicon), whose electrical properties lie between those of conductors, which allow the free flow of current, and insulators, which do not allow current to flow. (A small group of other devices under development, such as dye solar cells, utilise redox type reactions.) Conductors are often described as atoms floating in a sea of mobile electrons. In insulators, all the electrons are tightly bound to the material, so they cannot flow. Semiconductors act as insulators at absolute zero temperatures, but at higher temperatures, extra energy is available to free electrons temporarily. Energy to free electrons can also be provided by electromagnetic radiation. Silicon atoms form covalent bonds in a lattice structure. Since silicon has four electrons in its valence band, it needs four more electrons to be stable. It achieves this by sharing each of it’s electrons with another silicon atom in a covalent bond, hence effectively having eight electrons, as illustrated in Figure 1a.
Figure 1: Covalently bonded Silicon in a crystal lattice, and (b) A ‘hole’: A covalent bond breaks, an electron escapes and a positive charge remains
(a) (b)
If an electron in a semiconductor material is given enough energy, a covalent bond may be broken, and electrons will be free to participate in conduction. When an electron is removed from the covalent bond, there is no longer electrical neutrality in that area of the silicon lattice. There is a net positive charge left behind. The absence of an electron can be viewed as a positively charged ‘hole’ in an analogous fashion to a bubble in a liquid. It seems that the bubble is moving, but actually the liquid is being displaced. The hole is free to move around the lattice in a similar way to the bubble; via displacement of the electrons (or water molecules in the analogy). Charges that are free to move are called ‘charge carriers’. When a charge carrier is created, some energy is required to pull the electron from the covalent bond, and the electron gains some energy in order to escape the bond. The electrons in
A3-1 a semiconductor can be at one of two energy levels, the free level (conduction band) or the bound level (valence band). When an electron is freed from the crystal lattice, it is said to have moved from the valence band to the conduction band. Electrons in the valance band are bound to the crystal lattice, so they are unable to participate in conduction. Electrons in the conduction band are free to move within the lattice, in a similar manner to the ‘sea of electrons’ in a metal. Holes in the valence band are able to participate in conduction. There is a gap between the energy level of the conduction band and the valance band, called the bandgap (EG). EG is hence the additional energy required by an electron to move from the valence band into the conduction band. This concept is illustrated in Figure 2.
Figure 2: Band model for electrons in a solid
In silicon, at temperatures above absolute zero, charge carriers are always being created and also being ‘destroyed’ when electrons and holes meet and recombine into a bond. Energy is released when this happens, mainly in the form of heat. This energy is usually approximately equal to the bandgap – EG, since the electron falls back into the valence band.
Doping Silicon may be ‘doped’ by adding impurities of other atoms in such a way that they become part of the crystal lattice. Atoms with one more valance electron than silicon such as phosphorous, which is a valency 5 element, can be added to the silicon lattice to form an ‘n- type’ material, as illustrated in Figure 3. The n-type material has a surplus of electrons, because phosphorous has five electrons in its outer band which are available for bonding.
A3-2 Figure 3: Silicon crystal doped with (a) Phosphorous to produce n-type material, and (b) Boron to produce p-type material.
(a) (b)
Atoms with one less valence electron than Silicon, such as Boron (a valency 3 element), can be added to the lattice to form a ‘p-type’ material. This p-type material has insufficient electrons to achieve electrical neutrality. Hence p-type materials have excess holes (positive charges), and the material has an aggregate positive charge.
Figure 4: Free carriers and fixed positive charges in an n-type semiconductor
Figure has been removed due to copyright restrictions.
Source: (Honsberg & Bowden, 1999) When bonds are broken in doped material, electrons and holes are created and recombine. In doped material, there is always many more of one type of carrier than the other and the type of carrier with the higher concentration is called a ‘majority carrier’, while the lower concentration carrier is called a ‘minority carrier’. A doped material typically has 1017cm- 3 majority carriers (electrons are the majority carriers in n-type material), and 106cm-3 minority carriers (holes are minority carriers in n-type material) (Honsberg & Bowden, 1999).
A3-3 When carriers are generated (bonds are broken) in a doped material, there is one electron and hole pair created. If the material is a p-type material, there is so many holes already, that the number of holes does not increase significantly. There were very few electrons, so the increase in electrons is very significant, and the new numbers of electrons outweigh the old.
P-N Junctions P-n junctions, the basis for the photovoltaic effect, are formed by placing n-type and p- type materials together. The n-type and p-type materials have an imbalance of electrical charges: excess electrons on the n-type (negative) side, and excess holes on the p-type (positive) side. If charge carriers were not positively or negatively charged, they would randomly move around until they were perfectly mixed, in the same way that gases diffuse throughout a space when released into the air. This is called diffusion. Free charge carriers from the p and n-type materials will cross the junction by diffusion. Since there are less free holes in the n-type material, not many holes will be available to cross the junction by diffusion, but many electrons will. The opposite applies in the p-type material. The net effect is a movement of current from p-type to n-type material, as illustrated in Figure 5.
Figure 5: p-n junction, showing direction of diffusion current.
diffusion current
When electrons are freed from an n-type material, they leave behind positive charges fixed in the lattice which was previously electrically neutral, since the positive charge of the nucleus is not balanced out by the number of electrons. Holes freed from a p-type material result in fixed negative charges in the lattice. When an electron crosses the junction from the n-type material to the p-type material, it will become fixed in the lattice again, as it combines with one of the free holes, and they neutralise each other. The same applies to the holes crossing to the n- type material. A ‘depletion region’ which is devoid of free charges now exists around the junction of the two materials. Only fixed charges exist in the depletion region.
A3-4 Figure 6: Depletion region, electric field formed by fixed charges in the lattice, and resulting drift current. fixed positive charges
depletion electric region field
drift current fixed negative charges
An electric field is formed by the presence of the negative and positive charges on either side of the depletion region. Positive charges (holes) move in the direction of an electric field, while electrons move in the opposite direction. This is known as drift current. The electric field hence pushes free holes that move by diffusion into the depletion region back toward the p-type material, acting in the opposite direction to the diffusion current. This sets up a barrier to further diffusion, and an equilibrium condition is reached. The electric field exists only in the depletion region, and always propels electrons toward the n-type material. The n-type and p-type materials do not end up with equal numbers of free electrons and free holes due to diffusion, because there is an electric field (and hence drift current) stopping them.
Diodes The most basic building block of electronic semiconductor devices is the p-n junction diode, which is simply a p-n junction as discussed in the previous section. The p-n junction diode is like one-way street for current and is used as such in semiconductor circuits. The symbol for a diode is shown in Figure 7.
Figure 7: Diode, showing direction of applied voltage and current flow in ‘on’ state.
In p-n junction diodes, without an applied voltage, diffusion current and drift current balance each other out, leaving a zero net current. When a voltage is applied, the voltage effectively decreases the electric field in the depletion region. The drift current caused by the electric field is decreased, and the diffusion current can flow if the circuit is completed.
A3-5 Figure 8: p-n junction diode, showing electric field in depletion region, and applied voltage - + depletion electric field (reduced region by applied voltage)
applied voltage + If the applied voltage is below the ‘threshold voltage’ of the diode, the electric field in the depletion region prevents current flow. If a high enough positive voltage is applied across it, current flows freely. The diode is hence used as an on-off switch in circuits. The IV curve is a plot of the current versus the voltage. The IV curve of a diode is shown in Figure 9.
Figure 9: Current-Voltage Curve for a Diode
3.5
3
2.5
2
diode 1.5
Current (A) Current at voltage of around 300mV current begins to increase applied voltage rapidly increased, electric 1 applied voltage field in junction decreased, electric decreased, more 0.5 field in junction diffusion current increased, less diffusion current 0 0 50 100 150 200 250 300 350 400 Voltage (mV)
The equation describing the behaviour of a p-n junction with a voltage V applied across it is: = [ nkTqV )/( − ] II 0 exp 1 where
I0 = “dark saturation current” (Amps) V = applied voltage (Volts) n = diode ideality factor (no units) q = electronic charge of an electron = 1.6×10-19 Coulombs = 1eV (electron-Volts) k = Boltzmann’s constant = 1.38×10-23 (joule/degrees Kelvin) T = absolute temperature (degrees Kelvin)
A3-6 The current through the diode depends on the voltage across it, the temperature, Io (which is dependent on the material quality of the diode), and n, which is the diode ideality factor. The relationship between current and voltage is an exponential equation, which means current increases at a faster and faster rate with an increase in voltage. Looking at the relationship between the current and the voltage, it can be seen that the current is near to zero (no current flows) in the diode until the voltage gets to around 500mV. At this point, the current increases rapidly. At short-circuit, when there is no voltage across the diode, there is actually a tiny negative current generated thermally. This is equal to I0 and is called the “dark saturation current”.
How Solar Cells Work A solar cell is formed by creating a p-n junction, so has similar characteristics to a diode. At temperatures above absolute zero, there are thermally generated carriers (electron-hole pairs) in a semiconductor, because electrons can sometimes achieve the necessary energy to move into the conduction band. Through collisions, these carriers are continuously recombining, and new carriers are always being created. The material achieves an equilibrium state if the temperature remains constant. Electromagnetic radiation, however, can give electrons the energy required to create many more electron-hole pairs. This is how solar cells work: carriers are generated by sunlight. The energy in light exists in discrete packets called photons. Different wavelengths of light have photons of different energy. A photon of light must have energy greater than the bandgap EG, in order to move an electron from the valance band into the conduction band and create an electron-hole pair. Some wavelengths of light do not contain enough energy, so the photon passes through the material. Silicon is effectively transparent to these photons. Others have more than enough, but they still give only energy equal to EG to an electron to move it into the conduction band. Excess energy from a high energy photon will be dissipated as heat, effectively wasting the energy. The bandgap of silicon is 1.1eV. The energy (E) in light is inversely proportional to wavelength ( ), according to the equation: 24.1 eVE )( = μλ m)(
Photons with energy Eph < 1.1 (less than the band gap energy EG) do not have enough energy to create electron-hole pairs, passing through it as if it were transparent. Photons with energy equal to the band gap (Eph = 1.1) have just enough energy to create an electron-hole pair and are efficiently absorbed. Photons with energy Eph > 1.1 much greater than the band gap are absorbed and the excess energy is released mainly as heat.
A3-7 Since blue light (0.45 m) has energy of 2.76 eV, and red light (0.8 m) has energy of 1.55eV, the blue light will be much more readily absorbed by the silicon. That means that many blue photons will create charge carriers near the surface of the solar cell, while red photons will be more likely to get further into the material before generating a carrier. Red photons have a higher probability of passing through the material without being absorbed. When light falls on a solar cell, many free carriers are created. If the carriers recombine without crossing the junction (depletion region), the energy is released, and there is no net current flow – no electrical power is collected. Some of the carriers, however, will reach the p-n junction by natural diffusion processes. If the junction is reached by a minority carrier (electron in p-type material, or hole in an n-type material), the minority carrier will be pushed across the junction by the electric field, and will become a majority carrier due to the ‘drift current’ effect. This separation of the electron and hole across the junction prevents them from recombining, and will cause current to flow if the circuit is completed.
Figure 10: Electron and conventional current flow in a solar cell
If a load is connected to the solar cell, a voltage occurs across the terminals of the solar cell. This is equal to the voltage gain by the electrons as they move into the conduction band of the silicon (1.1V in the case of silicon, since the band gap is equal to 1.1eV) minus the voltage drop across the junction (which is dependent on the voltage across the terminals).
A3-8 Figure 11: Solar cell with load connected
So if the voltage drop across the solar cell is 0.5V, there is a voltage of 0.6V across the junction. Now electrical power is generated (power = current × voltage). If the voltage across the solar cell is high enough, it will reduce the effect of the electric field, as in diodes. The diffusion current will then flow more freely in opposition to the drift current, and the output current of the solar cell will be reduced.
Solar Cell Characteristics The IV curve of a solar cell is the same as that for a diode, since it is also a p-n junction. Illumination of the solar cell adds to the dark current in the diode, so the curve is shifted by an amount equal to the light generated current IL. When operating at a particular temperature, the equation for the IV characteristic of a solar cell is hence:
= II {}exp nkTqV )/( 1 −− I 0 L , where :
IL = light generated current = short circuit current V = voltage across the solar cell (Volts)
I0 = “dark saturation current” (Amps) n = diode ideality factor (no units) q = electronic charge of an electron = 1.6×10-19 Coulombs = 1eV (electron-Volts) k = Boltzmann’s constant = 1.38×10-23 (joule/degrees Kelvin) T = absolute temperature (degrees Kelvin)
A3-9 Figure 12: IV curve for a diode shifted down by IL
IV curve of solar cell & diode
3.5
2.5
1.5
0.5 Solar Cell Diode -0.5 0 100 200 300 400 Current (A) Current
applied voltage -1.5 applied voltage decreased, electric increased, electric field in junction IL field in junction increased, more -2.5 decreased, less carriers collected carriers collected (drift current) -3.5 Voltage (mV)
For a solar cell, with nkT/q approximately 0.026V (at 25 C), I0 = 0.002mA and IL = around 3 A, the equation becomes I = {}V 026.0/ −− 31exp000002.0 , and the IV curve is shown in Figure 13. Note that we flip the curve, so that the current, which is negative in the equation, is in the positive quadrant.
Figure 13: IV Curve for a solar cell, showing Voc and Isc
3.5
Isc 3
2.5
2
1.5 Current (A) Current 1
0.5
0 0 100 200 300 Voc 400 Voltage (mV)
The ‘equivalent circuit’ used to model a solar cell is a diode, with a current source in parallel which represents the light generated current. The current from a solar cell is equal to the light generated current (IL) minus the diode current.
A3-10 Figure 14: Equivalent Circuit of an Ideal Solar Cell
The open-circuit voltage (Voc) of a solar cell is the voltage that will be found across the terminals of the cell when they are open circuited. This is the maximum voltage from a solar cell, and since we have an open circuit, no current will flow.
Figure 15: Solar Cell at open circuit
Voc can be found from the equation above by setting the current I = 0. nkT C I S Voc = D L + 1ln T q E I 0 U
The short-circuit current (Isc) is the current that flows when the terminals of a solar cell are short-circuited. This is the maximum current that can be extracted from the solar cell. At short circuit, the voltage across the solar cell is zero. The 1.1V gained by electrons moving into the conduction band is entirely lost across the junction. Thus Isc is approximately equal to IL.
There is also a negligible contribution from the dark saturation current of the diode (I0): Isc =
IL+I0. For each point on the IV curve, the power output is the product of the current and the voltage (I×V). The maximum power point is the point on the IV curve where I×V is at its maximum. This is the point where the rectangle of the largest area can be drawn under the IV curve.
A3-11 Figure 16: IV Curve and Maximum Power Point for solar cell
3.5
3
2.5
2 Max Power 1.5 = 1.4 W
1
Current (A) and Power (W) and (A) Current 0.5
0 00.20.40.60.8 Voltage (V)
In the IV curve above, the maximum power point can be identified as the peak of the power curve. The efficiency of a solar cell is the power output as a ratio of the power input. PowerOutpu Wt )( Efficiency(%) = IncidentRa Wdiation )( Much of the light that falls on a solar cell is not converted to electrical energy. This is because: some light will be reflected, some light will not generate a charge carrier, electron hole pairs will be created that will recombine and do not contribute to the current, electron hole pairs will be created by photons with excess energy that will be wasted.
Parasitic Losses Parasitic effects are electrical losses in the solar cell. There are two kinds of parasitic losses, shunt resistance and series resistance. Shunt resistance is caused by the existence of an alternative (shunt) path for the current to flow, as indicated in Figure 17. When crystalline solar cells are made, the creation of the top (n-type) layer also creates an n-type path on the sides of the wafer. If this area is not removed correctly, a shunt path exists. Short circuits (shunt paths) may also be caused by imperfections in the depletion region. The lower the shunt resistance, the more current can be lost through this path.
A3-12 Figure 17: Equivalent Circuit with Shunt Resistance
If the shunt resistance is infinite, there is no shunt path. The lower the shunt resistance, the higher the conductance of the shunt path, and the more current can be lost. Shunting causes a reduction in the fill factor and the maximum power point, but unless the shunt is quite bad, the short-circuit current, Isc and the open-circuit voltage, Voc will not change appreciably. If the shunting is bad, the open-circuit voltage will start to deteriorate.
Figure 18: The Impact of Shunt Resistance on a Solar Cell
3.5
3
2.5
Rsh infinite 2 Rsh = 3 A Rsh = 1 1.5 Rsh = 0.3
1
0.5
0 0 0.2 0.4 0.6 V
Series resistance is caused by resistive losses in the solar cell material, the metal contacts and the metal-silicon interfaces. It can be modelled by the addition of a resistor in series with the solar cell, as shown in Figure 19.
Figure 19: Equivalent Circuit with Series Resistance
Series resistance reduces the maximum power point of the solar cell as illustrated by Figure 20.
A3-13 Figure 20: The Impact of Series Resistance on a Solar Cell
3.5
3
2.5
Rs = 0 2 Rs = 0.015 A Rs = 0.03 1.5 Rs = 0.08 Rs = 0.15
1
0.5
0 0 0.2 0.4 0.6 V
Optical Losses Optical losses in a solar cell result in some of the available light failing to enter the solar cell to generate charge carriers. These are caused by shading of the top surface (10-15%), and reflection from the surface of the cell. The reflection from silicon is approximately 33% of incident irradiation. Measures such as texturing the silicon with a chemical etch and the application of an anti-reflection coating are commonly used to reduce surface reflection.
Spectral Response Since blue light has a lot of energy, it will be easily absorbed and will be absorbed close to the surface of a solar cell. To reduce the series resistance losses at the silicon metal interface, the top n-type layer of a crystalline silicon solar cell is heavily doped. This means that charge carriers in this area recombine readily and are not converted to current. This is why blue light is not well utilised (Figure 21).
Figure 21: Spectral Response of a Solar Cell
Figure has been removed due to copyright restrictions.
Source: (Honsberg & Bowden, 1999)
A3-14
The infrared and red response is also poor, since some of the light will pass straight through the silicon, and some of it may recombine before it makes it all the way to the junction. Photons with more energy are more likely to be absorbed sooner (closer to the junction), and have a shorter distance to travel and are less likely to recombine. Treatment of the back surface to reduce recombination can also deplete the red response. Light of wavelength of 1.127 m and above does not have enough energy to overcome the bandgap, and is not absorbed at all.
A3-15 Appendix 5: Suni Encapsulation Procedure
This document describes a new encapsulation procedure under development at Suni Solar in Managua. The new procedure uses a glass-silicone-cell-silicone-plastic-glass structure. In the past, Suni has used glass-silicone-cell-silicone-Mylar™, but the substitution of glass for Mylar™ will circumvent the use of imported products for which the supply can be irregular and expensive. In addition, the use of two sheets of glass gives the module extra strength, and is being tested for large modules that can benefit from (greater than 50 Wp). The silicone used is Sylgard™ 184, which is purchased from the USA for around US$250 per 4 kg tub and transported in passenger luggage when visitors come from the US. The solar cells are reject cells obtained very cheaply from a US solar cell manufacturer. Normal window glass is used, since it is easily available and cheaper than tempered glass. Additionally, Suni finds that modules made with tempered glass are difficult to repair if the glass shatters.
Step 1: Assemble aluminium frame & glass front The module frame is made from square aluminium tubing. The square tubing, in place of ‘C’ shaped framing, has been adopted to give greater rigidity to the module. Since window glass is more susceptible to cracking due to flexing, the use of window glass instead of tempered glass makes the rigidity of the structure more critical. At each corner, two pieces of tubing with a 45 angle are joined to make a right angle by pop-riveting a right angled piece of aluminium to the inside corner of each piece.
Figure 1: Frame with glass attached with silicone sealant (strings of cells are only placed on glass in preparation for measuring).
3 mm Glass is attached to the frame using white silicone sealant. 5 mm glass has been used in the glass-Mylar modules manufactured in the past, but two sheets of 5 mm glass results
A5-1 in a very heavy module. Two sheets of 3 mm glass gives sufficient strength without excess weight.
Step 2: Solder strings of solar cells The solar cells used by Suni are large (15cm x 15cm) single-crystalline cells with an anti-reflection coating applied. The panels are used for 12V lead-acid battery charging, so have 36 cells (36 x 0.55V per cell = 19.8V). Since each cell is approx 0.018m2, 36 cells would give 0.65m2. If the efficiency of the cell is 14% at maximum power point and 1000W/m2, 0.65m2 would produce around 90Wp. Half cells are used to achieve the correct voltage for battery charging in a 45W module. Cells are cut in half using a diamond tipped disc attached to a hand tool. The cells used are reject cells that are broken or part of the cell is damaged or with defects. The good parts of the cells can be saved. The cells to be used are chosen by visually inspecting the available cells and choosing the best ones in terms of quality of screen printing, uniformity and colour of anti- reflection coating and avoidance of cells with cracks, scratches or other defects.
Figure 2: Cell-cutting using diamond-tipped blade.
Strings of cells are hand-soldered. Solder is melted onto the first third of the screen- printed silver busbars on the front of a cell. A piece of interconnect tape approximately two- thirds of a cell length is then pressed onto the solder using a hot soldering iron. Extra solder is added to the top of the interconnect tape to ensure good ohmic contact.
A5-2 Figure 3: Priming silver solder tab with solder prior to soldering interconnect tape to back of cell.
The cells being used by Suni have strips that are left bare when screen-printing the aluminium back contact of the cell. Silver solder tabs have been separately screen-printed along these strips, leaving a gap of about 5 mm between cells. The excess interconnect tape from the front of each cell is soldered to the solder tab at the back of the next cell to create a series string. Around 5mm of interconnect tape is left between each pair of cells to allow for stress-relief loops that will accommodate the expansion of the gap between cells at high temperatures, due to the higher expansion coefficient of glass compared to cells.
Figure 4: Stress-relief loop in cell interconnect.
A5-3 Figure 5: Joining cells in series.
Step 3: Test current & voltage of strings After the strings of cells are joined, they are taken into the sun to check the short circuit current and open circuit voltage of the strings. When cells are connected in series, the voltage of the string is equal to the sum of the voltages across each cell. The current is the same through the entire string, and is determined by the worst cell. It is hence imperative that cells are well matched, in order to avoid efficiency losses.
Figure 6: Resultant IV-curve (black), showing Isc and Voc for two cells (blue and red) connected in series.
A multimeter is used to measure the voltage across each string at open circuit. If, when measured, the voltage is not close to the expected value of around half a volt per cell (at around 33 C), each cell is tested to determine which cell(s) are bad.
A5-4 Figure 7: Measuring voltage across a string.
The multimeter is also used to measure the current running through a string. When a string is found to have lower current output than the other strings at the same irradiation level, each cell in the string is short-circuited using the multimeter, and the current is read off again.
Figure 8: Finding a bad cell in a string that produces lower than expected current.
In this way, it can be determined which cells are comparatively weaker than others, although the readings of Isc are in themselves inaccurate. Due to the internal resistance of the multimeter and the leads (around 0.3 ), the measurement taken is not at zero volts. For a string of 18 cells, the measurement is close to short circuit, so is a good indication, but as the number of cells in the string decreases, the accuracy of the reading decreases.
A5-5 Figure 9: IV Curve of (a) solar cell and (b) string of 18 cells showing possible error in current measurement due to internal resistance in multimeter and resistance of leads.
2.5 2.5
2 2
1.5 Current (A) 1.5 String of 18 Cells
Resistance = 0.3 (A) Resistance = 0.3 1 1
0.5 0.5
0 0 0 100 200 300 400 500 600 0246810 (mV) (V)
Step 4: Mix RTV The two-part RTV silicone (Sylgard™ 184) is prepared by mixing ¾ of a one-eighth a gallon vessel of base with 3-4ml of curing agent. It has been found that the curing time is 48 hours with the proportions used. The curing time can be halved to 24 hours if 6ml of curing agent is used, however the glass-silicone bond strength is poor and delamination results.
Figure 10: Preparing the RTV silicone
The silicone is rested for 15 minutes to allow the bubbles to come out of the liquid before use.
Step 5: Pour RTV onto glass Silicone is poured onto the glass, and the frame is lifted into a vertical position to allow the silicone to cover more of the glass surface.
A5-6 Figure 11: Using gravity to spread the silicone.
The module is then placed flat again, and the silicone is spread to cover the entire glass surface. A gap of a few millimetres is left at each side in order to save materials, since the cells will not be in this area of the module, and hence it is not integral to the moisture seal.
Figure 12: Spreading the silicone to cover the glass surface.
Most of the bubbles in the silicone are removed by blowing air onto the bubble to allow it to come to the surface, or by piercing it with a screwdriver.
A5-7 Figure 13: Removing bubbles from the silicone.
Step 6: Place strings into RTV Strings of cell are placed cell-by-cell into the wet silicone. Each cell is pressed and massaged into the viscous liquid to allow bubbles to escape from underneath the cell.
Figure 14: Placing strings into the module frame.
The front of the module is checked for any bubbles, and the cells are further massaged if there are bubbles.
Step 7: Pour second layer of RTV The second layer of RTV silicone is poured onto the back of the cells in the module frame. The silicone is spread with the fingers and pushed carefully around each cell. The cells may settle further into the first layer of silicone via the pressure applied at this stage. It is ensured that the cells are aligned properly at this stage, and the relief loop is left in the interconnect tape between cells.
A5-8 Figure 15: Massaging the silicone over the back of the cells.
The module is placed horizontally and the level is checked. The silicone is allowed to dry in this position indoors for 2 hours to ensure flatness. The module is then taken into the sun and the silicone cures for a further 10 minutes.
Step 8: Solder strings together & attach contacts Once the silicone has cured, the strings are soldered together in series using a double layer of interconnect tape.
Figure 16: Soldering the strings together.
A5-9 Figure 17: String connections are not encapsulated with the cells.
The strings are soldered together after the silicone has already cured, so are not included in the watertight seal. It was suggested that the soldering could be done while the silicone was still wet and then the interconnect pushed into the silicone, but the soldering is very difficult to perform when the silicone is wet, and the silicone interferes with the solder, increasing the likely hood of a bad contact. Another solution would be to solder the strings together after the silicone has cured, and then push them flat against the silicone & add a small amount of silicone to cover the interconnection. This would ensure the integrity of the moisture-proof seal.
Step 9: Apply white plastic and glass to back of panel White plastic is used to separate the silicone and the glass, since Suni considers it very important to be able to access the cells for repairs. In commercial manufacturing, it is assumed that the cells will never need replacing and that the moisture seal will last around 30 years. If the glass on a module is broken, the module is generally discarded (tempered glass shatters). If for any reason the moisture seal is breached and corrosion occurs, the module is under warranty.
A5-10 Figure 18: Applying White Plastic Barrier.
A soft cloth is wrapped around a wide spatula and as much air as possible is pushed out from between the plastic and the silicone. The plastic has the tendency to stick to the silicone.
Figure 19: Placing Piece of Glass onto Back of Module.
Glass is then added to the back of the module. An opening has been cut in the glass to allow the wires to exit the module. Although the glass does not participate in the watertight barrier of the module, it provides protection for the silicone, and additional mechanical strength to the module. In order to hold the glass in place, an aluminium “L” is attached to the frame. Additional plastic is placed under the frame in order to cushion the glass to prevent cracking when the “L” is riveted in place, and also to increase the watertightness of the module.
A5-11 Figure 20: Aluminium “L” Used to Hold Glass in Place.
The “L” of aluminium is riveted to the square tubing of the frame. Care must be taken not to put to much pressure on the glass, since the silicone surface underneath may be uneven and the glass is in danger of cracking.
Figure 21: “L” Riveted to Frame Holds Glass in Place.
Excess plastic is trimmed from the edges of the “L” with a Stanley knife. The last step is to melt plastic into the opening in the glass backsheet around the wires to further increase the watertightness of the module.
A5-12 Figure 22: Sealing the Hole in the Glass with Melted Plastic.
The Finished Module The finished module looks clean and neatly aligned. There are only tiny bubbles in the silicone encapsulant. It is not clear whether these will pose a problem in the long term or not. It is possible that over time the expansion and contraction of the bubbles as the temperature fluctuations cause the gases to expand and contact will cause the silicone to lift from the glass and allow moisture ingress. Keeping water out is one of the most crucial functions of the module encapsulation, since failures in modules are commonly caused by moisture entering the module followed by corrosion.
Figure 23: Finished Module.
A5-13 The makers were critical of this particular module, because there was a small amount of discolouration around the soldering of the interconnect tape that joins the strings. This was caused by burning of the silicone during the soldering. This is a purely aesthetic concern.
Figure 24: Messy Solder Joints & Discoloured Silicone.
A5-14 Appendix 6: Cost of Materials for Grupo Fenix Module Manufacture
Cost & Quantity of Materials Needed for a 50W panel in 2003 (14 Cordobas = US$1) Cost/Unit Quantity for Source Transport 50W panel Cells US$2/watt sold goes to 36 cells – 6 donated by Skyheat Crate from the Grupo Fénix. cm2 US regularly sent by an NGO. RTV silicone FBC funds paid for visitors from the Silicone out of project United States funds. In Sabana Grande, the material has never been purchased. The solar culture course pays for some materials Glass 20 Cordobas per square 50W panel = Bought especially for Transport would foot (3mm) and 22 4.5 square feet each order – no be difficult in Cordobas per square foot storage, since panel large quantities. (4mm), sizes are different. Bought from Ocotal or Managua. Broken bits of glass used to make 1.5V battery chargers. Frame $US25/21 feet 3 50W panels Alumicentro, Collected by Cutting – 2 Cordobas/cut can be made Managua employee: or 3 Cordobas for 45˚ from one tube Marco-Antonio or angle Mauro bring Al in a bus to Sabana new tube: 180 Grande. 180 Cordobas/21 feet (this is Cordobas for bus sawed in half to make 42 & taxi fares. feet) saw blades 12 Cordobas each 1/panel Hardware store for cutting frame sandpaper 5 Cordobas for paper 2 pieces paper Hardware store back, 15 Cordobas for back/panel or 1 cloth back piece/3 panels cloth back interconnect donated 3m/panel (was donated by Skyheat tape 1.5 when making smaller solder) Solder 120 Cordobas/1 lb Hardware store Collected by employee Sealant 60 Cordobas/296 ml one tube Hardware store Collected by makes about 6 employee panels Rivets 3 Cordobas/dozen 2 dozen/panel Hardware store Collected by employee L-profile 360 Cordobas/21 feet 6 inches/panel Mangaua Collected by employee small L- 45 Cordobas/21 feet 2.5m/panel Mangaua Collected by profile employee backing $US8/yard Paper supplier Collected by (drafting film) employee
A6-1 Appendix 7: SWRC/EU/UNDP Training Modules 2000
Delivery of training: The training took place over five days in the case of VEEC training and four days for the women’s group, with a morning and afternoon session in each day of the VEEC training and two or three sessions per day of the women’s training. Thirty participants were involved in activities such as ice-breakers as well as meditation, songs, plays and games throughout the sessions to relax, engage and involve all participants. Combinations of small group and large group discussions were used to bring the ideas of all participants into the discussion. All participants were encouraged to narrate their experiences and present their approaches. Solidarity was promoted through the awareness of shared experiences.
Summary of content – VEEC Training: Session 1. Development scene in villages & village dynamics What basic facilities are needed in the villages? Why these facilities are not in villages whereas there are so many agencies to provide it (government as the main agency)? Situation in the villages and the role of government / public representatives in it. Who is responsible for these situations? To what extent is the community responsible? Power vested in the hands of the public is transferred to whom? (Rich, poor, high caste, low caste, men, women) How can we change this scenario?
Session 2. Role of community in development Exploration of frictions in villages, vested interests of few who are interested in their own benefits: o Why the village as a whole does not prosper. o The need for organisation. o The role and nature of organisation. o The importance of participation of women.
Session 3. The role of government, SWRC and village level community organisation Opening a primary school. Electrification of houses. Starting a self-help group. Ensuring visit of ANM in village on a regular basis. The importance and need for resource mapping and survey.
A7-1 Session 4. SWRC Barefoot concept. Values, approach, priorities. Discussion.
Session 5. Project Introduction and Description What is solar energy? Uses of solar energy. Introduction of the EU-UNDP project of solar electrification. Role of the community in approaching it. Role of community in execution (resource mapping survey etc.). Aspects of projects (technical, REW, BSE, solar power usage) Financial sustainability (self-reliance, maintenance, replacement, salary of BSE…) Environmental aspects
Session 6. Resource mapping Conducting resource mapping practically in a village. Analysis and final map. Assessing need for the survey.
Session 7. Environmental issues The relationship between nature and life. The effect of pollution on human life and nature. Environmental problems in villages. The need for environmental awareness and awareness campaigns. Eco-friendly nature of solar energy and environmental aspects in the project.
Session 8. Action plan How they will proceed when they go back to their villages. Common points among villages. Each village prepares an action plan and 1-2 issues through which it will initiate organisation in the village.
Session 9. Free Session Questions from participants Songs and cultural programme
A7-2 Session 10. Training Evaluation Presentation of action plan Evaluation of the training
Summary of content – Women’s Group Training:
Session 1. Opening Up
Session 2. Perception of one’s own problems Problems that women perceive at the village level. Problems relating to women. Problems relating to low castes in village. Place of women in society. Why women don’t perceive their own problems. Power structures in the village.
Session 3. Position of women Role of women in family. Role of women in Panchayat. Role of women in agriculture. Role of women in religion. Role of women in decision making. Women in administration and bureaucracy. Including ratios of women/men, literacy, crime. Why these situations. Effect of power structures on women. Need for organisation of women.
Session 4. Organisation of women and its role Importance of women’s organisation Nature and ways of strengthening the organisation. Potential for field trips to gather information and help women.
Session 5. Women’s empowerment Examples of issues taken up by women and won.
Session 6. Women’s group role
A7-3 Preparation of chart “what women can do in the village”.
Session 7. Solar project Solar Energy. Project. Approach of SWRC. Role of village level organisation (VEEC) in project. How to reach people, especially poor. Resource mapping and survey. Financial aspects and sustainability. Productive applications of solar energy for better life.
Session 8. Action plan Preparation of action plan for work in the village. How to approach other women. Sensitive issues in the village. Methods for problem solving. Approach for decision making regarding solar project.
Session 9. Consolidation of training Topics discussed consolidated – need for strong women and role in development Evaluation of training. Cultural activities, unification.
Village Level Action Plan (Suggested): 1. Selection of the village. 2. Identification of the villagers. 3. Organisation of the women’s group. 4. Organisation of the village group. 5. Organisation of the VEECs. 6. Training of the women’s group. 7. Training of the VEECs. 8. Resource mapping and survey (village level). 9. Analysis survey. 10. Action plan of the VEECs. 11. Opening of bank account. 12. Selection of the BSEs.
A7-4 13. Monthly meetings of the VEECs (monitoring and follow up). a. Regular contributions. b. Accounts/stocks. c. Whether the SHS and lanterns are working properly. d. Need for components. e. Follow-up of productive applications. f. Follow-up of BSE’s work. g. Development activities. h. Minutes record. 14. Training of BSEs. 15. Productive activities. a. Which activities are feasible. b. Survey. c. Need for further information. d. Name of the village. 16. Solar workshop. a. Building of the workshop. b. Position of the construction. c. List of equipment required. d. Upgrading of the workshop. e. Opening of a new workshop. f. Master barefoot engineer. 17. Environment campaign. a. Identify environment problems b. Discussion with VEECs. c. Planning for the activities. d. Follow up of the activities. 18. Collaboration with other NGOs. a. Identify other organisations in the working area. b. Coordinate with organisations. c. Seek possibilities for joint activities. d. Organise joint activities. 19. Coordination with government officials and local government representatives. a. Organise meetings with government officials and local government authorities. b. Inform government officials and authorities about the project activities. c. Collect information on the development activities of he government.
A7-5 20. Marketing outlets. a. Ensure the quality of equipment fabricated at the workshop. b. Organise service for the maintenance of equipment. c. Establish shops for sales and repair of the equipment. d. Ensure availability of components.
A7-6 Appendix 8: Barefoot Solar Engineer Final Test Paper – Theoretical Component
1. Which of the following is not part of a solar unit? a. Solar charger b. Solar panel c. Battery d. Transistor
2. A 2.2 k MFR will have which colours? a. red, black, red, brown b. red, red, brown, black c. red, red, black, brown d. red, red, red, red
3. What would be the value of a resistor of the colours blue, grey, black, red? a. 65 k CFR b. 82 k MFR c. 68 k CFR d. 68 k MFR
4. A 100 CFR will have which colours? a. brown, black, brown b. brown, black, black c. brown, brown, black d. black, brown, black
5. A 56 k MFR will have which colours? a. green, blue, black, black b. green, blue, black, red c. green, blue, black, yellow d. green, black, blue, black
6. A 270 k MFR will have which colours? a. red, purple, red, black b. red, purple, black, red c. red, purple, black, orange d. red, purple, red, orange
7. What would be the value of a resistor of the colours brown, green, black, brown? a. 15 k MFR b. 150 k MFR c. 1.5 k MFR d. 1500 k MFR
8. What is the function of resistance? a. It increases voltage. b. It increases current. c. It controls the voltage. d. It does nothing.
9. What is the function of a diode?
A8-1 a. It helps current flow in both directions. b. It controls current flow in one direction. c. It controls current in the reverse direction. d. None of the above.
10. What is the function of a transistor? a. It increases voltage. b. It increases current. c. It does the function of an on-off switch. d. It decreases current.
11. What is an I.C.? a. It functions like a diode. b. It functions like a resistance. c. It functions like a transistor. d. All of the above.
12. What is a potentiometer? a. It functions like a diode. b. It functions like a transistor. c. It functions like a resistance regulation. d. It functions like an on-off switch?
13. What is the function of a transformer in an inverter? a. It increases a low voltage to a higher voltage. b. It decreases a high voltage to a lower voltage. c. It keeps the voltage level constant. d. All of the above.
14. What is the function of a fuse? a. It protects against a short circuit. b. It protects the circuit. c. It prevents overload. d. All of the above.
15. If the battery is fully charged, what would be the specific gravity? a. 1120 b. 1100 c. 1200 d. 1250
16. What would be the approximate life of a solar panel? a. 10 years b. 65 years c. 20 years d. 2 years
17. What would be the expected life of a 12 V, 7 Ah lead acid battery? a. 10 years b. 65 years c. 20 years d. 3 years
18. What would be the life of a 12 V, 75 Ah tubular plate battery? a. 10 years
A8-2 b. 5 years c. 20 years d. 2 years
19. How long would you keep a solar lantern on? a. Until discharged b. Until morning c. 3 hours d. 5 hours
20. How many cells are there in a 36 W solar module? a. 20 b. 30 c. 36 d. 24
21. What is not the function of a charge controller? a. Prevent low voltage b. Prevent overcharging c. Charge a battery from a solar panel d. Light a lamp
22. What is the low voltage cutoff? a. 11.2 V b. 12.4 V c. 14.4 V d. 14.8 V
23. What is the overcharging cutoff? a. 11.2 V b. 12.4 V c. 14.4 V d. 14.8 V
24. When is the voltage o.k.? a. 11.2 V b. 12.4 V c. 14.4 V d. 14.8 V
25. What is the correct order of connection for a solar home lighting system? a. panel, load, battery b. battery, load, panel c. panel, battery, load d. load, panel, battery
26. With what do you measure the specific gravity of a battery? a. multimeter b. solar panel c. thermometer d. hydrometer
27. What is the “solvent” in a battery? a. distilled water b. salty water
A8-3 c. mineral water d. water from a well or handpump
28. In which direction would you face the cells of a solar module? a. east b. west c. north d. south
29. What is the voltage of one cell of a solar module? a. 2 V b. 1 V c. 1.5 V d. 0.5 V
30. When should you put water in a battery? a. When the water level goes below the plate. b. Every day. c. Once a month. d. Once a year.
31. What is the reason for a solar lantern giving light for a short time? a. Panel not facing the right direction. b. Panel being covered by a layer of dust. c. Panel being upside down. d. All of the above.
32. Which of the following is the wattage of a module for a solar home lighting system? a. 60 W b. 37 W c. 10 W d. 30 W
33. What is the purpose of using a CFL? a. Low power usage. b. More light. c. Longer life. d. All of the above.
34. What is the wattage of a module for a solar lantern? a. 10 W b. 20 W c. 30 W d. 40 W
35. What is a solar cell made of? a. gold b. ? c. silicon d. copper
36. How many cells are there is a 12 V battery? a. 12 b. 24 c. 6
A8-4 d. 9
37. What is the specific gravity of the acid in a lead acid battery? a. 1200 b. 1100 c. 1230 d. 1300
38. What is the current through a 9 W, 12 V lamp? a. 0.7 A b. 1 A c. 2 A d. 0.5 A
39. What is the current through a 7 W, 12 V lamp? a. 0.58 A b. 0.68 A c. 0.78 A d. 0.5 A
40. If you have 2 batteries, each of 6V connected in series, what is the output voltage? a. 6 V b. 12 V c. 18 V d. 24 V
41. If you have 2 batteries, each of 6V connected in parallel, what is the output voltage? a. 6 V b. 12 V c. 18 V d. 24 V
42. If you have 4 batteries, each of 6V connected in series, what is the output voltage? a. 6 V b. 12 V c. 48 V d. 24 V
43. What is the wattage used by a 12 V, 1 A load? a. 11 W b. 14 W c. 13 W d. 12 W
44. What should be the frequency of an inverter? a. between 21 and 35 b. between 11 and 25 c. between 31 and 45 d. between 41 and 55
45. How many ohms are there in 1 k ? a. 500 b. 1000 c. 1500 d. 2000
A8-5
46. How many ohms in 1 M ? a. 10 000 b. 1 000 000 c. 1 500 000 d. 2 000 000
47. How many turns are there in the 28 gauge primary winding of the transformer in an inverter? a. 7 + 7 b. 8 + 8 c. 9 + 9 d. 10 + 10
48. How many turns are there in the 30 gauge base winding of the transformer in an inverter? a. 4 + 4 b. 3 + 3 c. 2 + 2 d. 1 + 1
49. How many turns are there in the 32 gauge secondary winding of the transformer in an inverter? a. 180 b. 160 c. 100 d. 120
50. A choke coil has: a. 65 turns of 20 gauge wire b. 65 turns of 30 gauge wire c. 65 turns of 32 gauge wire d. 100 turns of 34 gauge wire
51. What is a volt?
52. What is a watt?
53. What is an ampere?
54. What is a solar lantern?
55. What are the components in a solar unit?
56. What is the function of a solar panel?
57. What are the components in an inverter?
58. What are the tools you would require to work in an electronics workshop?
59. What are the precautions you should take when working with a multimeter?
60. What is an electrolyte?
A8-6 61. During the installation of a solar panel, what are the things you need to be careful about?
62. As a BSE, what would be your role in the organisation or community?
63. In your field of work, what is your duty towards the community?
64. What is your duty towards solar power?
65. What do you mean by technology?
66. What is an npn transistor?
67. What is a pnp transistor?
68. Name the different parts of a transistor.
69. What is the function of a capacitor?
70. What are the things to be careful about when soldering?
71. What is the purpose of using petroleum jelly on the terminals of a battery?
72. What are the things to be careful about when making connections to a battery?
73. What would you do to keep a hydrometer safe?
74. Clearly explain the difference between a zener diode and an ordinary diode.
75. What is the full form for IC? Also write the pin numbers on the figure below:
76. What is the full form of PCB?
77. What is the ampere rating of the fuse used in a solar lantern?
78. What is the ampere rating of the fuse used in a home lighting system?
79. What should be the distance between the battery bank and the lamps and what gauge wire should be used for the purpose?
80. What is the significance of the red LED being on and what is to be done when it is on?
81. What are the things to be checked if the green LED does not turn on while charging?
A8-7
82. How would you take care of a household solar panel?
83. What is the difference between MFR and CFR resistors?
84. Why do you use an MFR resistance?
85. In which direction is the flow of electricity in a diode?
86. Name the parts of a diode.
87. What is the full form of LED?
88. What are the different types of modules?
89. What is the difference between a series connection and a parallel connection?
90. What is the effect on the voltage and current if you divide cells in a solar panel into quarters?
91. Where do the following wire connections go:
Primary a Primary bc Primary d Base e Base fg Base h Secondary x Secondary y
92. What are the tests that can be conducted by a multimeter?
93. What is the purpose of using a pot core in a transformer?
94. Connect the flowing four 6 V batteries in series:
95. Connect the following two 12 V batteries in parallel:
A8-8 96. Show a 12 V output connection using four 12 V panels:
97. Show a 24 V output connection using four 12 V panels:
98. Why do you install a solar panel facing north?
99. Name the different parts of a home lighting system.
100. Who is a barefoot solar engineer?
A8-9 Appendix 9: Estimate Of Cost of Equipment for a New Rural Electronic Workshop (REW) in Ladakh in 2000
No. Description of Items Quantity Rate (Rs.) Total (Rs.) 1 Chisel 10 50 500 2 DE Spanner Set 20 30 600 3 De-soldering Pump 10 50 500 4 Drill Bit 6/64, 7/64, 9/64 20 50 1,000 5 Electronic Charge Controller 1 2500 2,500 6 Fitting Materials 1 set 5000 5,000 7 Hammer ½ Kg 10 100 1,000 8 Hand Drill Machine 2 200 400 9 Hex Saw 5 100 500 10 Iron Element 12 50 600 11 Iron Bit 12 20 240 12 Knife 10 50 500 13 Mini File Set 2 50 100 14 Multi Wire Cutter MT 31 6 75 450 15 Multi-meter (Analogue) 2 500 1,000 16 Multi-meter (Digital) 2 2500 5,000 17 Nose Pliers 10 75 750 18 Nut Driver 5 mm 10 50 500 19 Nut Driver 7 mm 10 50 500 20 PCB Card Tray 6 700 4,200 21 PCB Assembly Jig 1 3000 3,000 22 Rakes for Component 2 500 1,000 23 Screw Driver 825 10 50 500 24 Screw Driver 832 10 25 250 25 Screw Driver 872 10 30 300 26 Screw Driver 933 10 25 250 27 Screw Driver Set 812 5 125 625 28 Sitting tools 10,000 29 Soldering Iron DC 24V 30 W 10 100 1,000 30 Screw Driver 913 10 25 250 31 Test Panel 4 1000 4,000 32 Tin Box for Tools 6 500 3,000 33 Tweezers 10 10 100 34 Winding Machine 1 3500 3,500 35 Wire Cutter 06 8 75 600 36 Wire Stripper 6 100 600 37 Establishment of Solar Power Unit 1,50,000 (For 300 watt regular power supply to REW) Total 2,04,815
A9-1 Appendix 10: Details of repairs carried out on PV Systems installed in Leh district (1993-94)
Village No. of trips to Leh for repairs Total cost of (solar engineer) and cost of each trip repairs (Rs.) Parts replaced Chalunkha (Abbas 4 @ Rs. 400 0.90 5 resistors (47k) @ 0.18 Ali): 147 inverters (10) @ 14.70 37 units installed in distilled water 200 ltr @ Nov. 1992 2400 12 Total 2548 Digger and Tangyar: 5 @ Rs. 300 50 bulbs replaced in Nov. 39 units installed in 5500 92 @ 110 Sept. 92 7500 1 solar panel 294 20 inverters repaired 10 charge controllers 6000 repaired @ 600 90 ltrs distilled water @ 2160 12´2 30 units inverter to be 441 repaired Total 21695 Tsaga (Sonam 4 @ Rs. 400 distilled water 70 ltrs @ Tashi): 840 12 37 units installed in 44.10 3 inverters @ 14.70 Nov. 92 1200 2 chargers @ 600 Total 2084 Anley (Tundup 6 @ Rs. 400 1 battery replaced after 6 Dorjey): 3500 months 50 units installed in 50 ltrs distilled water @ 600 Oct. 92 12 3000 5 chargers @ 600 112.90 7 inverters @ 14.70 Total 7212.90 Man Merak (Sonam 12 @ Rs. 500 367.5 transistors @ 14.70 Stoban): 2.70 15 resistors @ 0.18 67 units installed in Oct. 92 50 ltrs distilled water @ 600 12 2 batteries replaced @ 7000 3500 Total 7971.25
Source: (Roy 1996, p.50)
A10-1 Appendix 11: Details, Parts and Costs for Systems Produced in the Barefoot College REWs
Solar Lantern (2006) 10W, 12V module (Rajasthan Electronics & Instruments Ltd. Jaipur) 12V, 7Ah EpZ-12 (12M7) Exide Powersafe Battery (made in India) www.exideindustries.com Price: 4 200 Rs (inc. installation and 1 year warranty including travel and parts)
Part Cost (Rs) Source Module (10W) 2500 Rajasthan Electronics & Instruments Ltd. Jaipur Battery (7Ah) 500 Exide Powersafe Battery (made in India) Circuit 300 Body 300 Delhi CFL 60 Wiring 40 Tube Holder 20 Fuse & Holder 15 Transport, Labour & Operating 465 Costs Total 4200
The moulded plastic casing for the solar lantern can be purchased from 4-5 companies in Delhi (they are already available for purchase, not custom made).
Charge Controller (2006): Part Cost (Rs) Metal box 60 Plastic body 45 Total 450
Lamp (2006): Part Cost (Rs) Tube 25 Tube Holder 20 Fuse & Holder 15 Wiring 30 Switch 10 Lamp Body 60 Transport, Labour & Operating 290 Costs Total 450
Small Fixed System (2006) 18W, 12V module (Rajasthan Electronics & Instruments Ltd. Jaipur) 12V, 3A Charge Controller (assembled by SWRC) 9W, 12V Compact fluorescent lamp (ballast assembled by SWRC)
A11-1 12V, 20Ah Exide Powersafe Battery (made in India) www.exideindustries.com Price: 8 500 Rs (inc. installation and 1 year warranty including travel and parts)
No. S Home Light System wp Qty Rate 1 SPV Module 12v 18 Wp 1 2 Battery 12v 20 Ah 1 3 Charge Controller 12v 8 AMP 1 4 Lamp 12v 9 W 1 5 SPV Module Stand 1Set 6 Wire 1.5 mm 2 Core 12 Mtr TOTAL 8500.00
37 W Home Light System (2006)
No. S Home Light System wp Qty Rate 1 SPV Module 12v 37 Wp 1 2 Battery 12v 40 Ah 1 3 Charge Controller 12v 8 AMP 1 4 Lamp 12v 9 W 2 5 SPV Module Stand 1Set 6 Wire 1.5 mm 2 Core 16 Mtr TOTAL 14500.00
Part Cost (Rs) Module 12V 37 Wp 5600 Battery 12 V 75 Ah 4000 2 CFL lamps @ 350Rs each 700 Battery Box 500 Charge Controller 450 Wiring, switches etc 160 Module Stand 150 2 CFL tube 900 @ 65Rs each 130 Sales Tax 243 Transport, Insurance 400 Labour & Operating Costs 1000 Spares 167 Total 13500
70 W Home Light Systems
No. S Home Light System wp Qty Rate 1 SPV Module 12v 70 Wp 1 2 Battery 12v 75Ah 1 3 Charge Controller 12v 8 AMP 1 4 Lamp 12v 9 W 4 5 SPV Module Stand 1 Set 6 Wire 1.5 mm 2 Core 25 Mtr TOTAL 26000.00
A11-2 Part Cost (Rs) Module 18 500 4 lamps 1800 Charge Controller 450 Wiring, switches etc 200 Battery Transport, Labour & Operating 2000 Costs Total 27 500
Larger Fixed System 75W, 12V module (Rajasthan Electronics & Instruments Ltd. Jaipur) 12V, 10A Charge Controller (assembled by SWRC) 4 x 9W, 12V Compact fluorescent lamps (ballasts assembled by SWRC) 12V, 75Ah Exide Powersafe Battery (made in India) www.exideindustries.com Designed to run all four lamps plus a fan for 4 hours, or a B&W TV instead of 2 lamps. Price: 27 500 Rs (inc. installation and 1 year warranty including travel and parts)
Field Centre System: 2.5kW, 48V array 3kVA inverter
For privately purchased systems, after 1 year warranty period, a service charge applies, including travel and spare parts.
A11-3 Components and Cost of Materials for a 12V Inverter Circuit Name of Materials Value Quantity Price Total 1. R1 820 Oms 1 1.00 1 2. Q 1.Q 2 MJE 3055 2 12.50 25 3. C 1 330/40 Volt 1 5.00 5 4. C 3 3900/J/2k Volt 1 7.00 7 5. C 2 6800/J/2k Volt 1 7.00 7 6. D 1 1 N 4007 1 .50 .50 7. L 1 Chock 65 turn 1 10.00 10 8. Q.1 Q2. Heat sink MJE 2 5.00 10 9. Pot Core 26/16/HP 1 35.00 35 10. Bobbin 26/16/HP 1 2.00 2 11. Bobbin Rode Type 1 .50 .50 12. Core Rode Type 1 .50 .50 13. Con.1 Cpu 5 Pin 1 10.00 10 14. Con.2 Cpu 2 Pin 3 12.00 36 15. P.C.B. Inverter 1 25.00 25 16. Screw Brass 2.5 MM 4 2.00 8 17. Screw Steel 2.5 MM 1/4 2 .50 1 18. Scerw pot Core 1’’ 1 .50 .50 19. Nut pot Core 1 .50 .50 20. Washer pot Core 1 .50 .50 21. Washer 2.5 MM MJE 2 .50 1 22. Winding Wier 20 SWG 65 Turn 1 20.00 20 23. Winding Wire 28 SWG 9+9 Turn 1 - 24. Winding Wire 30 SWG 2+2 Turn 1 - 25. Winding Wire 32 SWG 180 Turn 1 - 26. Cello Tap .50 .50 27. Sleaves .50 .50 28. C.F.L. 9 W 1 65.00 65 29. C.F.L. Holdar 4 Pin 1 15.00 15 30. Lamp Housing 1 70.00 70 31. Fues Holdar 1 15.00 15 TOTAL 372
A11-4 Components and Cost of Materials for a 12V 4A Charge Controller Components Value Reference Quanty Price Amount 1 P.C.B. Asper design 1 20 20 2 I.C. LM324 U1 1 7 7 3 I.C. PC817 U2 1 8.5 8.5 4 Capacitor Elco 10/25 V CI.C7.C9 3 2.7 8.1 5 Capacitor Cer 102 C2.C4.C10 3 0.3 0.9 6 Capacitor Cer 104 C6.C8 2 0.3 0.6 7 Resistance MFR 1k ¼ W R9R10 2 0.3 0.6 8 Resistance MFR 2k2 ¼ W R2 1 0.3 0.3 9 Resistance MFR 12k ¼ W R1.R11.R16 3 0.3 0.9 10 Resistance MFR 4k7 ¼ W R3.R17. BisideConn 3 0.3 0.9 11 Resistance MFR 10k ¼ W R6.R12.R16 3 0.3 0.9 12 Resistance MFR 1k2 ¼ W R24 1 0.3 0.3 13 Resistance MFR 62k ¼ W R8.R13 2 0.3 0.6 14 Resistance MFR 39k ¼ W R4 1 0.3 0.3 15 Resistance MFR 820e ½ W R5 1 0.7 0.7 16 Trimpot/Priset 4k7 ¼ W vRivR2 2 10 20 17 Zener Diode 5.1 V ZD1.2 2 1 2 18 Diode 4148 D1.2 ZD3.4.5. 5 0.35 2.75 19 Diode IN4007 D3 1 0.35 0.35 20 Diode 5408 1.3 1.3 21 Trangislar BC547 Q1 1 1 1 22 Mosfet irf Z44 Q1.3. 2 17.5 35 23 Regulator IC 7806 6v 1 7 7 24 Schotky Diode 5140 Q4 1 30 30 25 Connector Rellimet 4Pin 2.54 1 1 2 2 26 Connector NG 5PinNG 7.54 1 1 5.25 5.25 27 LED 5mm Red & Green 2 1 2 28 LED Holder 5mm(p) 2 0.65 1.3 29 Nylon lamp 6/12v 1 2.35 2.35 30 Nylon Holder 3mm(p) 1 0.65 0.65 31 Glass Fuses 3Amp 1 1 1 32 Fuses Holder 3Amp 1 7.8 7.8 33 Toggle Switch 1Amp 1 5 5 34 FM Socket FM 1 5 5 35 M/FS Box Charger Base/Top 1 50 50 36 C/Box 1 18 18 37 Heat Sink 48/2 ph 48 3 2 6 38 Luges 7253 4 0.6 2.4 39 Luges 7251 4 0.45 1.8 40 Fiber Wasar m8 1 1 1 41 Faber Wasar m4 4 0.4 1.6 42 Spaccr Nylon m4 4 0.15 0.6 43 Screw m3x15 4 0.2 0.8 44 Screw 3x10 4 0.15 0.6 45 Screw m3x12 4 0.3 1.2 46 Nut m3 20 0.15 3 47 Beads Nylon m3 5 0.1 0.5 48 P.V.C Wire 30 CM 1 12 49 P.V.C Wire 32/.2 Black 50 CM 1.2 12.2 50 P.V.C Wire 32/.2 Yellow 12 CM 0.5 0.5 51 P.V.C Wire 32/.2 Blue 17 CM 0.5 0.5 52 P.V.C Wire 14/36 Black 32 CM 1.3 1.3 53 P.V.C Wire 14/36 Red 32 CM 0.3 0.7 54 P.V.C Wire 14/36 Black 25 CM 0.7 0.7 55 P.V.C.Sleeves 2mm 25 CM 25 CM 0.4 0.4 56 P.V.C.Sleeves 4mm 20 CM 20 CM 0.6 0.6 57 P.V.C.Sleeves 5mm 20 CM 20 CM 0.5 0.5 58 jumper Wire 6 0.2 1.20 TOTAL 286-45
A11-5 Components and Cost of Materials for a Solar Lantern No. Description of Items Quantity Rate (Rs.) Total (Rs.) 1 P.C.B Lantern 1 20 20 2 MFR 1/4w 1 % 10k 0 3 .30 .90 3 MFR 1/4w 1 % 5.6k 0 1 .30 .30 4 MFR 1/4w 1 % 27k 0 2 .30 .60 5 MFR 1/4w 1 % 1k 0 3 .30 .90 6 MFR 1/4w 1 % 4. 7k 0 3 .30 .90 7 MFR 1/4w 1 % 39k 0 2 .30 .60 8 MFR 1/4w 1 % 8.2k 0 1 .30 .30 9 MFR 1/4w 1 % 2.7k 0 1 .30 .30 10 MFR 1/4w 1 % 100E 0 1 .30 .30 11 MFR 1/4w 1 % 1M 0 1 .30 .30 12 MFR 1/4w 1 % 2.2k 0 4 .30 1.20 13 MFR 1/4w 1 % 6.8k 0 1 .30 .30 14 MFR 1/4w 1 % 75E 0 1 .40 .40 15 MFR 1/4w 1 % 7.5k 0 1 .30 .30 16 MFR 1/4w 1 % 270E 0 2 .40 .80 17 MFR 1/4w 1 % 150E 0 2 .40 .80 18 1N 4007 MIC 10 .35 19 IN 5408 1 1.30 1.30 12 ELCOSE 10/ 25 V OR 50V 2 2.60 5.20 21 ELCOSE 100/40V 1 2.60 2.60 22 PPC CAP 562 K 200V 2 23 PPC CAP 10K2E/200V 3 2 6 24 PPC CAP 010/400V 1 2 2 25 BOX TYPE 5600/2KVA 1 7.92 7.92 26 TRANSISTOR BC 547B 2 1.20 2.40 27 TRANSISTOR C100 1 2.20 2.20 28 TRANSISTOR TIP 431 1 1.20 1.20 29 ZENER DIODE 5.1V 2 2.20 4.40 30 TRANSISTOR MJE 2955 1 12 12 31 TRANSISTOR 3055 2 9 18 32 IC LM 324 1 7 7 33 WW RESIS…05E 5W 1 10 10 34 HEATSINK Chanale 2 - 35 CHOCK 5x20 1 10.50 10.50 36 BRASS SCREW NUTS 1 37 TRANSFORMER 26/16 1 46.50 46.50 38 Pot Core 1SET - 39 CONNECTER 5 PIN CPU 1 40 CONNECTER 6 PIN CPU 1 41 CONN. 4 PIN RELIMATE 1 42 M 3 X 8 SCREW NUT 1 - 43 SWTCH 1 6.5 6.50 44 GLASS FUSE 1 1.15 1.15 45 CFL TUBE 1 57 57 46 CEL TUBE HOLDER 1 15 15 47 CONNECTINO WIRS 15ET 50 50 48 SOLAR LANTERN BOOY HBAEERY CLAMP 1 220 220 49 JHON PLUG MLF 1Set 12 12 50 SMF BATTERY 12 V7AH 1 410 410 51 SPV MODUEE 12V 10WP 1 1800 1800
TOTAL 2740.07
A11-6