BENCHMARKS OF GLOBAL CLEAN ENERGY MANUFACTURING, 2014–2016

Operated by the Joint Institute for Strategic Energy Analysis AcknowledgmentsCrosscutting Findings Trade Trends

Acknowledgments

The principal investigators—Debra Sandor and David Keyser—would like to acknowledge the U.S. Department of Energy (DOE) Office of Energy Efficiency and Renewable Energy (EERE), led by the Strategic Analysis Team, for its support in development of this report, including Brian Walker, Ben King, and Paul Spitsen. We also acknowledge the members of the Clean Energy Manufacturing Analysis Center (CEMAC) Advisory Committee for their advice throughout the development process. Key expert input was provided by John Reilly (Massachusetts Institute of Technology), Rebecca Hill (Colorado State University), and Ben Mandler (American Geosciences Institute). Many analysts within CEMAC made major contributions, including Ashwin Ramdas, Samantha Reese, Ahmad Mayyas, Scott Caron, Eric Lantz, Jon Weers, Robert Spencer, James McCall, Tian Tian, Tsisilile Igogo, and Billy Roberts. We also appreciate the editorial support of Beth Clark and Mike Meshek, and the design services of Liz Craig. Operational support by the Joint Institute for Strategic Energy Analysis (JISEA)’s Jill Engel-Cox and the National Renewable Energy Laboratory (NREL)’s Doug Arent and Gian Porro have been instrumental to CEMAC in producing this benchmark report.

ii Table of Contents

Contents

About this Data Book ...... 1

Crosscutting Findings ...... 5

Notable Trends ...... 5

Benchmark Data: Market Trends ...... 6

Benchmark Data: Trade Trends ...... 9

Benchmark Data: Value Added Trends ...... 10

Challenges: Comparing Clean Energy Manufacturing Trends Over Time ...... 14

Wind Turbines ...... 15

Wind turbine components and supply chain ...... 15

Notable Trends ...... 16

Benchmark Data: Market Trends ...... 17

Benchmark Data: Trade Trends ...... 22

Benchmark Data: Value Added Trends ...... 24

Wind Challenges and Opportunities: Cost Competitiveness and Advanced Manufacturing . . . . .27

Crystalline Silicon Photovoltaic Modules ...... 29

Crystalline silicon photovoltaic (PV) modules and supply chain ...... 29

Notable Trends ...... 30

Benchmark Data: Market Trends ...... 31

Benchmark Data: Trade Trends ...... 36

Benchmark Data: Value Added Trends ...... 40

PV Challenges and Opportunities: Balance of System Components ...... 43

iii Table of Contents

LED Packages ...... 44

Light-emitting diode (LED) packages and supply chain ...... 44

Notable Trends ...... 45

Benchmark Data: Market Trends ...... 46

Benchmark Data: Trade Trends ...... 49

Benchmark Data: Value Added Trends ...... 51

LED Challenges and Opportunities: Quality and Cost ...... 54

Lithium-Ion Battery Cells for Light-Duty Electric Vehicles ...... 55

Light-duty vehicle lithium-ion battery (LIB) cells and supply chain ...... 55

Notable Trends ...... 56

Benchmark Data: Market Trends ...... 57

Benchmark Data: Trade Trends ...... 62

Benchmark Data: Value Added Trends ...... 64

Raw Materials in the Manufacturing Supply Chain ...... 67

LIB Challenges and Opportunities: Recycling Critical Materials ...... 73

List of Acronyms and Abbreviations ...... 77

References ...... 78

Appendix ...... 84

iv Table of Contents

List of Figures and Tables

Benchmarks of Global Clean Energy Manufacturing report framework ...... 2

Clean energy technology end product global demand and production shares by economy, 2014–2016 ...... 6

Clean energy technology end product demand and production trends by economy, 2014–2016 ...... 7

Clean energy technology end product manufacturing capacity utilization, 2016 ...... 8

Clean energy technology end product trade, 2014–2016 ...... 9

Clean energy manufacturing total value added (tVA) by clean energy technology, 2014–2016 ...... 10

Clean energy manufacturing total value added (tVA) by value added component, 2014–2016 ...... 11

Clean energy manufacturing total value added (tVA) domestic and non-domestic contribution, 2014–2016 ...... 12

National gross domestic product and clean energy manufacturing contribution to economy-wide manufacturing, 2014–2016 13

Impact of global average selling price (ASP) on global demand trends, 2014–2016 ...... 14

Wind turbine supply chain demand and production shares by economy, 2014–2016 ...... 17

Wind turbine supply chain production and demand trends, 2014–2016 ...... 18

Wind turbine supply chain manufacturing capacity utilization, 2014–2016 ...... 20

Wind turbine generator sets (nacelles and blades) trade, 2014–2016 ...... 22

Wind turbine generator set (nacelles and blades) trade flows, 2014–2016 ...... 23

Wind turbine supply chain total value added (tVA) by value added component, 2014–2016 ...... 24

Wind turbine supply chain total value added (tVA) domestic and non-domestic contribution, 2014–2016 ...... 26

PV module supply chain demand and production shares by economy, 2014–2016 ...... 31

Global PV module supply chain demand and production for key PV module economies, 2014–2016 ...... 32

PV module supply chain manufacturing capacity utilization, 2014–2016 ...... 34

PV module supply chain trade, 2014–2016 ...... 36

PV module supply chain trade flows, 2014–2016 ...... 38

PV module supply chain total value added (tVA) by value added component, 2014–2016 ...... 40

PV module supply chain total value added (tVA), domestic and non-domestic contribution, 2014–2016 ...... 42

Modeled cost breakdown trends for PV systems installed in the United States (inflation adjusted), 2010–2017 ...... 43

LED package supply chain demand and production shares, 2014–2016 ...... 46

LED package supply chain demand and production for key LED package economies, 2014–2016 ...... 47

LED package supply chain manufacturing capacity utilization, 2014–2016 ...... 48

LED package trade, 2014–2016 ...... 49

LED package trade flows, 2014–2016 ...... 50

LED package supply chain total value added (tVA) by value added component, 2014–2016 ...... 51

LED package supply chain total value added (tVA), domestic and non-domestic contribution, 2014–2016 ...... 53

v Table of Contents

LED Applications, 2014–2016 ...... 54

Automotive LIB cell supply chain demand and production shares, 2014–2016 ...... 57

Automotive LIB cell supply chain demand and production for key LIB cell economies, 2014–2016 ...... 58

Automotive LIB cell supply chain manufacturing capacity utilization, 2014–2016 ...... 60

LIB cell (for all applications) trade, 2014–2016 ...... 62

LIB cell (for all applications) trade flow, 2014-2016 ...... 63

Automotive LIB cell supply chain total value added (tVA) by value added component, 2014–2016 ...... 64

Automotive LIB cell supply chain total value added (tVA) domestic and non-domestic contribution, 2014–2016 ...... 66

Cobalt supply chain supporting lib manufacturing ...... 67

Cobalt reserves and mines production, 2014–2016 ...... 68

Cobalt demand to support EV LIB cathode production, 2014–2016 ...... 69

Cobalt material exports, 2014–2016 ...... 70

Cobalt mining total value added (tVA), 2014–2016 ...... 71

Cobalt refining total value added (tVA), 2014–2016 ...... 72

Global li-ion battery recycling capacity, 2016 ...... 73

vi About this Data Book

About this Data Book The Benchmarks of Global Clean Energy Manufacturing methods for assessing and comparing clean energy report provides an assessment of the global state of clean technology supply chains.6 The analysis presented in this energy manufacturing between 2014 and 2016.1 Researchers benchmark report focuses exclusively on the manufacturing examined four technologies—wind turbine components aspects of the larger clean energy value chain and (blade, tower, nacelle), crystalline silicon (c-Si) solar examines each technology in terms of four manufacturing photovoltaic (PV) modules, light-duty vehicle (LDV) lithium- supply chain links: raw material, processed material, ion battery (LIB) cells, and light-emitting diode (LED) subcomponents, and end products. packages for lighting and other consumer products—across Just one piece of the larger clean energy economy, manufacturing supply chains that include processing raw manufacturing is the linchpin between technology materials, making required subcomponents, and assembling development and its deployment in the marketplace final products.2 (see Benchmarks of Global Clean Energy Manufacturing The impacts of the manufacturing supply chain for these report framework on p. 2). Upstream, innovation in the four technologies are assessed in terms of three common development stage has economic value in the form of benchmarks: market size (including manufacturing capacity intellectual property, research, and corporate management. and production), global trade flows, and manufacturing Downstream, installation, systems integration, and value added, and across 13 economies that comprise the operations bring economic value through employment, primary manufacturing hubs for the technologies: Brazil, services, property taxes, improved efficiency, decreased Canada, China, Denmark, Germany, India, Japan, Malaysia, energy consumption, and reduced negative environmental Mexico, South Korea, Republic of China (referred to impacts. While development and deployment of throughout this report as Taiwan), the United Kingdom, and technologies7 make tremendous contributions to the the United States.3 New methodologies were developed economy, this report focuses on the value added by and to generate the data sets for each benchmark, while opportunities found in the manufacturing supply chain. accommodating the variations in clean energy technology While there is a wide array of clean energy in the global manufacturing4 supply chains and data availability. marketplace today, this report uses four technologies Throughout this report, general drivers for benchmark as proxies for broader market trends. Wind turbine trends in the context of an ever-changing clean energy components, c-Si PV modules, lithium-ion battery cells, manufacturing landscape have been identified, but specific and LED packages were selected for this report because analysis of trends over the study period were not included they all experienced significant cost reductions, demand in the scope of effort. Nonetheless, the data and insights growth, and had adequate data to analyze during the report provided by these benchmarks can help guide research period. The specific materials and subcomponents in the agendas, inform trade decisions, and identify manufacturing analysis were selected based on standard criteria including opportunities by location and technology. data availability; uniqueness, or role as an enabling process/ product; involvement in global trade; impact on overall cost, Focus and Framework and contribution to quality (see the methodologies report8 5 The Clean Energy Manufacturing Analysis Center (CEMAC) for details on selection criteria). developed and uses a common framework and standardized

1 For more information, previous versions of the benchmark report, related reports, and key figures, see the CEMAC “Benchmarks of Global Clean Energy Manufacturing” website at https://www.jisea.org/benchmark.html. 2 Throughout this report, clean energy manufacturing refers to aggregated metric values for the four end products, unless stated otherwise. 3 Where data are available, a rest of world designation is used to present data from other economies beyond the thirteen. Throughout this report, global refers to aggregated metric values across the 13 economies examined and the rest of the world, unless stated otherwise. 4 Throughout this report, clean energy manufacturing technologies refers to the four end products examined in this report (wind turbine components, crystalline silicon solar photovoltaic modules, light-duty vehicle lithium-ion battery cells, and light-emitting diode packages for lighting and other consumer products), unless stated otherwise. 5 CEMAC is a program under the Joint Institute for Strategic Energy Analysis (JISEA). More information about CEMAC can be found at https://www.jisea.org/ manufacturing.html. 6 For details about the benchmark methodology, see CEMAC’s Benchmarks of Global Clean Energy Manufacturing, 2014-2016: Framework and Methodologies (Sandor et al. 2021). 7 Data related to the deployment of some of the technology end products examined in this report can be found in the International Renewable Energy Agency’s Statistics Time Series (IRENA n.d.) 8 Benchmark methodologies are detailed in CEMAC’s Benchmarks of Global Clean Energy Manufacturing, 2014-2016: Framework and Methodologies (Sandor et al. 2021).

1 About this Data Book

Benchmarks of Global Clean Energy Manufacturing report framework Value chain for clean energy technologies Installation/ System Operation & Development Manufacturing Construction Integration Maintenance

Manufacturing Supply Chain Links 4 Technologies Subcompone essed Materials nts Proc Clean Ener gy Techn erials ology w Mat End Pr Ra oduct Polysilicon, Silver Paste, Glass, c-Si PV Wafer, c-Si PV Cell, Silica, Silver Ore Specialty Polymers Frame, Encapsulant c-Si Solar PV Module

Iron, Neodymium, Steel, Fiberglass, Carbon Fiber, Permanent Magnets, Generators, Wind Turbine Components: or Dysprosium Ores Neodymium and Dysprosium Alloys Gear Assemblies, Steel Components Blades, Tower, Nacelle

Lithium, Cobalt, Cathode Materials, Separators, Housings, Light Duty Vehicle Nickel, Graphite Ores Anode Materials, Electrolytes Metal Foils, Tabs Li-ion Battery Cell

Gallium, Indium, Sapphire Substrates, Trimethyl LED Chips LED Package Yttrium Ores Gallium (TMG), Trimethylindium (TMI), YAG Phosphors

Economies Brazil • Canada • China • Denmark • Germany • India • Japan • Malaysia • 13 Mexico • South Korea • Taiwan • United Kingdom • United States

3 Benchmarks Market Trends Trade Trends $ Value Added Trends

3 Years 2014 • 2015 • 2016

The 13 economies were benchmarked based on market • New visualizations to present the key benchmark trends, size, manufacturing capacity across the supply chain, and with select visualizations published in the report, and data availability. Three common points of reference, or additional visualizations available online benchmarks—market size (including manufacturing capacity • The addition of Denmark to the group of manufacturing and production), global trade flows and manufacturing value hubs examined, in recognition of its contribution to wind added—provide a standardized basis for: turbine component manufacturing and trade • Comparing key economic aspects of clean energy • A new methodology to calculate the indirect value technology manufacturing on national and global levels added metric that now includes value streams from • Tracking changes as markets and manufacturing process both non-direct domestic and international intermediary evolve. components, processes, and services. A more detailed description can be found in the next section. What’s New • Throughout the report all costs have been normalized to This is CEMAC’s second benchmark report, expanded to 2014 dollars [US$(2014)] to allow comparison over the summarize trends between 2014 and 2016. To address period. stakeholder feedback and incorporate additional years of • Having secure access to raw materials is increasingly data, this benchmark report presents some new features important to manufacturers who use global supply chains. and formats: For the first time, the benchmark report is able to track • A more concise summary of the key insights from the raw materials for light-duty vehicle lithium-ion battery benchmark analysis with the format modified from a cells. However, while important, in this report raw material full-length technical publication to a comprehensive yet flows were not tracked for the other three benchmark easy-to-digest report technologies because of insufficient data.

2 About this Data Book

Understanding Benchmark Reporting production trends by economy, 2014–2016 on p. 7 for example.) Market Benchmarks 3. Stacked bar charts show excess manufacturing capacity, Clean Energy Market Size: This benchmark provides production, and capacity utilization for each link in the insight into the relative concentration of demand for supply chain. (See Clean energy technology end clean energy technologies across the globe. Market size product manufacturing capacity utilization, 2016 on p. 8 (or market demand) data were collected from existing for example.) secondary sources to estimate the market size for each aterials Subcomponents Processed M Clea technology across the manufacturing supply chain and in n Energy T Trade Benchmark ls echnolo ateria gy End each economy. When they were available actual production Raw M Prod This benchmark provides insight into global clean uct data for each subsequent downstream intermediate9 energy trade activity and interconnectedness across the formed the basis of demand estimates for key supply chain manufacturing supply chain. Balance of trade (exports intermediates. When data were not available, typically for minus imports) is a key component of national GDP. The smaller industries (LED packages and LDV Li-ion battery value of trade flows is derived from imports and exports cells), the demand for intermediates was approximated data tracked by international harmonized trade codes used by assuming the production volume of the end product is by the U.S. International Trade Commission (U.S. ITC) and equivalent to the demand for each upstream intermediate International Trade Centre (ITC)10. While official trade data product. The monetary value of demand was determined for the final products are often available, the upstream data by applying estimates of average global unit prices to allow are often intertwined with much larger industry sectors and comparison across technologies and economies. difficult to extract for the specific technology of interest. Clean Energy Manufacturing Capacity and Production: With the exception of PV cells and polysilicon (two of the This benchmark provides insight into the clean energy intermediates for PV modules), global trade data are only manufacturing capacity and production around the world available for the end products included in this report. and highlights opportunities for expansion to meet demand. Trade data are displayed in two visualizations: Manufacturing capacity and production were estimated to highlight the economies that make the largest contributions 1. Bar charts—which show imports (negative values), in each category and to understand where excess capacity exports (positive values), and balance of trade (BOT) is located around the world for each technology. Like market numerically—allow readers to quickly identify import size data, data were collected from existing secondary and export trends among benchmarked economies. sources, and monetary values were determined by (See National gross domestic product and clean applying estimates of average global unit prices to (1) allow energy manufacturing contribution to economy-wide comparison across technologies and economies and (2) manufacturing, 2014–2016 on p. 13 for example.) provide input for the value added benchmark based on the 2. Interactive chord charts that highlight the trade flows production value of each technology and intermediate. among benchmarked economies (for one year and, one Market data are presented in three visualizations in this report: technology link in each view) are available online.11 1. Stacked bar charts display the global distribution Value Added Benchmarks of production and demand across the supply chain, This benchmark provides insight into the contribution highlighting the contribution (in %) of each economy. and importance of clean energy manufacturing to (See Clean energy technology end product global national economies. Value added (VA) from clean energy demand and production shares by economy, 2014–2016 manufacturing contributes to an economy’s GDP and consists on p. 6 for example.) of wages, returns to capital (e.g., income to property owners), 2. Combination line and bar charts show magnitude and and taxes. Manufacturing VA from clean energy technologies trends of production and demand in end product units is estimated using the estimated production value for each (e.g., megawatts for PV modules) across the supply chain. intermediate across the supply chain in combination with (See Clean energy technology end product demand and social accounting data from the Organization for Economic

9 Throughout this report, intermediates refer to the specific materials and components included in each supply chain link of the four end products. 10 For more information about USITC, see the USITC website at https://www.usitc.gov/ and the International Trade Centre’s market analysis tools at https://www.intracen.org/itc/market-info-tools/market-analysis-tools/. 11 “Clean Energy Trade Benchmark,” JISEA, https://www.jisea.org/benchmark.

3 About this Data Book

Cooperation and Development (OECD) Structural Analysis inputs such as polysilicon in solar module production are (STAN) Input-Output (I-O) database.12 sourced domestically, both the dVA and iVA accrues to domestic industries or businesses that supply those inputs. Total value added (tVA) from clean energy manufacturing If inputs are imported, the iVA accrues to businesses in the is generally highest in economies with the highest levels of economy of origin and is not included in the dVA calculation. production and is composed of two components: VA retained is calculated by dividing the domestic total value • Direct value added (dVA) comes solely from domestic added (tVA) by the revenues from domestic manufacturing clean energy manufacturing. This contribution to (aka direct output). For example, if solar module national GDP includes payments to manufacturing manufacturing generated $100 million in revenue in a specific workers, property-type income such as profits earned economy, and domestic dVA is $30 million and domestic iVA by owners and investors, and taxes paid on production is $20 million, VA retained is 50%: (30 +20)/100. less government subsidies within a single economy. Value added data are presented in two different For example, if solar module manufacturing generated visualizations: $100 million in revenue in a specific economy, and 70% of that went to intermediate inputs (payments for both 1. Bar charts represent dVA, iVA, and tVA for benchmarked domestic and non-domestic goods and services used in economies by technology supply chain links for each of production), the remaining 30% would be the direct value the three years. (See Clean energy manufacturing total added. value added (tVA) by value added component, 2014–2016 on p. 11 for example.) • Indirect value added (iVA) has two subcomponents: 2. Bar charts show the share of tVA accrued from 1. Domestic iVA comes from the broader supply chain that domestic and non-domestic production of clean energy provides domestic inputs13 used by manufacturers. technologies for each benchmarked economy. (See Clean 2. Non-domestic iVA comes from goods and services energy manufacturing total value added (tVA) domestic exported to support manufacturing that takes place in and non-domestic contribution, 2014–2016 on p. 12 for other economies. The GDP of the economy that exports example.) these goods and services benefits from the wages, profits, and taxes that support manufacturing in that Data Confidence exporting economy. This report provides a unique perspective of the clean For example, a module manufacturer may purchase energy manufacturing value proposition. The data polysilicon from a polysilicon producer. This producer and needed to estimate the benchmarks at the desired level its contribution to GDP would be included in the indirect of disaggregation are not available for all technologies effect, either as domestic iVA, if the polysilicon was included in the benchmark report. By applying technology- manufactured domestically and as non-domestic iVA if specific engineering assumptions and analysis best the polysilicon was manufactured in another country. The practices, along with consultation and review by experts non-domestic (inter-country) iVA indicates the globalization from industry and academia, we estimated benchmark and interconnectedness of benchmarked economies with metrics across the manufacturing supply chain. However, respect to clean energy manufacturing supply chains, and our level of confidence in data reported here varies. Details the domestic iVA indicates the strength of domestic supply of the data confidence and specific assumptions used chains. for each technology are provided in CEMAC’s benchmark methodology report.14 Value added retained (VA retained) estimates the fraction of revenue an economy retains from in-economy production of clean energy technologies. VA retained varies across economies as a result of different wage rates, tax rates, government subsidies to industries, and company profitability. It can also be influenced by how much is spent on inputs, either imported or sourced domestically. When

12 Further information about the OECD STAN I-O database, including the data used in the benchmark report, can be found at http://www.oecd.org/sti/ind/ stanstructuralanalysisdatabase.htm. 13 Domestic inputs are payments by a domestic business or industry to other domestic businesses and industries for goods or services used in production. 14 Benchmark methodologies are detailed in CEMAC’s Benchmarks of Global Clean Energy Manufacturing, 2014-2016: Framework and Methodologies (Sandor et al. 2021).

4 Crosscutting Findings

Crosscutting Findings Looking across the manufacturing supply chains of the four period, while actual unit sales (physical units) increased due technologies—wind turbine components, crystalline silicon to rapidly dropping prices for end products. For example, (c-Si) photovoltaic (PV) modules, light-emitting diode (LED) while production of the four clean energy technologies packages, and light-duty electric vehicle (EV) lithium-ion increased significantly from 2014 to 2016 in physical units, battery (LIB) cells—provides perspective on the collective/ the associated tVA decreased as a result of rapidly declining aggregate impacts and trends of clean energy manufacturing prices over the period. This situation is also reflected in between 2014 and 2016. the decline in global end product imports and exports in aggregate dollar terms over the period. Notable Trends Wind Manufacturing Capacity and Utilization At $50.6 billion in 2016, tVA from wind component production Manufacturing capacity expansion to meet anticipated across the 13 economies analyzed was the highest among demand growth was driven in part by domestic policies that the four benchmarked clean energy technologies. In addition, set targets for renewable energy production and provide because tVA generally follows production trends, of the incentives to offset costs. Overall capacity utilization relative clean energy technologies studied, manufacturing of wind to global production for clean energy benchmark technology turbine components contributed the most value added to end products except LED chips declined from 2014 to 2016. the benchmarked economies. A reduction in wind turbine While all economies added manufacturing capacity, China production over the period drove the small overall decline in added the largest amount of new clean energy manufacturing tVA from the four clean energy technologies (in aggregate) capacity during the analysis period. Low manufacturing across the benchmarked economies. capacity utilization rates may imply that these industries could boost production in current manufacturing facilities China to meet future demand growth or that new investment China accounted for the largest demand for and production is required to modernize manufacturing processes to of each of the four clean energy technology end products, accommodate new technologies. Production increases that with tVA three to four times higher than that for each of are not accompanied by increased demand, however, can the next three economies (the United States, Japan, and place downward pressure on prices. For example, oversupply Germany). China also contributed the highest levels of non- in PV module and LED chip supply chains contributed to domestic iVA to other economies. Policies focused on building falling prices for these components over the period. manufacturing capacity and domestic supply chains and concerted efforts to increase production beyond domestic Global Supply Chains demand contributed to increased exports over the period and Across the benchmarked economies, indirect value added also helped China secure its position as the only benchmarked (iVA) from clean energy manufacturing was greater than economy able to meet domestic demand for the four end 15 direct value added (dVA), indicating that clean energy products with domestic production alone. manufacturing supply chains added more value, both domestically and globally, than the manufacture of the end United States products. All benchmarked economies received iVA from The United States moved from third- to second-highest tVA the production of the four technologies in other economies from clean energy manufacturing of the four technologies as a result of global supply chains that link these economies. over the period. The United States generally retained the In general, most individual economies did not have the highest shares of revenue from in-economy production of manufacturing capacity to meet their own demand for the clean energy technologies as dVA (VA retained), as a intermediates and services across the entire supply chain result of a combination of factors that varied across the four and relied on trade networks to fill the gaps. technologies relative to the other benchmarked economies, including robust domestic supply higher wages, greater Price and Volume profits to domestic shareholders, and fewer subsidies. For some clean energy technologies in some economies, total More detailed information can be found on the following value added (tVA) and market demand decreased over the pages.

15 See, for example, Usha C.V. Haley, George T. Haley, “How Chinese Subsidies Changed the World”, Harvard Business Review, April 25, 2013, https://hbr.org/2013/04/how- chinese-subsidies-changed. 5 Crosscutting Findings Market Trends

CH1-Benchmark Crosscutting Data: Story Market Trends

CC-1. 2014-2016 CC-2. 2014-2016 CC-3. 2014-2016 CEM CC-4. 2014 -2016 Trade CC-9. Demand/Price FYI Calc for CC2 Clean energyProduction an dtechnologyProduction end- Deman dproductExcess Ca pglobalacity demandFlows by Tech Endand productionTrends sharesdiscussion by economy, Demand Shares Trends Product 2014–2016

PV Module Wind LED Chip LIB Cell Country 100% China 7% 11% United States 20% 26% 30% Brazil 32% 36% 80% 42% 43% 39% 45% 48% Canada 29% Denmark 14% 26% 59% 60% 14% Germany 5% 22% 22% India 9% 22% 21% 15% 15% 14% 14% Japan Demand Shares 5% 40% 22% 14% Malaysia 20% 4% 5% 10% 10% 12% 7% 9% 6% Mexico 9% 27% 11% 4% 7% 24% Rest of world 20% 4% 21% 22% 16% 13% 22% 22% South Korea 12% 17% 17% 13% 5% 13% 13% 11% Taiwan 6% 7% 0% 6% 6% 5% United Kingdom 100% 13%

80% 38% 40% 38% 46% 44% 44% 19% 50% 53%

74% 71% 70% 60% 5% 6% 14% 16% 9% 13% 15% 5% 40% 5% 12% 15% Production Shares 40% 11% 12% 14% 10% 5% 7% 6% 17% 9% 14% 10% 14% 14% 22% 17% 20% 4% 5% 9% 9% 23% 6% 8% 23% 14% 19% 15% 5% 8% 7% 10% 9% 0% 5% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Breakdown (in %) of global demand (top) and production (bottom) by economy for benchmarked clean energy technology end products. Note that LED chip data are presented in place of LED package data. (Due to a lack of availability of economy-specific demand data for LED packages, the benchmark analysis assumes that demand for LED packages is equal to production throughout the report).

Of the benchmarked economies in 2016, China had the largest Demand for LED packages (used in manufacturing a wide shares of demand for and production of the four clean energy variety of products from lighting to televisions) was particularly technologies. China increased its share of global demand for concentrated, with nearly 100% of aggregate demand coming PV modules, LED chips, and LIB cells, while its wind turbine from only five economies—Japan, South Korea, Malaysia, component share held steady. Taiwan, and China—where many of the final consumer products that contain LEDs are assembled. In 2016, 58% of PV demand was found outside China, while 70% of production occurred in China. C-Si PV module production Lithium-ion battery cell demand was also fairly concentrated, outside China was dispersed across all but three of the with about 75% of aggregate demand located in four economies included here, with Malaysia and Japan being the economies—China, the United States, Japan, and Germany— next largest producers. the top four automotive manufacturers globally. Over the period, the distribution of demand and production shares Wind turbine component demand and manufacturing were shifted significantly for LDV LIB cells. The demand shares from generally colocated on a regional basis due to transportation non-benchmarked countries increased from 5% to 22%, as U.S. challenges associated with their size and weight. Outside China, demand shares declined from 59% to 22% over the period. On wind turbine production was led by the United States, Germany, the production side, LIB production shares dropped to 17% in India, Denmark, and Brazil. Japan (from 40%) and 9% in Korea (from 23%).

6 Market Trends Crosscutting Findings

Clean energy technology end product demand and production trends by economy, 2014–2016 y a orea l a n an a d mark ay si a wan h K a i Indi a Ch in a Braz i Ja p United Me xi co T Ca n Kingdom Ma l De n Ger m Sou t United States 8.0K

e 20.0K 6.0K u l 4.0K Mo d V $ million 10.0K P 2.0K

0.0K 0.0K

30.0K 8.0K

6.0K 20.0K

W in d 4.0K $ million 10.0K 2.0K

0.0K 0.0K

4.0K 2.0K 1.5K 3.0K 1.0K 2.0K $ million LED Chip 0.5K 1.0K 0.0K 0.0K 2.0K 2.0K 1.5K

1.0K LIB Cell

$ million 1.0K 0.5K

0.0K 0.0K 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Demand (color-coded lines) and production (gray bars), both in US$(2014), for four clean energy technology end products by economy. Note the variable scale, which is used to help visualize data trends across the widely varying market size for the four technologies; China data are on different scales than those of the other benchmarked countries. Note that LED chip data are presented in place of LED package data, due to a lack of availability of economy-specific demand data for LED packages.

The manufacturing of the four clean energy technology end Of the benchmarked countries, only China had sufficient products contributed to global markets of widely varying sizes production to meet domestic demand for the four clean in 2016, ranging from the $42 billion wind industry to the $6 technology end products over the period. In 2016, Germany and billion automotive lithium ion battery cell industry. India had sufficient production to meet domestic demand for PV modules and wind turbine components. Germany was also Between 2014 and 2016, China had the highest demand for and able meet its demand for LIB cells. production of the four benchmarked clean energy technologies. Over the period, global demand (on a dollar basis) for the four The smallest shortfalls between domestic production and clean technology end products in total decreased slightly—from demand appeared for wind turbine components, as these large $98.2 billion to $97.1 billion, with a peak of $116.8 billion in 2015. components tend to be manufactured relatively close to where Total demand (on a dollar basis) grew only in the United States demand is located. The largest production-demand gaps were and India. Wind and c-Si PV end products constituted the observed for PV modules in the United States, Japan, and India; largest contribution to demand for clean technologies across LED chips in Japan, Malaysia, and South Korea; and LIB cells in the 13 economies. the United States.

7 Crosscutting Findings Market Trends

CH1- Crosscutting Story

CC-1. 2014-2016 CC-2. 2014-2016 CC-3. 2014-2016 CEM CC-4. 2014 -2016 Trade CC-9. Demand/Price FYI Calc for CC2 Production and Production - Demand Excess Capacity Flows by Tech End Trends discussion Clean energyDemand Shares technologyTrends end product manufacturingProduct capacity utilization, 2016

Wind PV Module LED Chip LIB Cell Capacity Utilization Capacity Utilization Capacity Utilization Capacity Utilization 0% 50% 100% 0% 50% 100% 0% 50% 100% 0% 50% 100%

China 43% 53% 70% 27%

0K 20K 40K 0K 20K 40K 0K 20K 40K 0K 20K 40K $ million $ million $ million $ million

Brazil 54% 0%

Canada 66% 75%

Denmark 74%

Germany 68% 51% 97%

India 45% 36%

Japan 42% 91% 82% 31%

Malaysia 83%

Mexico 44% 87%

South Korea 37% 74% 70% 18%

Taiwan 21% 75%

United Kingdom 61% 24%

United States 77% 58% 86% 20%

0K 10K 20K 30K 0K 10K 20K 30K 0K 10K 20K 30K 0K 10K 20K 30K $ million $ million $ million $ million

72% 74% 66% 59% 60% 58% 64%

a cit y 54% 49% 55% 41%

Ca p 40%

a l 28% Utilization o b 20%

G l 20% 0% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Bars show manufacturing capacity (lighter shading) and utilized manufacturing capacity (i.e., production, darker shading) in US$2014 for the benchmarked economies in 2016. Vertical lines and associated numerical values show capacity utilization (production as a % of manufacturing capacity). Trend lines show global capacity utilization percentage for 2014–2016 (bottom). Note that China is displayed on a different scale.

Global manufacturing capacity increased for all four Increasing manufacturing capacity over the period indicates technologies over the period. There was excess manufacturing that manufacturers were anticipating continuing increased capacity for nearly all of the four clean energy technologies, in demand for the clean energy end products. Low capacity virtually all of the 13 economies. Capacity utilization was highest utilization implies that these industries boosted production for LED packages and lowest for LIB cells but was still lower in current manufacturing facilities to meet potential demand than the typical economy-wide manufacturing facility capacity growth from increased technology adoption. However, without utilization rate of around 80% (Federal Reserve Bank of St. increased demand, persistent low capacity utilization rates can Louis 2018). place downward pressure on pricing.

8 Trade Trends Crosscutting Findings

CH1- Crosscutting Story Benchmark Data: Trade Trends CC-1. 2014-2016 CC-2. 2014-2016 CC-3. 2014-2016 CEM CC-4. 2014 -2016 Trade CC-9. Demand/Price FYI Calc for CC2 Production and Production - Demand Excess Capacity Flows by Tech End Trends discussion Clean energyDemand Shares technologyTrends end product trade, 2014–2016Product

PV Module Generator sets (Nacelle + Blades) LED Packages and Chips LIB Cell Brazil -$15M -$254M -$35M -$306M Canada -$49M -$614M -$50M -$53M China $7,472M $294M -$2,987M $2,051M Denmark $4M $3,743M -$25M -$11M Germany -$286M $1,660M $340M -$488M India $22M $83M -$76M -$126M Japan -$6,070M -$84M $3,110M $1,496M 2014 Malaysia $1,622M $0M $1,615M $58M Mexico $513M -$527M -$366M $0M South Korea $1,599M -$45M -$339M $1,806M Taiwan $1,006M -$5M $1,917M $18M United Kingdom -$1,150M -$590M -$144M $82M United States -$3,822M -$407M -$397M -$976M Brazil -$42M -$98M -$29M -$198M Canada $15M -$614M -$54M -$72M China $10,316M $573M -$4,386M $3,170M Denmark -$112M $3,268M -$23M -$13M Germany $18M $1,926M $205M -$880M India -$186M $2M -$131M -$200M Japan -$4,194M -$85M $2,520M $1,384M 2015 Malaysia -$65M $0M $1,749M -$5M Mexico $912M -$389M -$468M -$226M South Korea $550M -$12M $334M $1,610M Taiwan $537M -$15M $3,153M $46M United Kingdom -$1,103M -$328M -$146M $154M United States -$5,712M -$84M -$486M -$921M Brazil -$250M -$32M -$27M -$168M Canada $36M -$93M -$52M -$93M China $7,381M $1,144M -$2,298M $3,630M Denmark -$43M $3,035M -$11M -$22M Germany -$533M $1,925M $160M -$1,067M India -$322M $1M -$178M -$296M Japan -$3,034M -$72M $2,299M $1,822M 2016 Malaysia $469M $0M $1,753M $70M Mexico $819M -$540M -$451M -$165M South Korea $1,335M -$51M -$46M $1,888M Taiwan $298M -$34M $1,567M $45M United Kingdom $460M -$325M -$132M $240M United States -$7,855M -$106M -$144M -$865M

Imports Exports Imports Exports Imports Exports Imports Exports

South United United Brazil Canada China Denmark Germany India Japan Malaysia Mexico Korea Taiwan Kingdom States $10.0B

$0.0B

B ala n ce of T rad e -$10.0B 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Bar chart (top) shows imports (negative values), exports (positive values), and balance of trade (exports less imports) in US dollars US$(2014) by economy for four clean energy end products: wind turbine nacelles and blades PV modules, LED chips and packages, and lithium-ion cells. Line chart (bottom) shows balance of trade trends for the four end products. Note that unlike other figures, imports and exports for PV modules are not broken out by chemistry (e.g., c-Si) and lithium-ion batteries are not broken out by end-use (e.g., light duty vehicles).

From 2014 to 2016 aggregate exports for the 13 economies 2016, wind turbine component imports experienced the largest declined 7.3% from $ 39.5 billion to $36.6 billion while imports decline at 48.5%; imports of LED packages declined by 16.4%; declined 9.2% from $51 billion in 2014 to $46.3 billion in 2016. and imports of PV modules remained relatively flat, declining China was the largest exporter of benchmark technologies while by just 0.7%. Imports of lithium-ion battery cells increased by the United States was the largest net importer. Exports of PV 13.3%. Some net importers of end products, such as the United modules, LED packages, and wind turbine nacelles and blades States, were major exporters of upstream processed materials declined (20.2%, 18.5% and 9.2%, respectively), while exports and/or subcomponents for the same technologies, illustrating of LIB cells expanded (19.4%) over the period. From 2014 to the complexity of clean technology manufacturing and trade.

9 Crosscutting Findings $ Value Added Trends

Benchmark Data: Value Added Trends

Clean energy manufacturing total value added (tVA) by clean energy technology, 2014–2016

China United South United States Japan Germany Korea India Taiwan Denmark Malaysia Brazil Canada Kingdom Mexico 15.0K

14.0K 45.0K PV Module Wind Turbine 13.0K LED Package Li-ion Battery Cell 40.0K

12.0K 11.76K 11.28K

11.0K 10.92K 35.0K 10.28K 10.0K 9.53K )

30.0K 9.06K 9.0K 8.20K 8.11K 8.0K 25.0K

ue Add e d ($ milli o n 7.0K Va l 6.34K a l 6.18K 6.09K o t 20.0K 5.89K

T 6.0K

5.0K 15.0K 3.79K 4.0K 3.76K 3.59K 3.38K 3.32K 3.30K 3.07K

10.0K 2.85K

3.0K 2.75K 2.63K 2.64K 2.62K 2.57K 2.54K 2.34K 2.0K 5.0K 1.51K 1.23K 1.15K 1.11K 1.07K 1.04K 1.0K 0.40K 0.33K 0.32K 0.0K 0.0K 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Total value added (tVA) trends in US$(2014) million from manufacturing of four clean energy technology end products in 13 key economies. Economies are ordered by tVA. Note that tVA for China is displayed on a different scale. Data are listed in order of 2016 tVA.

Total value added (tVA) from production of the four benchmark driven by decreases in revenues from wind turbine component technologies increased by $89.6 billion from 2014 to $102.4 manufacturing. billion in 2015, and then dropped to $87.3 billion in 2016. The United States moved from third to second highest tVA While tVA decreased on a dollar basis, physical unit sales for among the economies, jumping from $9.1 billion in 2014 to $10.3 the benchmark technologies actually increased because of billion in 2016, due largely to increasing wind turbine and lithium- significant technology price declines (see discussion on global ion battery pack manufacturing. Japan moved from second to average selling prices at end of Crosscutting section). third due to a drop in tVA from $11.3 billion in 2014 to $9.5 billion From 2014 to 2016, China accrued the largest tVA from in 2016, mainly from decreased tVA from PV module and LED manufacturing the four clean energy end products16 in all package production. The greatest amount of tVA growth from three years. China’s tVA grew from $39.0 billion in 2014 to 2014 to 2016 was experienced by Germany and the United States, $46.7 billion in 2015, and then dropped to $36.8 billion in 2016, with increases of $1.9 billion and $1.2 billion, respectively.

16 Only final products are included to avoid double-counting of indirect VA numbers. For example, solar cells are used to make modules, so cells are part of indirect VA for modules. Adding tVA for cells and modules would double-count cells. Final products are defined as solar PV modules; LED cells; LIB cells for vehicles; and nacelles, generators, and towers for wind turbines.

10 Value Added Trends $ Crosscutting Findings CC - Crosscutting VA Story

CC5. 2014 - 2016 Total CC6. 2014-2016 CEM CC7. 2014-2016 CE FYI dVA, dO, and FYI Calcs for CC5 VA by Clean Energy Direct and Indirect VA TVA Impact Shares VAretained discussion Clean energyTechnology manufacturingand VA Ret atotalined by C.. value added (tVA) by value added component, 2014–2016

United South United China Germany Japan India Denmark Malaysia Brazil Taiwan Canada Mexico States Korea Kingdom 15.0K

14.0K 45.0K

13.0K

40.0K 12.0K

11.0K 35.0K

10.0K ) 30.0K 9.0K

8.0K 25.0K

ue Add e d ($ milli o n 7.0K Va l

a l 20.0K o t 6.0K T

5.0K 15.0K

4.0K

10.0K 3.0K

2.0K 5.0K 1.0K

0.0K 0.0K 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Total value added (tVA) for the period in US dollars (2014$) from manufacturing of four clean energy technology end products in the 13 economies. Darker shading indicates direct value added (dVA), and lighter shading indicates indirect value added (iVA). Economies are listed in order of dVA in 2016. Note that tVA for China is displayed on a different scale than the other economies.

Across the economies analyzed, indirect value added (iVA) Note that this benchmark report does not include a detailed from the four benchmark technologies was greater than direct decomposition of sources of indirect value added. value added (dVA), demonstrating the larger amount of indirect The United States, the United Kingdom, Canada, and value added through processing materials, manufacturing Germany retained the greatest shares of tVA as a portion intermediary components, and providing services throughout of manufacturing revenue. This metric reflects the extent of the supply chain instead of from directly manufacturing domestic supply chains as well as prevailing wages, domestic the clean energy technologies themselves. China’s dVA profits, and taxes less subsidies. and iVA from manufacturing the benchmark technologies was significantly greater other than all other economies.

11 Crosscutting Findings $ Value Added Trends

Clean energy manufacturing total value added (tVA) domestic and non-domestic contribution, 2014–2016

South United United Country Brazil Canada China Denmark Germany India Japan Malaysia Mexico Taiwan Korea Kingdom States Brazil Canada 100% China Denmark Germany India 90% Japan Malaysia Mexico 80% South Korea Taiwan United Kingdom United States 70%

60%

50% Percentage of t otal VA

40%

30%

20%

10%

0% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

For each economy (listed across the top), color-coded bars show the share of tVA accrued from domestic and non-domestic production of clean energy technologies for 2014 to 2016. Domestic bars (generally, the largest in each column) represent the share of tVA (iVA plus dVA) from domestic production. Non-domestic bars represent the share of tVA (iVA only) from production in other economies (dVA only occurs in the economy where production occurs).

As a consequence of global supply chains associated with the was the largest supplier of materials and components to all the production of the four benchmark technologies, all analyzed other countries allowing it to contribute the most iVA to the economies received indirect value added from the production other benchmarked economies. In the benchmarked economies, of intermediate material, subcomponents, or services related the greatest share of tVA is generally accrued from domestic to end product manufacturing of PV modules, wind turbine production of clean energy technologies. Exceptions were the components, LED packages, and lithium-ion battery cells United Kingdom and Taiwan, where iVA accrued from clean in other economies. For example, in 2016, the United States energy manufacturing in China was greater than tVA accrued received $3.4 billion in iVA from manufacturing in the other from domestic production. economies, comprising 33.3% of the $10.3 billion U.S. tVA. China

12 Value Added Trends $ Crosscutting Findings

National gross domestic product and clean energy manufacturing contribution to economy- wide manufacturing, 2014–2016

CEM Contribution to Economy-Wide Manufacturing VA (%) Country 0.10% 0.20% 0.30% 0.40% 0.50% 0.60% 0.70% 0.80% 0.90% 1.00% 1.10% 1.30% 1.40% 1.50% Name Year 1.20% Denmark 2014 2015 2016 Malaysia 2014 2015 2016 Taiwan 2014 2015 2016 Brazil 2014 2015 2016 Canada 2014 2015 2016 China 2014 2015 2016 South Korea 2014 2015 2016 India 2014 2015 2016 Germany 2014 2015 2016 United 2014 States 2015 2016 Japan 2014 2015 2016 United 2014 Kingdom 2015 2016 Mexico 2014 2015 2016 $0B $2,000B $4,000B $6,000B $8,000B $10,000B $12,000B $14,000B $16,000B $18,000B GDP

Total bar length shows national gross domestic product (GDP) in US(2014$), gray shading indicates portion of GDP contributed by all manufacturing in a given economy (bottom axis). Squares indicate the percentage of tVA from domestic clean energy manufacturing (does not include iVA from non-domestic manufacturing) as a fraction of GDP from economy-wide manufacturing (top axis). Data are presented in the order of each economy’s domestic clean energy manufacturing share of national GDP.

Domestic clean energy manufacturing is a small contributor (0.03% to 1.2%) to national gross domestic product in all economies analyzed. The four benchmark technologies contributed the most to manufacturing sectors in Denmark, Malaysia, and Taiwan (the three smallest economies analyzed in this report).

13 Crosscutting Findings

CH1-Challenges: Crosscutting Story Comparing Clean Energy Manufacturing Trends Over Time CC-1. 2014-2016 CC-2. 2014-2016 CC-3. 2014-2016 CEM CC-4. 2014 -2016 Trade CC-9. Demand/Price FYI Calc for CC2 Production and Production - Demand Excess Capacity Flows by Tech End Trends discussion Impact ofDema globalnd Shares averageTrends selling price (ASP) on globalProduct demand trends, 2014–2016

Wind PV Module LED Chip LIB Cell 40K 10K 6K 60K 8K 5K 30K

6K 4K 40K 20K 3K 4K Demand ($ million) Demand ($ million) Demand ($ million)

Demand ($ million) 2K 20K 10K 2K 1K

0K 0K 0K 0K 80K 30 200K 150K 60K 150K 20

100K 40K 100K Demand (MW/year)

Demand (GWh/year) 10 Demand (MW/year)

50K 20K Demand (packages/year) 50K

0K 0K 0K 0

0.07 0.70 1.10 350

0.06 300 1.00 0.60 0.05 250 ASP ($/MW) ASP ($/GWh) ASP ($/MW) ASP ($/package) 0.90 0.50 0.04 200

0.80 0.40 0.03 150 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Lines (top) show global demand for the end products in US$(2014) million/year. Gray bars (middle) show demand in physical units/ year (i.e., MW for Wind and PV, number of packages for LEDs, and GWh for LIB). Lines (bottom) show estimated global average selling price for each benchmarked technology over the period. Note that each technology is presented on a different scale.

Because demand, production, and manufacturing capacity declined by 26% on a dollar-per-year basis over the same are measured in different physical units for each of the four period. For the other technologies, while the general unit-per- technologies (e.g., gigawatt-hours for LIB cells and megawatts year and dollar-per-year trends were the same, the magnitude for PV modules), the market benchmarks reported here are of the rates of change were different. The decline in wind normalized to a dollar-per-year basis to enable comparison and demand on a dollar-per-year basis was great enough to aggregation across technologies. However, this approach may impact the aggregate demand trend for the four clean energy not fully discern trends for physical units when the global average technology end products. selling price (ASP) changes significantly during the period. Consideration of ASP trends can provide additional context For example, global demand for wind turbines increased by when interpreting aggregated market and VA benchmark 6.2% on a megawatt-per-year basis from 2014 to 2016 but results and trends in this report.

14 Manufacturing Wind Turbines

Wind Turbines Wind is the second largest source of renewable electricity generation behind hydropower, with approximately 563 GW deployed globally at the end of 2018 (IRENA n.d.). From the beginning of 2014 to the end of 2016, roughly 167 GW of new capacity was added globally, growing the cumulative capacity from 300 GW to 467 GW (an increase of 56%) (IRENA n.d.). This growth represents approximately $350 billion of new wind investment for the same period (IRENA n.d.).

Wind turbine components and supply chain

Wind turbine supply chain alignment with Clean Energy Manufacturing Analysis Center (CEMAC) benchmark framework. Boxes highlight components included in the benchmark analysis. No raw materials were included in the analysis because of a lack of data that could link specific materials to end product manufacturing. Illustration by Josh Bauer, NREL

The modern wind turbine is composed of more than 8,000 individual subcomponents (EWEA 2009). The majority of these subcomponents (especially smaller ones) are produced and transported globally. However, once assembled into intermediates or end products, they often remain “in-country” because challenges caused by their size and weight. Approximately 90% of the value of these subcomponents is reflected in estimated prices for three main components nacelles, blades, and towers (Moné et al. 2015). This analysis tracks these three high value components in addition to steel.

Continued development of offshore wind and more moderate wind-speed resource areas have created opportunities for innovation, including taller towers, longer blades, and lower-weight nacelles and rotors. These advances generally expand the accessible wind resource. Significant evolution of the global supply chain is anticipated as manufacturers evaluate further cost-cutting measures, such as consolidation and lower cost centers, and as they begin to deploy smart factories and advanced manufacturing methods.

15 Manufacturing Wind Turbines

Notable Trends Preference for Domestic Manufacturing For wind energy technologies, domestic market Key drivers of wind turbine supply chain trends include: demand drives domestic manufacturing of the end products and, to a lesser extent, the upstream supply • Declining wind turbine component prices resulting chain. This domestic production alleviates cost and from maturing supply chains logistic challenges associated with transporting large • Expiration and reduction of subsidies to promote wind turbine component imports. As a result, most wind turbine manufacturing and deployment of the tVA from wind turbine component production was accrued from domestic manufacturing, with less • Preference for domestic production due to cost and inter-economy trade than typically seen with other logistic challenges associated with transporting technologies. With the exception of the United States, large wind turbine components. key wind turbine manufacturing economies’ domestic production was able to meet domestic demand Wind Turbine Component Prices and across the supply chain. tVA from manufacturing wind Competitiveness turbine components was greatest for China, followed Global average capacity-weighted installed wind costs by the United States, Denmark, and Germany, the four (in US$(2014)) decreased from $1,655/kW in 2014 largest wind turbine component producers. to $1,518/kW in 2016 (8% decrease) (IRENA 2020). In the United States, the average capacity-weighted Cost Competitiveness and Advanced project cost from 2014 to 2016 fell from $1,743/kW to Manufacturing $1,620/kW (7% decrease) (Wiser and Bolinger 2019). Continuing advances in wind turbine technology In 2016, the average generation-weighted levelized (larger rotors, taller towers, deployment in lower- power purchase agreement price in the United States quality wind areas), combined with advanced was $24.34/MWh, while today prices are estimated to manufacturing approaches (including on-site additive be below $18.46/MWh in some parts of the country manufacturing and 3D printing, automation, robotics, (Wiser and Bolinger 2019). Despite the price declines advanced sensing, and smart/adaptable floorplans), through 2016, wind turbines remained the most capital have the potential to circumvent issues related to intensive of the technology end products evaluated component transportation. over the period. As a result of higher production revenues from wind component manufacturing, More detailed information can be found on the economies derived greater value added from following pages. manufacturing wind components than from other clean energy technology intermediates.

Expiration of Subsidies Uncertainty surrounding renewal of the U.S. renewable electricity production tax credit (PTC) contributed to 2015 peaks in wind energy technology demand, production, trade flows, and value added. Drops were seen in 2016 in all these metrics across the economies studied, due in part to the PTC expiration and reduction of similar subsidies in China and Germany. Declining subsidies and market maturation can also put increased pressure on the supply chain to lower prices.

16 Market Trends Manufacturing Wind Turbines

Wind Turbine Supply Chain, 2014-2016 Market and Trade Trends Benchmark Data: Market Trends Wind 1. 2014-2016 Wind 2. Wind 3. 2014 - 2016 Wind 4. 2014-2016 App/Wind 2. WInd - FYI calc for wind 2 FYI calc for wind 4 Wind Production and 2014-2016Production Wind Demand, Wind Genset Gensets Detailed BOT WindDeman dturbine (% contributi.. supplyDemand Tre chainnds demandProduction, M fgand Capa.. productionImport/Export/BOT T ..shares by economy, 2014–2016

Nacelle Blades Tower Generator Steel Brazil Canada 100% 4% 4% 4% 4% 4% 5% 5% 5% 5% 5% China Denmark Germany India 80% Japan

43% 43% 43% 43% 43% Malaysia 45% 45% 45% 45% 45% 48% 48% 48% 48% 48% Mexico South Korea 60% Taiwan

nd Sh are s United Kingdom United States 10% 10% 10% 10% 10%

De m a Rest of World 10% 10% 10% 10% 10%

40% 9% 9% 9% 9% 9% 7% 7% 7% 7% 7% 15% 15% 15% 15% 15%

20% 9% 9% 9% 9% 9% 14% 14% 14% 14% 14% 17% 17% 17% 17% 17% 17% 17% 17% 17% 17% 13% 13% 13% 13% 13% 0%

100% 4% 4% 5% 6% 6% 6% 8%

80% 44% 46% 49% 46% 49% 50% 50% 45% 49% 45% 45% 47% 47% 50% 50%

60% on Sh are s t i 8% 8% 12% odu c 11% 6% 10% 5% 5% 6% P r

40% 18% 7% 9% 12% 12% 7% 6% 6% 16% 7% 11% 11% 10% 11% 8% 5% 7% 6% 4% 4% 7% 7% 10% 7% 7% 7% 5% 4% 7% 5% 5% 20% 5% 8% 6% 7% 10% 13% 13% 12% 10% 27% 12% 24% 20% 20% 20% 13% 16% 15% 15% 13% 9% 9% 8% 8% 0% 6% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Breakdown (in %) of global demand (top) and global production (bottom) by economy for benchmarked wind turbine supply chain intermediates.

China accounted for the largest share of demand and Wind turbine component demand and production production across the wind turbine supply chain over were generally colocated on a regional basis due to the period. Germany and the United States showed transportation logistic challenges associated with moderate increases in shares of global demand and their size and weight. Outside China, wind turbine production for nacelles, blades, and towers. production was led by the United States, Germany, India, Denmark, and Brazil.

17 Manufacturing Wind Turbines Market Trends

Wind Turbine Supply Chain, 2014-2016 Market and Trade Trends

Wind 1. 2014-2016 Wind 2. Wind 3. 2014 - 2016 Wind 4. 2014-2016 App/Wind 2. WInd - FYI calc for wind 2 FYI calc for wind 4 Wind Production and 2014-2016Production Wind Demand, Wind Genset Gensets Detailed BOT WindDemand (%turbine contributi.. supplyDemand Tren chainds productionProduction, Mfg Capa.. andIm pdemandort/Export/BOT T.. trends, 2014–2016

China Germany United States Denmark India Brazil 10K 30K

20K Nacelle 5K MW/yr 10K

0K 0K 10K 30K

20K Blades 5K MW/yr 10K

0K 0K 10K 30K

20K Tower 5K MW/yr 10K

0K 0K 10K 30K

20K Generator 5K MW/yr 10K

0K 0K 10K 30K

20K Steel 5K MW/yr 10K

0K 0K 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Demand (color-coded lines) and production (gray bars) trends (in MW) by economy for wind turbine supply chain intermediates. Economies are listed in order of 2016 production levels Economies are color-coded throughout this benchmark report. China data are displayed on a different scale than those of the other countries. Because of a lack of Chinese demand data for wind turbine components and intermediates, demand and production were assumed to be equal. Because China made little inroads into the broader western market during 2014–2016, this can be considered a fairly robust assumption.

18 Market Trends Manufacturing Wind Turbines

Global demand (in MW) for wind turbines increased in India’s demand for wind increased in both in 2015 and 2015, followed by a modest downturn in 2016 (with the 2016, supported by the implementation of several expiration of renewable energy policies in China, the favorable policies, including state-level feed-in tariffs United States, and Germany), for a net increase of 6% beginning in 2014. from 2014 to 2016.17 Global production (in MW) of wind turbine supply chain intermediates followed a similar Following demand, global production for the wind global trend. With the exception of the United States, turbine supply chain increased from 51.5 GW in 2014, key wind turbine manufacturing economies generally and to 63.5 GW in 2015, followed by a decline to 54.6 had sufficient domestic production to meet domestic GW in 2016. Germany and the United States saw the demand across the supply chain. highest growth in production for nacelles, blades, and towers over the three-year period. Germany remained Global demand for wind turbine components (nacelles, the second-largest exporter of nacelles and blades, blades, and towers) increased by 23% from 51.5 GW behind Denmark, which saw a dip in blade and nacelle in 2014 to 63.5 GW in 2015, and then decreased 14% production. Denmark’s nacelle, blade, and generator to 54.6 GW in 2016. China had the highest demand production rose in 2015, before dipping in 2016. for and production of wind turbine and supply chain intermediates among the 13 economies. Domestic market demand tended to drive domestic production of the end products and, to a lesser extent, The three economies with the highest levels of the upstream supply chain, over the period. Of the key demand—China, the United States, and Germany— wind economies considered, only the United States did all followed a trend of demand increasing in 2015, not have sufficient domestic manufacturing capacity followed by a drop in demand in 2016. The 2015 uptick to meet domestic demand for all intermediates, other in China may have been driven by a race to build, as than nacelles, in 2016. many provinces had renewable energy targets tied to the end of that year. The follow up to a year of unusually high demand, combined with a reduction in national feed-in tariffs, likely contributed to the downturn in 2016 (BNEF 2017a).

The United States saw an uptick in demand from approximately 4.8 GW in 2014 to approximately 8.6 GW in 2015, most likely associated with an anticipated expiration of the federal production tax credit. When the credit was extended, the United States saw a slight decrease, but 2016 levels were still higher than 2014 levels (BNEF 2018a). In Germany, with the announcement that onshore wind feed-in tariffs were being replaced by market premiums for new wind projects, demand for wind in 2016 dropped to 2014 levels (BNEF 2017b). Germany also reduced its target for offshore wind capacity by 2020 from 10 GW to 6.5 GW during the period (BNEF 2018b).

17 Due to the steep drop in the global ASP of wind turbine components, on a megawatt-per-year basis, global demand for wind turbines increased by 6.2%, but on a dollar- per-year basis, demand declined by 26% over the period. See Challenges: Comparing Clean Energy Manufacturing Trends Over Time on p. 14 for details.

19 Manufacturing Wind Turbines Market Trends

Wind turbine supply chain manufacturing capacity utilization, 2014–2016

2014 2015 2016

Capacity Utilization Capacity Utilization Capacity Utilization

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%

China Nacelle 52% 48% 39% Blades 58% 67% 45% Generator 98% 69% 43% Tower 99% 70% 47% Steel 73% 70% 72%

0K 10K 20K 30K 40K 0K 10K 20K 30K 40K 0K 10K 20K 30K 40K $ million $ million $ million

Brazil Nacelle 57% 69% 48% Blades 49% 58% 41% Generator 106% 74% 82% Tower 94% 50% 54% Steel 71% 69% 63% Denmark Nacelle 100% 69% 56% Blades 100% 69% 81% Generator 0K 0K 0K Tower 90% 57% 48% Steel 0K 0K 0K Germany Nacelle 57% 54% 65% Blades 60% 64% 88% Generator 106% 74% 67% Tower 85% 85% 74% Steel 70% 67% 68% India Nacelle 30% 36% 40% Blades 38% 59% 38% Generator 55% 45% 40% Tower 85% 31% 95% Steel 80% 75% 76% United Nacelle 57% 92% 73% States Blades 51% 68% 83% Generator 70% 62% 54% Tower 77% 90% 99% Steel 78% 71% 69% 0K 2K 4K 6K 0K 2K 4K 6K 0K 2K 4K 6K

100% 106% 85% 69%

a cit y 70% 68% 54% 74% 57% 60% 67% Ca p 50% 48% 57% 48% a l 51%

Utilization 44% o b G l

0%

Bars show manufacturing capacity (lighter shading) and utilized manufacturing capacity (i.e., production, darker shading) in US$(2014) for key wind turbine economies. Vertical lines and associated numerical values show capacity utilization (production as a % of manufacturing capacity). Trend lines show global capacity utilization percentage for 2014–2016 (bottom). Note that China is displayed on a different scale.

20 Market Trends Manufacturing Wind Turbines

In aggregate, global manufacturing capacity increased and capacity utilization declined for all wind supply chain intermediates, reflecting expanded manufacturing capacity built up in anticipation of increased demand. Capacity utilization rates in the individual benchmarked economies generally indicate an ability to expand production to meet wind demand growth.

Global manufacturing capacity for nacelles was estimated at 89.7 GW in 2014, 117 GW in 2015, and 115 GW in 2016. Corresponding production was estimated at 51.5 GW in 2014, 63.5 GW in 2015, and 54.6 GW in 2016, reflecting a global excess of manufacturing capacity that had been built up in anticipation of increasing demand. Globally, blade, generator, and tower capacity grew over the period while capacity utilization grew for these intermediates. China and India increased manufacturing capacity for nacelles, blades, towers, and generators in 2015, before curtailing nacelle capacity in 2016. Denmark’s capacity utilization declined each year for nacelles and towers, while capacity utilization for blade production increased in 2016.

21 Manufacturing Wind Turbines Trade Trends

Wind Turbine Supply Chain, 2014-2016 Market and Trade Trends Benchmark Data: Trade Trends Wind 1. 2014-2016 Wind 2. Wind 3. 2014 - 2016 Wind 4. 2014-2016 App/Wind 2. WInd - FYI calc for wind 2 FYI calc for wind 4 Wind Production and 2014-2016Production Wind Demand, Wind Genset Gensets Detailed BOT WindDeman dturbine (% contributi.. generatorDemand Trends sets (nacellesProduction, Mfg Cap a..and Imblades)port/Export/BOT trade,T.. 2014–2016

Year 2014 2015 2016

Brazil -$254M -$98M -$32M

Canada -$614M -$614M -$93M

China $294M $573M $1,144M

Denmark $3,743M $3,268M $3,035M

Germany $1,660M $1,926M $1,925M

India $83M $2M $1M

Japan -$84M -$85M -$72M

Malaysia $0M $0M $0M

Mexico -$527M -$389M -$540M

South Korea -$45M -$12M -$51M

Taiwan -$5M -$15M -$34M

United Kingdom -$590M -$328M -$325M

United States $407M -$84M -$106M

Imports Exports Imports Exports Imports Exports

South United United Brazil Canada China Denmark Germany India Japan Malaysia Mexico Korea Taiwan Kingdom States $4,000M

$2,000M Balance of Trade $0M 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Bar chart (top) shows imports (negative values), exports (positive values), and balance of trade (exports positive and imports negative) in US$(2014) by economy for wind turbine generator sets. Line chart (bottom) shows balance of trade trends for wind generator sets.

Denmark, Germany, and, increasingly, China were In 2016, the 13 benchmarked economies exported a the top exporters of wind turbine generator sets total of almost $6.4 billion of wind generator sets, (nacelle and blades) over the period. Net exports from down 9.3% from $7.0 billion in 2014. Imports were also Denmark declined, while China’s net exports increased. down 48.5%, from $2.9 billion in 2014 to $1.5 billion in Mexico, the United Kingdom and Canada were the 2016. The declining cost of wind turbine technologies largest importers. over the period contributed to this contraction in trade.

22 10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report Wind Trade Flow

 (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

(/index.html)

Trade TrendsGlobal summary of wind generatorManufacturing set (including blades Wind and nacels)Turbines trade flow, in 2014 USD

Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html)

2014 2015 2016

Imports Brazil Exports $269.0 M $15.0 M Wind turbine generator set (nacelles and $628.2 M Canada $591.2 M

United States of America $13.9 M blades) trade flows, 2014–2016 $184.5 M

$8.6 M China While global exports and imports of wind generator sets $4.0 M

$302.3 M (nacelles and blades) declined over the period, the trade United Kingdom

$593.5 M $547.2 M network among the benchmarked economies remained

$726 K intact and generally followed expected flows, given the Taiwan

$5.9 M distribution of manufacturing capacity across economies. Germany

10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report Wind Trade Flow

K 7 4 $

 (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

K 3 4 5 $ Between 2014 and 2016, the largest changes in wind Malaysia

$2.2 B

turbine genset trade flow were observed in Denmark,

K 0 $

M .7 2 4 China, Canada, and the United States. Denmark showed $

the largest change in net exports, declining from $3.7 (/index.html) Mexico M .5 9 6 5 $ Denmark

Global summary of wind generator set (including blades and nacels) trade flow, in 2014 USD

M .9 8 billion in 2014 to $3.0 billion in 2016. Exports from $

M .2 4 5

Batteries (/benchmark/battery.html)$ LED (/benchmark/led.html) PV Wind (/benchmark/wind.html)

Denmark to Germany and the United Kingdom dropped Korea South

M .7 1 $

K 1 0 2 $

M .0 3 8 $

M .3 5 8 $

B .7 3 2014 2015 2016 $ Japan by $270 million and $240 million, respectively, in part due 2014 India Imports to feed-in tariff cuts implemented in early 2016. China Brazil Exports

$98.2 M $521 K $621.5 M Canada

$147.3 M showed the largest increase in net exports—from $300 Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). United States of America $7.9 M million to $1.1 billion. With decreased domestic demand, $231.7 M $1.5 M

$9.0 M China due in part to feed-in tariff reductions, China’s imports https://www.jisea.org/benchmark/wind.html 1/2 United Kingdom $574.5 M from Denmark declined, while its exports to Mexico and $337.2 M $ 4 1 .6 M

the rest of the world increased. Canada’s wind capacity $504 K

Taiwan addition dropped from 1.9 GW to 0.7 GW, leading to $15.2 M

reduced imports from the United States, Germany, and Germany

K 7 1 $

Denmark over the period. The United States moved from 10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report Wind Trade Flow

Malaysia

K 6 1 5 $

a net exporter to net importer as exports to Brazil and  (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

M .8 6 2

Canada declined and imports from Germany increased. $ $ 2 .4 B

K 0 $

Mexico’s growing market primarily imported from China Mexico

M .8 5 1 4 $

and Denmark in 2016. (/index.html)

M .7 1 $

Global summary of wind generator set (including blades and nacels) trade Denmark flow, in 2014 USD

M .1 4 1 $ South Korea South

Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html)

M .1 2

Large wind turbine components are generally installed $

M .6 6 8 $

M .3 2 $ M .6 4 $

Japan 2014 2015 2016 B .3 3 $ in the country or region where they are manufactured, 2015 India

Imports because of either country-specific local content policies Brazil Exports $33.3 M $1.3 M Canada $103.0 M Back to the Benchmarks of Global Clean Energy Manufacturing$16.1 M (/benchmark.html). or high transportation and logistics costs associated with United States of America $10.1 M $122.5 M moving oversized components (e.g., assembled nacelles, $1.7 M $6.1 M China large blades) to project sites, so trade for these is limited https://www.jisea.org/benchmark/wind.htmlUnited Kingdom 1/2 $331.3 M compared to the other technologies. By contrast, raw $1.1 B $ 2 0 9 .9 M

$519 K materials, processed materials, and many of the thousands Taiwan of smaller wind turbine parts tend to be produced and $34.8 M shipped globally. Germany

$74 K

Malaysia

$329 K

Gray chords show wind generator set trade flows in US$(2014)

K 9 $ among benchmarked economies. Width of bars are scaled to total $ 2 .1 B

flows, with lighter shading showing imports and darker shading M .8 9 $

Note that gray chords only capture the trade among Mexico M .1 0 4 5 showing exports. $

the 13 benchmarked economies; trade flows with the rest of the world

K 9 8 5 $

are not shown. Denmark

M .1 1 5 $

South Korea South

K 4 1 1 $

M .1 2 7 $

M .8 6 $ M .9 7 $

Japan

B .0 3 $

2016 India

23 Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html).

https://www.jisea.org/benchmark/wind.html 1/2 Manufacturing Wind Turbines $ Value Added Trends

WindBenchmark Value Added Data: Story Value Added Trends

Wind 6. WInd-x Wind Wind 7. Wind Total FYI Wind tVA calc for FYI - dVA, dO, VA Value Added (dVA, iVA Value Added Impacts Wind 6 retained Wind turbine supply achainnd tVA) total valueof Glo baddedal Production ((tVA).. by value added component, 2014–2016

China Germany United States Denmark India Brazil

15,320 4,219 15.0K 13,677 4.0K 3,511 3,624 3,279 3,280 3,265 10,207 2,655 10.0K 1,873 Nacelle Nacelle $ million $ million 2.0K 1,590 1,556 1,432 1,526 1,226 1,351 5.0K 955

0.0K 0.0K 1.5K 5,750 1,332 6.0K 1,247

3,893 4,131 1.0K 917 4.0K 884 820 777 648 673 636 594 590 Blades Blades 550 $ million $ million 480 471 2.0K 0.5K 274

0.0K 0.0K 6,732 2.0K 1,881

6.0K 1,558 1,512 4,401 1,269 3,932 4.0K 1.0K 863 917 Tower Tower 828 805 $ million $ million 651 545 592 460 430 2.0K 353 216 0.0K 0.0K 4,612 1,503 1.5K 4.0K 3,549 3,616 1,006 1,029 1.0K 896 761 680

$ million 607 $ million

Generator 2.0K Generator 504 538 0.5K 340 303 320

0.0K 0.0K 10 12 12

1.5K 1,437 163 1,244 1,196 132 126 116 1.0K 104 107 0.1K 77 Steel Steel 72 67 $ million $ million 63 59 64 0.5K

0.0K 0.0K 2 2 2 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Total value added (tVA) for the period in in US$(2014) across the supply chain for key wind turbine economies. Darker shading indicates direct value added (dVA), lighter shading indicates indirect value added (iVA), and numerical values show total value added (tVA). Economies are listed in order of tVA from nacelle production in 2016. Note that tVA data for China are displayed on a different scale.

24 Value Added Trends $ Manufacturing Wind Turbines

China, Germany, and the United States accrued more total value added (tVA) from manufacturing wind turbine components than the other benchmarked economies over the period. Indirect value added (iVA) was higher than direct value added (dVA) across the wind turbine supply chain, with the exception of towers in Denmark and blades in China. tVA (iVA plus dVA) is generally greatest in the economies with the highest levels of production revenue. In 2016, China, the United States, and Germany, respectively, accrued $18.7 billion, $6.0 billion, and $5.8 billion in tVA , in aggregate from manufacturing nacelles, blades, and towers for wind turbines. Across the supply chain, China’s tVA was about equal to the sum of the tVA of all other benchmarked economies combined, with the exception of steel. China, the United States, and India generally experienced peaks in tVA for wind intermediates in 2015, with the exception of generators in India and steel in all economies. iVA and dVA generally showed similar trends over the period. iVA was higher than dVA for all benchmarked economies across the wind turbine supply chain, with the exception of tower manufacturing in Denmark and blade manufacturing in China, demonstrating that, in general, the supply chains that support wind turbine component manufacturing domestically and internationally were more important to the benchmarked economies than domestic manufacturing of these components and end product.

Between 2014 and 2016, both dVA and iVA from domestic production of steel (primarily for towers) declined slightly in the economies considered. Presumably, this reflects steel’s classification as a global commodity and wind turbines’ fractional value relative to other steel applications (e.g., construction, vehicles).

25 Manufacturing Wind Turbines $ Value Added Trends Wind Value Added Story

Wind 6. WInd-x Wind Wind 7. Wind Total FYI Wind tVA calc for FYI - dVA, dO, VA Wind turbine supplyValue Adde dchain (dVA, iVA totalVa valuelue Added addedImpacts (tVA)Wind 6domestic andre non-domestictained contribution, 2014–2016and tVA) of Global Production (..

Brazil China Denmark Germany India United States Country 100% Brazil

VA Canada China 50% Denmark Blades Germany India Co nt r ib u ti o n t 0% 100% Japan

VA Malaysia Mexico 50% South Korea

Nacelle Taiwan United Kingdom Co nt r ib u ti o n t 0% United States 100% VA

50% Tower

Co nt r ib u ti o n t 0% 100% VA

50% Generator

Co nt r ib u ti o n t 0% 100% VA

50% Steel

Co nt r ib u ti o n t 0% 2014 2016 2014 2016 2014 2016 2014 2016 2014 2016 2014 2016

For each economy (listed across the top) color-coded bars show the share of tVA accrued from domestic and non-domestic production of wind turbine supply chain intermediates for 2014 to 2016. Domestic bars (generally the largest in each bar) represent the share of tVA (dVA + iVA) from domestic production. Non-domestic bars represent the share of tVA (iVA only) from production in other economies (dVA only occurs in the economy where production occurs).

Most of the tVA received from the production of wind VA retained) from manufacturing wind turbine turbine nacelles, blades, towers, generators, and steel components—ranging from 16% to 48% across the in 2016 came from domestic manufacturing of those supply chain. The highest levels of VA retained components. With its rapidly growing wind turbine occurred in the United States from manufacturing component manufacturing capacity, production in blades (48%), generators (42%) and nacelles (38%); by China generally contributed the most to non-domestic Denmark for towers; and by Brazil for steel production. iVA in the benchmarked economies across the supply VA retained reflects the extent of domestic supply chain. chains as well as prevailing wages, domestic profits, and taxes less subsidies. Benchmarked economies retained varying shares of tVA as a portion of manufacturing revenue (i.e.,

26 Manufacturing Wind Turbines

Wind Challenges and Opportunities: Cost Competitiveness and Advanced Manufacturing

Potential pathways to 50% LCOE reduction in 2030 by LCOE parameter

$50 46 5 $45 8 $40 Net 21% CapEx $35 reduction 3 through $30 wind plant Net 22% increase 1 5 economies of in energy Net 25% OpEx $25 Net 5% cost 23 scale, turbine production reduction of capital from $20 scaling with through turbine reduction from Net 26% cost of capital less material scaling, advanced increased reduction from $15 use, and enhanced operation and certainty of increased plant life due LCOE (US$[2015]/MWh) LCOE ecient control strategies maintenance future plant to improved design and advanced operation $10 manufacturing and reducing strategies performance and maintenance wind plant losses and $5 strategies reduced risk $0 2015 Baseline CapEx AEP OpEx Financing Plant Life 2030 Conceptual Plant Plant Source: Dykes et al. 2017 CapEx: capital expenditure, OpEx=operating expenses, AEP=annual energy production

Competitive pressure and the potential development and Bolinger 2017). This trend of growth in rotor sizes of offshore wind and more moderate wind-speed and growth in the size and scale of wind turbines in resource areas continue to create opportunities for wind general as a means of system level cost reduction technology innovation, including taller towers, longer is expected to continue and may become an even blades, and lower-weight nacelles and rotors. A 50% more significant opportunity in the offshore market reduction in wind’s levelized cost of electricity (LCOE) by segment. 2030, as shown above, is expected to require both larger • Towers: The average hub height of turbines in 2016 rotor diameters and taller towers. Some of the impacts of was 83 meters, 48% taller than turbines built two these advancements include: decades earlier. The increased turbine height enables • Assembled Turbines: Further price reductions the machines to tap into a higher quality and more are predicated on less costly and more efficient consistent wind resource which improves their total subcomponents, improved manufacturing processes, energy production (Wiser and Bolinger 2017). and economies of scale. For example, BNEF found • Generators: Alongside blade length and tower height, that between 2014 and 2016, nominal turbine prices generator size has also increased while decreasing decreased from $1,200/kW to $1,070/kW (11% in weight, enabling turbines to consistently produce 18 decrease), and fell to roughly $700/kW in 2019. more energy. • Blades: The average rotor diameter of wind turbines installed in 2016 was 108.0 meters, 127% larger than those of 1998–1999 and 13% larger than the average blade size of the previous five years (2011–2015) (Wiser

18 Bloomberg New Energy Finance. 2019. 2H 2019 Turbine Price Index.

27 Manufacturing Wind Turbines

Accompanying these cost reduction opportunities is segmentation, advanced manufacturing, and advanced the potential for increased deployment in both lower logistics to overcome the escalating costs associated and higher wind speed areas. Expansion into lower- with transporting large components. Advanced quality wind resource regions, in particular, is expected manufacturing—which includes additive manufacturing to require more detailed site-specific and terrain- (3D printing), automation, robotics, advanced sensing, optimized solutions that present further manufacturing and smart/adaptable floorplans—has the potential and logistics challenges. In some cases, the increase in to enable novel solutions (onsite manufacturing, site component size can limit transport options and cause specific designs) that are not feasible using current intermediary subcomponents to be manufactured manufacturing processes. Significant evolution of the domestically in order to reduce transport costs. global supply chain is anticipated as manufacturers This trend is expected to continue as components begin to deploy these methods and evaluate further increase in size. Original equipment manufacturers cost-cutting measures, such as consolidation and lower will need to consider innovative solutions such as cost centers.

28 Manufacturing c-Si PV Modules

Crystalline Silicon Photovoltaic Modules Solar energy is the third largest source of renewable electricity generation behind hydropower and wind, with approximately 487 GW deployed globally at the end of 2018 (IRENA n.d.). From the beginning of 2014 to the end of 2016, roughly 155 GW of new photovoltaic (PV) capacity was added globally, growing the cumulative capacity from 136 GW to 291 GW (an increase of 114%). This growth represented approximately $468 billion of new PV investment for the same period (IRENA n.d.).

Crystalline silicon photovoltaic (PV) modules and supply chain

Crystalline silicon (c-Si) photovoltaic (PV) module alignment with Clean Energy Manufacturing Analysis Center (CEMAC) benchmark framework. Boxes highlight components included in the benchmark analysis. No raw materials were included in the analysis for a lack of data. Illustration by Josh Bauer, NREL

Crystalline silicon-based technologies are considered mature and account for more than 90% of the PV market (Fraunhofer ISE 2018). Other PV technologies, such as cadmium telluride and copper indium gallium selenide thin film-based modules, are commercially available via only a limited number of producers. Researchers focus on key elements of the manufacturing supply chain for the crystalline silicon (c-Si) PV module manufacturing process—polysilicon, PV wafers, and PV cells. The market for PV modules in 2016 favored polycrystalline over monocrystalline 63% to 37% (ITRPV 2018).

The global manufacturing network for PV is well established, but opportunities for innovation remain, including cost-effective production of advanced cell architectures, enhancement of polysilicon (poly-Si) purity, and other improvements to cell and module efficiencies.

29 Manufacturing c-Si PV Modules

Notable Trends of the end product. In addition, because none of the four economies could meet its own domestic demand Key drivers of PV module supply chain trends over the for all PV module components, trade of intermediates period included: occurred. Net exporters of PV module end products in some cases were net importers of intermediates such • Declining cost of solar projects, driven by as polysilicon. manufacturing capacity increases, manufacturing cost reductions, module efficiency improvements, China and module and supply chain intermediate surpluses in the global market With the highest demand, production, and manufacturing capacity of the economies considered, • Global demand growth, in part driven by declining China drove global PV module supply chain market prices and improved module efficiencies trends and also received the highest tVA from PV manufacturing. The economic benefit of PV • Shifting markets that stimulate policies and manufacturing in China also contributed iVA to all the equipment contracting approaches, including benchmarked economies. China’s policies focused installation targets, tax credits, auctions, and targeted on building manufacturing capacity and domestic tariffs (Kavlak, McNerney, and Trancik 2018). supply chains, along with import tariffs levied by other economies on specific PV supply chain components Manufacturing Capacity and Inventory and intermediates, helped shape the global market Increased demand for PV modules was driven in part and trade landscape. by domestic policies that set targets for deployment of renewable technologies or provided incentives to Balance of System Components offset costs. For example, in the United States, PV A range of balance of system (BOS) components— installations grew through the period, as installers inverters, structural hardware, and electrical anticipated the Investment Tax Credit expiration hardware—are required for PV module system (BNEF 2015), supported by significantly increased installation and operation. Markets for these U.S. imports of PV modules from China from 2014 to components tend to be highly fragmented, with 2016. Global manufacturing capacity expansion (led U.S. PV installations sourcing BOS equipment from by China) to meet anticipated increases in demand domestic and foreign suppliers. The price of these resulted in excess global manufacturing capacity components, for both residential and utility-scale for all benchmarked photovoltaic (PV) supply chain applications, dropped significantly over the period. intermediates and contributed to significant price drops over the period. While China maintained its More detailed information can be found on the position as the largest PV module and cell exporter, following pages. increased Chinese demand and an international buildup of manufacturer and installer inventory levels contributed to reduced Chinese exports in 2016.

Global Supply Chains In most economies, indirect value added (iVA) from manufacturing PV modules and intermediates was greater than direct value added (dVA) over the period, indicating that participation in the broader supply chain that supports PV module manufacturing at home and abroad was more important to the benchmarked economies than domestic production

30 Market Trends Manufacturing c-Si PV Modules

PV Module Supply Chain, 2014-2016 Market and Trade Trends Benchmark Data: Market Trends PV 1: 2014-2016 PV PV 2.. 2014 - 2016 PV PV 3: 2014 - 2016 PV PV 4: 2014 PV Trade App/PV2: 2014 - 2016 App/PV 3: 2014-2016 App/PV 4: 2014-2015 FYI Production and Production&Demand Demand, Production, Bar and Trend Chart Detailed PV Module Detailed PV Cell BOT Detailed Polysilicon for Calc PVDemand module (%) contribu.. supply(MW) chain demandMfg Capacity (uandpdate.. production sharesBOT for by Key Coueconomy,ntries for key 2014–2016 countries key countries for P..

PV Module PV Cell PV Wafer Polysilicon Brazil 100% Canada China Denmark 26%

80% 32% Germany 42% India Japan 67% 61% 68% Malaysia 70% 71% 74% 60% 77% Mexico 82% 85%

22% South Korea 7%

20% Taiwan

Demand Shares United Kingdom 40% 11% 6% United States

7% Rest of World 14% 5% 6% 5% 5% 14% 21% 20% 6% 6% 9% 9% 4% 22% 14% 17% 16% 5% 16% 8% 8% 12% 8%

0% 5% 100%

80% 44% 48% 49% 67% 68% 61% 70% 71% 74% 60% 77% 82% 85% 17% 16% 15% Production Shares 40% 5% 6% 16% 5% 5% 17% 6% 20% 19% 6% 9% 9% 4% 14% 18% 17% 16% 5% 12% 8% 10% 8% 8%

0% 5% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Breakdown (in %) of global demand (top) and global production (bottom) by economy for benchmarked PV module supply chain intermediates. Note that, due to lack of disaggregated data, demand for the upstream intermediate is estimated to equal the production of the preceding intermediate (e.g., demand for polysilicon equals production of PV wafers).

China accounted for the highest shares of global intermediates. Demand shares for PV modules showed demand and production of PV modules and supply the greatest change over the period, with Japan’s chain intermediates. PV module demand and demand declining while China and U.S. demand polysilicon production were broadly distributed across increased. The United States lost polysilicon production the benchmarked economies, reflecting China’s shares to China and South Korea over the period. gap in production relative to demand for these two

31 Manufacturing c-Si PV Modules Market Trends

PV Module Supply Chain, 2014-2016 Market and Trade Trends

GlobalPV 1: 2014-2016 PVPV modulePV 2.. 2014 - supply2016 PV chainPV 3: 2014 - demand2016 PV P Vand 4: 2014 PproductionV Trade App/PV 2for: 2014 -key 2016 PVApp module/PV 3: 2014-201 6economies,App/PV 4: 2014-201 5 FYI Production and Production&Demand Demand, Production, Bar and Trend Chart Detailed PV Module Detailed PV Cell BOT Detailed Polysilicon for Calc 2014–2016Demand (%) contribu.. (MW) Mfg Capacity (update.. BOT for Key Countries for key countries key countries for P..

China Malaysia South Korea Japan Germany United States Taiwan 60K 15K e

u l 40K 10K Mo d MW/yr V P 20K 5K

0K 0K 60K 15K

l 40K 10K Ce l V MW/yr P 20K 5K

0K 0K 60K 15K

er 40K 10K Wa f V MW/yr P 20K 5K

0K 0K 60K 15K

40K 10K y silic on MW/yr o l P 20K 5K

0K 0K 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Demand (color-coded lines) and production (gray bars) trends (in MW) by PV module supply chain intermediates for top PV module economies. Economies are listed in order of 2016 production levels. China data are displayed on a different scale than the other benchmarked countries.

Global demand and production (in MW) for PV Between 2014 and 2016, global demand for PV modules and supply chain intermediates grew from modules increased by 80% from 43.3 GW to 77.9 GW. 2014 to 2016. China had the highest demand for and After China, the United States, Japan, and India had production of PV modules, cells, and wafers, more the highest demand for PV modules over the period. than offsetting shifting demand and production in the The biggest increases were observed in China, with other benchmarked economies. With the exception of demand almost tripling from 11.2 GW to 32.7 GW, Germany, none of the 13 economies was able to meet and in the United States with demand increasing domestic demand for all supply chain intermediates in two-and-a-half times, from 6.3 GW to 16.0 GW. U.S. 2016 with domestic production alone. demand increased in part due to the anticipated 2015 expiration of the federal investment tax credit (ITC)

32 Market Trends Manufacturing c-Si PV Modules

and its subsequent extension in 2016. Japan’s annual demand for PV modules decreased from 9.6 GW to 8.3 GW, due in part to mid-2016 revisions to feed-in tariff laws, along with new curtailment rules in early 2015. Further upstream in the supply chain, global demand increased by 51.6% for PV cells (to 70.3 GW), 60.6% for wafers (to 65.7 GW), and 45.8% for polysilicon (to 69.7 GW), reflecting increased PV module demand. Malaysia, Taiwan, and South Korea had the highest demand for these supply chain intermediates over the period.

To meet increased demand, annual global production of PV modules grew by 51.6% from 46.3 GW to 70.3 GW between 2014 and 2016. For PV modules, the biggest increases in production were observed in China (up 45.2%), Malaysia (more than tripling), and South Korea (increasing two-and-a-half times). Between 2014 and 2016, global production of PV cells, wafers, and polysilicon increased by 60.8%, 45.7%, and 38.2%, respectively. Production increases in PV cells and wafers were driven by China, Malaysia, Taiwan, and South Korea. These gains offset decreases in Japan’s and South Korea’s PV wafer production. Growth in polysilicon production over the period was driven by increases in China, Germany, and South Korea, which offset decreases in U.S. polysilicon production, driven by reduced exports to China (due to tariffs) and price drops (due to oversupply).

Production of PV modules in China, Malaysia, South Korea, and Taiwan was greater than domestic demand over the period; PV modules from the economies were exported to economies where demand exceeded domestic production, including Japan, Germany, and the United States. The United States had the largest gap between production and demand for PV modules, increasing from $4.3 billion in 2014 to $7.5 billion in 2016. Japan showed the second largest gap, decreasing from $5.4 billion to $3.1 billion over the period.

33 PV ModuleManufacturing Supply Chain, c-Si PV 2014-2016 Modules Market andMarket Trade Trends Trends

PV 1: 2014-2016 PV PV 2.. 2014 - 2016 PV PV 3: 2014 - 2016 PV PV 4: 2014 PV Trade App/PV2: 2014 - 2016 App/PV 3: 2014-2016 App/PV 4: 2014-2015 FYI Production and Production&Demand Demand, Production, Bar and Trend Chart Detailed PV Module Detailed PV Cell BOT Detailed Polysilicon for Calc Demand (%) contribu.. (MW) Mfg Capacity (update.. BOT for Key Countries for key countries key countries for P.. PV module supply chain manufacturing capacity utilization, 2014–2016

2014 2015 2016

Capacity Utilization Capacity Utilization Capacity Utilization

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%

China PV Module 56% 54% 53% PV Cell 60% 63% 61% PV Wafer 72% 74% 80% Polysilicon 77% 93% 97% 0K 10K 20K 30K 40K 50K 0K 10K 20K 30K 40K 50K 0K 10K 20K 30K 40K 50K

$ million $ million $ million

Malaysia PV Module 62% 90% 87% PV Cell 65% 80% 81% PV Wafer 75% 59% 76% Polysilicon 18% 60% 60% South Korea PV Module 68% 71% 79% PV Cell 55% 82% 83% PV Wafer 65% 20% 28% Polysilicon 77% 86% 100% Germany PV Module 49% 47% 57% PV Cell 71% 74% 84% PV Wafer 97% 75% 83% Polysilicon 97% 99% 96% Japan PV Module 95% 90% 87% PV Cell 80% 86% 90% PV Wafer 95% 81% 74% Polysilicon 60% 59% 88% Taiwan PV Module 33% 16% 19% PV Cell 71% 80% 80% PV Wafer 71% 89% 91% Polysilicon 0% 0% 0% United PV Module 40% 72% 61% States PV Cell 84% 61% 69% PV Wafer 69% 84% 20% Polysilicon 81% 77% 74% 0K 1K 2K 3K 0K 1K 2K 3K 0K 1K 2K 3K

$ million $ million $ million

PV Module PV Cell PV Wafer Polysilicon 90% 80% 78% 66% 73% 76% 85% 68% 60% 62% 73% 58% 54% ati on 55% 40% Utili z ob al C apacit y

G l 20% 0% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Bars show manufacturing capacity (lighter shading) and utilized manufacturing capacity (i.e., production, darker shading) in US$2014 for key PV module economies. Vertical lines and associated numerical values show capacity utilization (production as a % of manufacturing capacity). Trend lines show global capacity utilization percentage for 2014–2016 (bottom). Note that China is displayed on a different scale.

34 Market Trends Manufacturing c-Si PV Modules

The benchmarked economies generally exhibited excess manufacturing capacity relative to global demand across the supply chain, indicating some potential to expand production in existing facilities to meet future demand growth or that new investment is required to modernize manufacturing processes to accommodate new technologies. In aggregate, global manufacturing capacity increased for all PV module supply chain intermediates in anticipation of increased demand. Globally, capacity utilization increased for all PV module supply chain intermediates (cells, wafers and polysilicon), and declined for PV modules over the period.

Global manufacturing capacity for PV modules was estimated at 80.3 GW in 2014, 107 GW in 2015 and 129 GW in 2016. Corresponding production was estimated at 46.3 GW in 2014, 59.3 GW in 2015, and 70.3 GW in 2016, reflecting a global excess of manufacturing capacity. All benchmarked PV supply chain intermediates had excess manufacturing capacity and reduced manufacturing costs over the period, which contributed to reduced prices in all four intermediates (e.g., PV module prices dropped 20%–25% over the period). China grew its manufacturing capacity and production for all PV module supply chain intermediates over the period, maintaining the highest capacity for each across the 13 economies. In 2016, both Malaysia and South Korea increased manufacturing capacity and production of PV modules and cells. Globally, capacity utilization was generally higher for the intermediates than the PV module end product.

35 Manufacturing c-Si PV Modules Trade Trends

PV Module Supply Chain, 2014-2016 Market and Trade Trends

PBenchmarkV 1: 2014-2016 PV PData:V 2.. 2014 - 2016 Trade PV PTrendsV 3: 2014 - 2016 PV PV 4: 2014 PV Trade App/PV2: 2014 - 2016 App/PV 3: 2014-2016 App/PV 4: 2014-2015 Production and Production&Demand Demand, Production, Bar and Trend Chart Detailed PV Module Detailed PV Cell BOT Detailed Polysilicon for Demand (%) contribu.. (MW) Mfg Capacity (update.. BOT for Key Countries for key countries key countries PV module supply chain trade, 2014–2016

2014 2015 2016 Brazil -$15M -$42M -$250M Canada -$49M -$15M -$36M China $7,472M $10,316M $7,381M Denmark -$4M -$112M -$43M Germany -$286M $18M -$533M India -$22M -$186M -$322M Japan -$6,070M -$4,194M -$3,034M Malaysia $1,622M -$65M -$469M

PV Module Mexico $513M $912M -$819M South Korea $1,599M $550M $1,335M Taiwan $1,006M $537M -$298M United Kingdom -$1,150M -$1,103M -$460M United States -$3,822M -$5,712M -$7,855M Brazil $0M -$1M -$13M Canada -$289M -$255M -$150M China $1,849M $1,709M $2,962M Denmark -$2M -$14M -$5M Germany $81M $237M $648M India -$466M -$1,433M -$2,405M Japan -$233M $108M $660M

PV Cell Malaysia -$122M $1,038M $831M Mexico -$346M -$718M -$618M South Korea -$425M -$752M -$953M Taiwan $3,929M $3,259M $2,476M United Kingdom -$694M -$744M -$297M United States $17M $32M -$63M Brazil $0M $0M $0M Canada $3M $2M $1M China -$2,019M -$1,898M -$2,136M Denmark -$5M $14M $11M Germany $936M $1,015M $947M India -$92M -$88M -$39M Japan -$932M -$730M -$945M Malaysia -$38M -$14M $26M

Polysilicon Mexico $0M -$1M $0M South Korea $845M $918M $1,021M Taiwan -$609M -$270M $1,684M United Kingdom -$143M -$105M $7M United States $1,780M $1,201M $1,619M

Imports Exports Imports Exports Imports Exports

South United United Brazil Canada China Denmark Germany India Japan Malaysia Mexico Korea Taiwan Kingdom States $10,000M T

B O $0M PV Module

T $2,000M

B O $0M PV Cell

$2,000M

T $0M B O

Polysilicon -$2,000M 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Bar chart (top) shows imports (negative values), exports (positive values) and balance of trade (exports less imports) in US$(2014) for PV module supply chain intermediates. Line chart (bottom) shows balance of trade trends for PV module supply chain intermediates. Note that trend lines are shown on different scales for PV modules, PV cells, and polysilicon.

36 Trade Trends Manufacturing c-Si PV Modules

In the benchmarked economies, the balance of trade (exports minus imports) remained relatively stable across the PV module supply chain, with a few exceptions including PV modules in Japan and the United States, PV cells in China and Taiwan, and polysilicon in Taiwan. Some net importers of end products, such as the United States, were also major exporters of upstream processed materials and/or subcomponents for the same technologies, illustrating the complexity of clean technology manufacturing and trade.

In 2016, the 13 economies exported $29.8 billion in PV modules, cells, and polysilicon, down from $31.5 billion in 2014. Imports of these products also dropped from $27.6 billion in 2014 to $26.7 billion in 2016. The rapidly declining cost of PV supply chain components over the period contributed to this contraction in trade.

China maintained its position as the largest PV module and cell exporter—primarily to the United States, Japan, and India—with 2016 decreases in exports driven in part by a buildup of manufacturer and installer inventory levels. South Korea and Malaysia emerged as growing exporters of PV modules and cells, possibly as a result of Chinese manufacturers exporting out of these countries to avoid import tariffs levied by the European Union and the United States (BNEF 2017c).

The United States and Japan were the two largest net importers of PV modules over the period. U.S. PV module net imports doubled, while Japan’s net imports decreased by 50%. In 2016, the United States was the second largest exporter of poly-Si (after Taiwan), maintaining a relatively steady value of polysilicon exports despite a dip in 2015. China continued to be the largest net importer of polysilicon.

37 Manufacturing c-Si PV Modules Trade Trends 10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report PV Module Trade Flow10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report PV Module Trade Flow10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report PV Module Trade Flow

 (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis) (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis) (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

(/index.html) PV module supply chain trade(/index.html) flows, 2014–2016 (/index.html) Global summary of PV module trade (alone), in 2014 USD Global summary of PV module trade (alone), in 2014 USD Global summary of PV module trade (alone), in 2014 USD

Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html) Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html) Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html) 2014 2015 2016 PV2014 modules2015 2016 2014 2015 2016 Imports Imports Imports Brazil Exports Brazil Exports $122.7 M Brazil Exports $437.8 M $14.8 M $56 K Canada $42.2 M $181 K Canada $249.8 M $148 K Canada $268.6 M $158.4 M $220.8 M $69.7 M $78.8 M $74.7 M $103.8 M

$20.7 M $94.1 M $110.9 M

United States of America United States of America

$4.1 B $5.8 B United States of America

China China China $138.0 M

$123.8 M $8.0 B

United Kingdom $1.3 B United Kingdom $1.2 B $63.6 M

$7.9 B United Kingdom $7 $523.1 M

.5 B B .6 1 $ $1.0 B $10.5 B

$553.5 M

M .5 3 8 7 $

Taiwan

Germany Germany

Taiwan

B .2 1 $

$16.0 M

M .3 8 5 3 $

$25.6 M

Germany

M .2 0 5 2 $

Taiwan

M .6 0 6 $

B .3 1 $

B .7 1 $

B .2 1 $

B .1 1 $

M .7 7 5 $

M .9 9 1 $ Denmark

M .4 1 5 5 $

Denmark

Malaysia Malaysia

M .1 5 1 $

M .3 7 2 1 $

M .9 4 1 $

M .3 9 6 1 $

M .4 8 3 $

Denmark

10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report PV Cell Trade Flow 10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report PV Cell Trade Flow 10/5/2020 Malaysia Joint Institute for Strategic Energy Analysis: Benchmark Report PV Cell Trade Flow

M .9 2 8 $

M .6 5 1 $

 (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis) (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis) (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

M .4 4 2 4 $

M .3 6 2 5 $

M .8 3 2 9 $

M .6 8 9 $

M .4 8 4 8 $

M .9 7 8 2 $

India

India

Mexico M .7 2 0 1 $

M .5 0 2 1 $

M .2 3 1 $ Mexico

M .1 2 1 $

India M .5 1 0 1 $

Mexico

M .5 9 2 $

B .7 1 $

B .7 3 $ M .6 0 0 9 $ B .7 6 $

B .7 1

2014 2015 2016 $

B .8 4 $

M .6 0 5 3 $

M .5 9 4 1 $

Japan

South Korea South Korea South

M .3 4 3 6 $ M .7 0 9 3 $ Japan M .7 9 1 6 $

Japan

South Korea South

M .2 9 5 6 (/index.html) $ (/index.html) (/index.html)

Global summary of PV cell trade (alone), in 2014 USD Global summary of PV cell trade (alone), in 2014 USD Global summary of PV cell trade (alone), in 2014 USD

Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html) Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html) Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html) Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). 2014 2015 2016 PV2014 cells2015 2016 2014 2015 2016 Imports Imports Imports https://www.jisea.org/benchmark/pv-module.html Brazil https://www.jisea.orExportsg/benchmark/pv-module.html1/2 Brazil https://www.jisea.orExportsg/benchmark/pv-module.html1/2 Brazil Exports 1/2 $159 K $8 K Canada $1.0 M $0 K Canada $12.6 M $0 K Canada $296.7 M $261.7 M $1.3 B $150.6 M $1.2 B $114.3 M $2.6 B $7.6 M $98.1 M $7.2 M $69.6 M $333 K United States$97.5 M of America United States$66.6 M of America United $States132.1 M of America

$7.5 M $4.4 M $23.6 M

China United Kingdom United Kingdom China United Kingdom $320.6 M $698.3 M China $751.9 M $3.4 B

$2.6 B

$4.0 B

$ 6 0 3 .6 M $ 3.0 B Taiwan

$4.2 B

M .1 1 9 5 Taiwan $

$4.4 B Taiwan

Germany

M .3 2 0 8 $ Germany

$106.2 M

$94.0 M Germany

B .1 1 $ 8 4 0 .7 M $

B .2 1 $

B .2 1 $125.0 M $

M .8 9 3 5 $

M .6 3 8 8 $

M .7 5 1 $

Malaysia

Denmark

M .7 1 $

M .1 7 $ Malaysia

Denmark

M .7 8 4 2 $

M .0 2 2 1 $ Malaysia

M .2 2

10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report PV Polysilicon Trade Flow$ 10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report PV Polysilicon Trade Flow10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report PV Polysilicon Trade Flow M .7 1 $

Denmark

K 4

   $       M .9 1 3 (mailto:[email protected]) (https://twitter.com/jisea1) (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis) (mailto:[email protected]) (https://twitter.com/jisea1) (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis) (mailto:[email protected])$ (https://twitter.com/jisea1) (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis) B .4 1 $

M .0 8 0 5 $

M .7 9 2 $

B .4 2 $

M .5 6 1 $

Mexico

India

M .0 5 9 4 $ Mexico

M .9 9 4 6 $ M .5 5 7 3 $

M .9 2 $ Mexico

India India

M .7 8 2 $

M .5 1 5 $

M .1 8 1 7 $

M .9 3 7 1 $

M .4 7 2 6

2014 2015 $ 2016 M .8 9 4 1 $

M .6 4 2 $

B .0 1 $ B .0 1 $

M .5 8 9 5 $

M .7 7 1 3 $

M .5 7 7 9 $ M .1 5 3 7 $ M .9 3 7 7 $

South Korea South Japan

South Korea South

M .6 1 0 9 $

Japan Japan (/index.html) (/index.html) Korea South (/index.html)

Global summary of PV polysilicon trade, in 2014 USD Global summary of PV polysilicon trade, in 2014 USD Global summary of PV polysilicon trade, in 2014 USD

Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html) Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html) Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html) Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). 2014 2015 2016 Polysilicon2014 2015 2016 2014 2015 2016 Imports Imports Imports https://www.jisea.org/benchmark/pv-cell.html https://www.jisea.org/benchmark/pv-cell.html1/2 https://www.jisea.org/benchmark/pv-cell.html1/2 1/2 Brazil Canada $2.2 B Exports Brazil Canada Exports Brazil Canada Exports $1.9 B $2.1 B $1.7 B $2.3 B $1.3 B $50 K $130 K $228 K $17 K $0 K $307 K $40 K $0 K $210 K $5.0 M $1.9 M $968 K

United States of America United States of America United States of America China $77.6 M China China

$76.4 M $77.7 M $54.2 M

$54.8 M $36.8 M United Kingdom $158.8 M

$30.2 M

$603.4 M United Kingdom United Kingdom $294.2

$198.0 M

M .9 3 5 3 $ $224 $139

M $599.3 M .4 M .6 M

$479.9 M

M .6 6 1 1 $

M .0 2 9 $

Taiwan

Germany Germany

Taiwan Germany

Taiwan

$873.0 M

B .0 1 $

$1.1 B $898.1 M

M .6 4 0 1 $

B .1 1 $

B .4 1 $

M .4 5 $ M .5 0 4 1 $

M .8 0 9 $

Denmark

M .6 8 $ M .6 1 1 $

Denmark

Malaysia

Denmark M .0 6 1 $

M .3 8 1 1 $

M .1 3 2 $

Malaysia Malaysia

M .5 0 3 $

M .7 8 3 $ M .4 8 2 1 $

M .7 4 1 1 $

K 4 6 4 $

K 0 7 4 $ M .5 8 8 $ M .2 2 9 $

M .2 1 $ India

K 0 $ M .4 1 $

Mexico K 6 $

M .2 1 $ India India

Mexico

K 0 3 7 $ K 3 9 $ Mexico

B .3 1 $

B .2 1 $

B .4 1 $ B .3 1 $

B .1 1 $ B .4 1

2014 $ 2015 2016

Japan

Japan

Japan South Korea South

South Korea South

South Korea South

M .2 1 9 3 $

M .6 1 9 4 $ M .2 8 1 4 $

M .7 1 8 3 $ M .1 2 6 2 $ M .2 3 3 3 $

Back to the BenchmarksGray of Global chords Clean Energy Manufacturingshow (/benchmark.html) trade flows. for PV moduleBack to thesupply Benchmarks of Globalchain Clean Energy intermediates Manufacturing (/benchmark.html). in US$(2014) amongBack to the Benchmarksbenchmarked of Global Clean Energy Manufacturing economies. (/benchmark.html)Width . of bars are scaled to total flows, with lighter shading showing imports and darker shading showing exports.Note that gray chords only capture the https://www.jisea.org/benchmark/pv-polysilicon.htmltrade among the 13 benchmarked economies;https://www trade.jisea.org/benchmark/pv-polysilicon.html flows1/2 with the rest of the world are not shown.https://www.jisea.or g/benchmark/pv-polysilicon.html1/2 1/2

38 Trade Trends Manufacturing c-Si PV Modules

A dynamic trade network connects the economies that $2.6 billion. During the period, the United States, South manufacture PV modules, cells, and polysilicon. Trade Korea, and Germany began exporting polysilicon to flow data suggest robust global trade among the Taiwan to avoid Chinese tariffs. Despite its lack of any economies that generally followed expected patterns production capacity, Taiwan then exported poly-Si given the distribution of manufacturing capacity produced by other economies to China. The U.S. BOT across economies. Trade flow for PV modules, wafers, 2015 dip was driven by decreases in exports from and polysilicon was impacted by changes to policies $1.9 billion to $1.3 billion, as a result of reductions in regarding subsidies, tax credits, tariffs, and grid polysilicon production from $940 million to $620 integration of renewables. million, possibly due to oversupply and a related 32% drop in prices. Between 2014 and 2016, the largest changes in PV module trade flow were observed in China, Japan, and the United States. China’s balance of trade (BOT) 2015 peak was driven by an increase in exports from $7.9 million in 2014 to $10.5 million in 2015. Potential drivers of the 2015–2016 drop in exports to $7.5 million include buildup of inventory in 2015 and expiration of a PV subsidy in 2016. Japan’s 2014–2016 BOT increase was driven by decreases in imports from $6.7 billion to $3.7 billion, triggered by reduction in demand from $6.8 billion to $4.1 billion. Potential reasons for the drop include a decline in solar project commissioning after mid-2016 revisions to feed-in tariff laws, along with the early-2015 introduction of new curtailment rules that dampened demand. The U.S. BOT decrease was driven by increases in imports to meet growing demand, as well as a rush to beat solar investment tax credit (ITC) expiration.

The biggest changes in BOT for PV cells were observed in China, India, and Taiwan. In 2016, China’s PV cell exports rebounded from a 2015 drop, due to decreases in imports from $2.6 billion to $1.2 billion and triggered in part by the 2016 expiration of its PV subsidy. Taiwan’s BOT decrease was driven by decreases in exports from $4.0 billion to $2.6 billion, potentially due to manufacturers’ pausing production based on oversupply and low cell prices.

Between 2014 and 2016, the largest changes in polysilicon trade flow were observed in Taiwan, Japan, and the United States. Potential drivers include the higher quality of polysilicon imported from other countries and the failure of Chinese polysilicon production to keep up with increasing domestic demand. Taiwan’s BOT increase was driven by increases in exports of polysilicon from $470 million to

39 Manufacturing c-Si PV Modules $ Value Added Trends

PV Value Added Story Benchmark Data: Value Added Trends PV 6. Value Added bar PV 7. Total Value FYI 2014-2016 PV FYI - PV Intercountry FYI calculations for PV total (iVA, dVA) Added Impacts of supply chain dva, dO Indirect Value Added, 6 discussion PV module supply chain totalGlob avaluel Production (added%).. and v(tVA)a retained by valuePV Suppl yadded Chain, All C.. component, 2014–2016

China Japan South Korea United States Malaysia Taiwan Germany

15,19115,355 4,269 4,315 15.0K 13,453 4.0K 3,605 e

u l 10.0K 2,465 2,492 2,492 2,509 2,536 2,113 Mo d V $ million $ million 2.0K P 1,608 1,625 1,527 1,628 1,646 1,440 1,398 1,413 1,327 5.0K

0.0K 0.0K

6,415 2,143 6,144 1,993 2,009 6.0K 5,572 2.0K 1,925 1,918 1,888

l 1,330 4.0K Ce l 1,119

V 1,008 P 909 923 $ million $ million 1.0K 848 590 628 2.0K 514 569 540 429

0.0K 0.0K 6.0K 5,354 1,497 5,143 1.5K 1,349 4,682 1,285

4.0K 1,004 958 991 er 1.0K

Wa f 728

V 630 659 $ million $ million P 591 562 552 2.0K 459 472 0.5K 397 250 227 231

0.0K 0.0K 3.0K 2,628 1.0K 958 2,264 843 772 2,074 735 697 2.0K 669 640 583 587

y silic on 0.5K $ million $ million o l P 1.0K 203 211 231 93 103 0.0K 0.0K 25 12 11 14 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Total value added (tVA) for the period in US$(2014) across PV module supply chain for key PV economies. Darker shading indicates direct value added (dVA), lighter shading indicates indirect value added (iVA), and numerical values show the total value added (tVA). Note that tVA data for China are displayed on a different scale.

40 Value Added Trends $ Manufacturing c-Si PV Modules

Total value added (tVA) to China’s economy from that of any other benchmarked economy, the other production of PV modules, cells, and polysilicon was economies generally had higher VA retained, led by much greater over the period than that of any other the United States, Germany and Japan (37% - 65% benchmarked economy. Japan’s tVA was second across the supply chain). VA retained reflects the highest tVA, but 10% to 30% that of China’s tVA across extent of domestic supply chains as well as prevailing the supply chain. tVA from global production of PV wages, domestic profits, and taxes less subsidies. modules generally declined over the period, while tVA increased slightly or remained flat for PV cells, PV wafers, and polysilicon. Indirect value added (iVA) was generally greater than direct value added (dVA) for the key PV module economies.

Over the period, tVA from global production of PV modules generally decreased—in part due to sharply declining prices that reduced production revenues— while tVA increased slightly or remained flat for PV cells, wafers, and polysilicon. Between 2014 and 2016, tVA from PV modules declined or remained flat for all benchmarked economies. Between 2014 and 2016, tVA from polysilicon generally increased, with the exception of the United States, where a downward production trend was triggered by declining prices and Chinese tariffs on U.S. polysilicon imports. tVA is generally highest in economies with the highest levels of production revenue. In 2016, the 13 benchmarked economies produced $57.6 billion in tVA from the production of c-Si PV modules, cells, wafers, and polysilicon, down 3.4% from $59.6 billion in 2014. $14.5 billion of the 2016 aggregate tVA was dVA from domestic manufacturers. iVA was higher than dVA for all benchmarked economies across the supply chain, with the exception of polysilicon production in South Korea, Malaysia, and the United States, demonstrating, in general, that the supply chains that support PV module manufacturing domestically and internationally were more important to the benchmarked economies than the domestic manufacturing of the intermediates (PV wafers, PV cells and polysilicon) and end product (PV modules).

Benchmarked economies retained varying shares of tVA as a portion of manufacturing revenue (i.e., VA retained) over the period from manufacturing PV modules and intermediates. While the value added to China’s economy from manufacturing across the PV module supply chain was much greater than

41 Manufacturing c-Si PV Modules $ Value Added Trends

PV module supply chain total value added (tVA), domestic and non-domestic contribution, 2014–2016

China Germany Japan Malaysia South Korea Taiwan United States Country 100% Brazil Canada VA

a l China o t e

u l Denmark Germany

Mo d 50%

V India P Japan Malaysia Co nt r ib u ti o n t T Mexico 0% South Korea 100% Taiwan

VA United Kingdom a l United States o t l Ce l

V 50% P Co nt r ib u ti o n t T 0% 100% VA a l o t er

Wa f 50% V P Co nt r ib u ti o n t T 0% 100% VA a l o t

y silic on 50% o l P Co nt r ib u ti o n t T 0% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

For each economy (listed across the top) color-coded bars show the share of tVA accrued from domestic and non-domestic production of PV module intermediates for 2014 to 2016. Domestic bars (generally the largest in each bar) represent the share of tVA (iVA plus dVA) from domestic production. Non-domestic bars represent the share of tVA (iVA only) from production in other economies (dVA only occurs in the economy where production occurs).

While all benchmarked economies received iVA part by growth in manufacturing capacity, impacted all from production of PV modules, cells, wafers, and benchmarked economies over the period and generally polysilicon in other economies over the period, most contributed more iVA than the other economies. of the tVA came from domestic manufacturing of those intermediates. Production in China, driven in

42 Manufacturing c-Si PV Modules

PV Challenges and Opportunities: Balance of System Components

Modeled cost breakdown trends for PV systems installed in the United States (inflation adjusted), 2010–2017

Utility-Scale PV, Utility-Scale PV, Residential PV (5.7kW) Commercial PV (200 kW) Fixed Tilt (100 MW) One-Axis Tracker (100 MW) $8 7.24 Soft Costs - Others (PII, Land Acquisition, Sales Tax, Overhead, and Net Pro t) $7 Soft Costs - Install Labor 6.34 Hardware BOS - Structural and Electrical Components Inverter $6 Module 5.36 5.44 $5 4.97 4.48 4.57 4.59

$4 3.92 3.91 3.44 3.42 3.18 3.15 2.98 $3 2.80 2.78 2.76 2.66 2.39 US$(2017) per Watt DC Watt US$(2017) per 2.27 2.17 2.15 2.04 1.97 $2 1.85 1.89 1.82 1.45 1.54 1.03 1.11 $1

$0 2010 2011 2012 2013 2014 2015 2016 2017 2010 2011 2012 2013 2014 2015 2016 2017 2010 2011 2012 2013 2014 2015 2016 2017 2010 2011 2012 2013 2014 2015 2016 2017

Source: Fu et al. 2017

This benchmark report focuses on the impacts of the suppliers in each segment collectively provided manufacturing supply chains of selected clean energy more than 80% of the market in 2017, there were a technologies but does not currently address the large number of small racking suppliers. Foreign and manufacture and trade of the other equipment required domestic racking companies supplied to the U.S. market, to actually use the technology. For example, the report with some foreign parent companies operating U.S. does not consider the impacts of the supply chains subsidiaries with local manufacturing. As shown in the of balance of system (BOS) components—inverters, figure, 2017 hardware BOS structural and electrical and structural (e.g., racking and frames) and electrical component costs ranged from $0.20/W for utility-scale hardware—required for PV module system installation (fixed-axis) to $0.36/W for residential PV systems, a and operation. large reduction from 2010 values of $0.43/W for utility- scale (fixed-axis) and $0.56/W for residential. PV installations in the United States have relied on BOS equipment from both domestic and foreign sources. U.S. Steel makes up the largest amount of raw material used inverter manufacturing capacity was at a five-year low in in PV racking. Due to concerns about tariffs with some 2017, with companies consolidating their global supply Asian countries, many providers used North American chains to reduce costs and China and Europe supplying steel (GTM Research 2017). The supply of other BOS most PV inverters for the U.S. market. Still, approximately equipment, such as wiring and combiner boxes, was also 30% of all inverter capacity installed in the United States highly fragmented, with U.S. manufacturing companies in 2017 was domestically sourced. sourcing material (e.g., aluminum and steel) from a global marketplace. The U.S. PV racking market has been extremely fragmented by market segment. While the top five

43 Manufacturing LED Packages

LED Packages The light-emitting diode (LED)—a solid-state semiconductor device that emits light when an electric current is passed through it—is one of today’s most energy-efficient and rapidly-developing lighting technologies, according to the U.S. Department of Energy.19 Between 2014 and 2016, global production of LED packages and chips increased by 37.8% from 162.0 billion to 223.4 billion packages annually. Between 2015 and 2016, the market entered a period of oversupply, with price pressure driving average selling prices down by 30%–40%. The global market for LED packages used in luminaires, one specific application, expanded from $7.9 billion in 2014 to $9.1 billion in 2016. Global revenue from LED lighting systems for all applications is expected to total $216 billion by 2024 (Navigant Research 2015).

Light-emitting diode (LED) packages and supply chain

Light-emitting diode package alignment with the Clean Energy Manufacturing Analysis Center (CEMAC) benchmark framework. Boxes highlight components included in the benchmark analysis. No raw materials were included in the analysis because of a lack of data. Illustration by Josh Bauer, NREL

Researchers focus on the LED package as the end product of the supply chain framework, even though these act as components in a variety of other products, including automobiles, personal electronics, displays, and lighting. This simplified version of the manufacturing supply chain includes the sapphire substrate, LED chips, and LED packages.

Multiple opportunities exist for innovations that could lead to improvements in LED product efficacy, quality, and/ or price. These include new chip and package designs; improvements in package substrate, encapsulant, optic, and phosphor materials; as well as novel processing techniques such as wafer bonding, substrate removal, and wafer-level processing (Navigant Consulting 2014).

19 For more information, see the DOE “LED Lighting” web page at https://www.energy.gov/energysaver/save-electricity-and-fuel/lighting-choices-save-you-money/ led-lighting.

44 Manufacturing LED Packages

Notable Trends imports from Taiwan dropped, and a global excess of manufacturing capacity was established. China Key drivers of LED package supply chain trends was also the largest net importer of LED chips and include: packages.

• Demand growth related to regulated phaseouts of Quality and Cost incandescent lighting LEDs are used in a wide range of products. Each • Declining prices resulting from oversupply and application has a unique set of specifications defining industry consolidation. color temperature, lumens (brightness), and voltage requirements. Due to differences in LED specifications Industry Consolidation for various applications, LED quality and performance Historically, regulated phaseouts of incandescent varies. In 2016, Chinese manufacturers commanded lighting to meet domestic energy efficiency targets in the greatest share of the lower-cost, low-lumens LEDs the benchmarked economies, as well as and declining used for television and personal lighting, while U.S. LED costs, have driven demand growth for LEDs used and European companies retained market share for in lighting. Global demand for light-emitting diode brighter, higher-lumen equipment that yielded larger (LED) packages, chips, and sapphire substrate grew profit margins (Bradsher 2014). rapidly between 2014 and 2016, led by China. Demand More detailed information can be found on the in other individual benchmarked economies was following pages. relatively flat. Even with increased demand, oversupply (due in part due to China’s large and increasing manufacturing capacity) combined with industry consolidation drove down the price of LEDs.

Global Supply Chains Across the LED supply chain for all benchmarked economies, indirect value added (iVA) from manufacturing was greater than direct value added (dVA), indicating that participation in the broader supply chain that supports LED manufacturing at home and abroad was more important to the benchmarked economies than domestic production of the intermediates end product. Production of LED packages and chips in Japan and China contributed the largest share of non-domestic iVA to the benchmarked economies. Only Taiwan had sufficient domestic production to meet domestic demand across the supply chain. Other economies relied on trade to fill the supply chain gaps.

China and Japan China’s concerted efforts to expand production of LED packages and chips contributed to its displacement of Japan as the top global producer over the period. As new Chinese foundries came online, Chinese

45 Manufacturing LED Packages Market Trends

LED Package Supply Chain, 2014-2016 Market and Trade Trends Benchmark Data: Market Trends

LED 1. 2014-2016 LED LED 2. 2014-2016 LED LED 3. 2014 - 2016 LED LED-4. 2014-2016 LED App/LED 2: 2014-2016 FYI Calc for LED 2 and FYI Calc for LED 4 Production and Production/Demand in Demand, Production, Package LED Detailed LED3 discussion LEDDema packagend (%) contribu.. supplymillion packages/yr chain demandMfg Capacity (uandpdate.. productionImport/Export/BOT Tshares,.. Imports 2014–2016/Exports

LED Package LED Chip Sapphire Substrate Brazil 100% Canada China 11% 11% Denmark 20% 20% Germany 30% 30% 80% India

45% 45% Japan 29% 29% 54% Malaysia 26% Due to a lack of availability 26% Mexico 60% of economy-specific 22% 22% South Korea

nd Sh are s demand data for LED Taiwan 15% packages, demand for LED 15% 14% 14% 14%

14% United Kingdom De m a packages is not included in 14%

40% 14% United States this chart. 12% Rest of World 15% 15% 27% 27% 24% 24% 14% 21% 20% 21% 22% 22% 16% 13% 13% 13% 11% 13% 11% 0% 100% 11% 20% 22% 29% 29% 80% 30% 38% 44% 29% 53% 19% 26% 17% 60% 18% 22% on Sh are s 15% t i 15% 17% 15% 14% 15% odu c 16% P r 40% 14% 12% 17% 14% 27% 14% 24% 37% 21% 35% 20% 32% 23% 19% 15% 13% 13% 11% 6% 0% 5% 4% 4% 2014 2015 2016 2014 2015 2016 2014 2015 2016

Breakdown (in %) of global demand (top) and global production (bottom) by economy for benchmarked LED package supply chain intermediates. Due to a lack of availability of economy-specific demand data for LED packages, demand for LED packages is not included in this chart.

China grew its shares of global demand and LED package production in Japan and South Korea production of LED packages and chips over the period, constituted more than 55% of total global package accounting for the highest share of each among the output in 2014, with China and Taiwan together economies in 2016. Taiwan accounted for the highest contributing another 24%. By 2016, China was the shares of global production of sapphire substrate. highest-volume package producer with a 30% share, Shifts in shares over the period were driven by China’s with Japan, South Korea, and Taiwan together concerted efforts to expand production of LED contributing 54%. packages and chips.

46 Market Trends Manufacturing LED Packages

LED Package Supply Chain, 2014-2016 Market and Trade Trends

LED 1. 2014-2016 LED LED 2. 2014-2016 LED LED 3. 2014 - 2016 LED LED-4. 2014-2016 LED App/LED 2: 2014-2016 FYI Calc for LED 2 and FYI Calc for LED 4 LEDPro dpackageuction and supplyProduction /chainDemand in demandDemand, Productio andn, productionPackage for keyLED De tLEDailed packageLED3 economies,discussion 2014–2016Demand (%) contribu.. million packages/yr Mfg Capacity (update.. Import/Export/BOT T.. Imports/Exports

China Japan Malaysia South Korea Taiwan United States

120K

100K /yr 80K

kage Due to a lack of availability of economy-speci c demand data for LED packages, demand for LED packages is not included in this chart. 60K LED Pa c

M illi o n p ackage s 40K

20K

0K

120K

100K /yr 80K

60K LED Chip

M illi o n p ackage s 40K

20K

0K

120K

100K e /yr 80K Su bst ra t

re 60K M illi o n p ackage s Sa pp h i 40K

20K

0K 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Demand (color-coded lines) and production (gray bars) trends (in millions of packages) for LED package supply chain intermediates by economy. Economies are listed in order of 2016 production levels. Due to a lack of availability of economy-specific demand data for LED packages, demand for LED packages is not included in this chart.

Global demand for and production of LED packages, Between 2014 and 2016, annual global production and chips, and sapphire substrate grew significantly over demand for LED packages increased by 37.8% from the period, led by China. Production and demand 162.0 billion to 223.4 billion packages. Over the period, in other individual benchmarked economies was the most significant increases in LED package and chip relatively flat. Only Taiwan had sufficient domestic production were observed in China, with many new production to meet domestic demand across the foundries coming online in that country in 2015. supply chain.

47 LED PackageManufacturing Supply LED Chain, Packages 2014-2016 MarketMarket and Trends Trade Trends

LED 1. 2014-2016 LED LED 2. 2014-2016 LED LED 3. 2014 - 2016 LED LED-4. 2014-2016 LED App/LED 2: 2014-2016 FYI Calc for LED 2 and FYI Calc for LED 4 Production and Production/Demand in Demand, Production, Package LED Detailed LED3 discussion Demand (%) contribu.. million packages/yr Mfg Capacity (update.. Import/Export/BOT T.. Imports/Exports LED package supply chain manufacturing capacity utilization, 2014–2016

2014 2015 2016

Capacity Utilization Capacity Utilization Capacity Utilization 0% 50% 100% 0% 50% 100% 0% 50% 100% China LED Package 100% 100% 100% LED Chip 60% 58% 70% Sapphire Substrate 74% 8% 9% Japan LED Package 100% 100% 100% LED Chip 90% 79% 80% Sapphire Substrate 74% 76% 63% South Korea LED Package 100% 100% 100% LED Chip 71% 61% 78% Sapphire Substrate 74% 30% 35% Malaysia LED Package 100% 100% 100%

LED Chip 0K 0K 0K

Sapphire Substrate 0K 0K 0K Taiwan LED Package 100% 100% 100% LED Chip 83% 66% 77% Sapphire Substrate 74% 45% 44% United LED Package 100% 100% 100% States LED Chip 94% 86% 80% Sapphire Substrate 74% 7% 3% 0K 2K 4K 6K 0K 2K 4K 6K 0K 2K 4K 6K $ million $ million $ million

LED Package LED Chip Sapphire Substrate 100% 100% 100% 100% 74% 72% 74% 64% 50%

21% 17% 0% ob al C apacit y Utili z ati on

G l 2014 2015 2016 2014 2015 2016 2014 2015 2016

Bars show manufacturing capacity (lighter shading) and utilized manufacturing capacity (i.e., production, darker shading) in US$2014 for LED package economies. Vertical lines and associated numerical values show capacity utilization (production as a % of manufacturing capacity). Trend lines show global capacity utilization percentage for 2014–2016 (bottom); because of a lack of economy-specific manufacturing capacity data for LED packages, manufacturing capacity of LED packages is assumed equal to production.

China’s large and increasing manufacturing capacity Global manufacturing capacity for LED chips was contributed to an excess of manufacturing capacity estimated at $10.4 billion in 2014, $10.4 billion in 2015 for LED chips and sapphire substrate relative to global and $11.5 billion in 2016. Corresponding production was demand, indicating some potential to meet future estimated at $7.4 billion in 2014, $6.7 billion in 2015, demand growth by expanding production. Globally, and $8.5 billion in 2016, reflecting a global excess of capacity utilization for LED chips was relatively flat manufacturing capacity. Globally, capacity utilization (64% to 74%) while capacity utilization for sapphire was much higher for LED chips than for sapphire substrate dropped from 74% to 17% over the period. substrate, as sapphire substrate manufacturing capacity grew much faster than production over the period.

48 Trade Trends Manufacturing LED Packages

LED Package Supply Chain, 2014-2016 Market and Trade Trends Benchmark Data: Trade Trends LED 1. 2014-2016 LED LED 2. 2014-2016 LED LED 3. 2014 - 2016 LED LED-4. 2014-2016 LED App/LED 2: 2014-2016 FYI Calc for LED 2 and FYI Calc for LED 4 Production and Production/Demand in Demand, Production, Package LED Detailed LED3 discussion LEDDema packagend (%) contribu.. trade,million p ackages/yr2014–2016Mfg Capacity (update.. Import/Export/BOT T.. Imports/Exports

2014 2015 2016 Brazil -$35M -$29M -$27M

Canada -$50M -$54M -$52M

China -$2,987M -$4,386M -$2,298M

Denmark -$25M -$23M -$11M

Germany $340M $205M $160M

India -$76M -$131M -$178M

Japan $3,110M $2,520M $2,299M

Malaysia $1,615M $1,749M $1,753M

Mexico -$366M -$468M -$451M

South Korea -$339M $334M -$46M

Taiwan $1,917M $3,153M $1,567M

United Kingdom -$144M -$146M -$132M

United States $397M $486M -$144M

Imports Exports Imports Exports Imports Exports India China Brazil Japan United Mexico Taiwan Canada Kingdom Malaysia Denmark Germany South Korea United States

$2,000M

$0M

-$2,000M Balance of Trade -$4,000M 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Bar chart (top) shows imports (negative values), exports (positive values) and balance of trade (exports less imports) in US$(2014) by economy for LED packages. Line chart (bottom) shows balance of trade trends for LED packages.

Between 2014 and 2016, China was the largest net decreased by 18.5% (from $17.1 billion to $14.0 billion) importer of LED packages, while Japan, Malaysia, and the value of global imports decreased 16.4% (from and Taiwan were the largest net exporters. In the $13.8 billion to $11.5 billion), reflecting lower prices and benchmarked economies, the balance of trade oversupply in the LED market. (exports minus imports) remained relatively stable, with a few exceptions including China, Taiwan, the Japan, Malaysia, and Taiwan had trade balances (BOT) United States, and South Korea. of $2.3 billion, $1.8 billion, and $1.6 billion, respectively for LED chips and packages in 2016, while China’s BOT Over the period, the value of global exports of LED was negative $2.3 billion in 2016. packages among the benchmarked economies

49 10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report LED Trade Flow

 (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

(/index.html)

Manufacturing LED Packages Trade Trends Global summary of LED trade flow, in 2014 USD

Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html)

2014 2015 2016

Imports Brazil Exports Canada LED package trade flows, 2014–2016 $35.7 M $198 K $73.9 M $7.1 B $1.8 B $23.7 M

While global exports and imports of LED packages United States of America $1.4 B declined over the period, the trade network among

China the benchmarked economies remained intact $56.9 M

United$2 0Kingdom0.5 M and generally followed expected flows, given the distribution of manufacturing capacity across $2.9 B

$4.1 B

economies. China’s expanding LED package and chip Taiwan M .5 0 9 7 $ manufacturing capacity contributed to trade shifts in Germany 10/5/2020 $1.0 B Joint Institute for Strategic Energy Analysis: Benchmark Report LED Trade Flow

 (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

B .1 1

benchmarked economies. $

B .1 2 $

M .0 9 2 $ Denmark

Between 2014 and 2016, the largest changes in trade Malaysia

M .4 4 $

M .4 7 7 4 $

flow of LED chips and packages were observed in (/index.html)

M .6 4 2 $

M .2 5 8

China, Japan, and Taiwan. Many new Chinese foundries Global summary of LED trade flow, in 2014 USD $

M .5 9 $ India

Mexico M .1 0 9 3 $

Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html)

M .9 4 7 9

came online in 2015, abruptly reducing imports from $ M .7 1 9 8 $

B .3 1 $

B .0 4 2014 2015$ 2016

Japan Taiwan and contributing to the sharp reduction in Korea South 2014 China’s net imports (from -$4.4 billion to -$2.3 billion) Imports Brazil Exports $29.4 M $164 K Canada $74.1 M and Taiwan’s net exports (from $3.2 billion to $1.6 $1.9 B $8.7 B Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). $20.2 M United States of America billion) from 2015 to 2016. New capacity in China also $1.4 B contributed to net exports reductions in Japan from $49.1 M https://www.jisea.org/benchmark/led.html China 1/2

United Kingdom $3.1 billion in 2014 to $2.3 billion in 2016. $194.7 M

$4.2 B Gray chords show LED package trade flows in US$(2014) among

$4.3 B

benchmarked economies. Width of bars are scaled to total Taiwan

flows, with lighter shading showing imports and darker shading

B .0 1 $

$1.0 B showing exports. Note that gray chords only capture the trade 10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report LED Trade Flow Germany

among the 13 benchmarked economies; trade flows with the rest of  (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

B .2 1 $

B .2 2

the world are not shown. $

M .7 7 2 $ Malaysia

Denmark

M .0 3 1 4 $

M .3 4 $

(/index.html)

M .0 8 1 $

M .0 2 4 1

Global summary of LED trade flow, in 2014 USD $

Mexico

M .3 6 8 4 $

M .8 0 1 $ India

B .6 1

Batteries (/benchmark/battery.html)$ LED (/benchmark/led.html) PV Wind (/benchmark/wind.html)

M .5 7 5 7 $

B .2 1 $ B .3 3 $ Japan

2015 Korea 2014South 2015 2016

Imports Brazil Exports $27.3 M $262 K Canada $1.3 B $69.9 M $5.4 B Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). $18.1 M United States of America $1.5 B

https://www.jisea.org/benchmark/led.html$46.0 M China 1/2

United Kingdom $177.8 M

$2.4 B

$3.1 B

Taiwan

B .0 1 $ Germany

$847.9 M

B .1 2 $

B .2 1 $

Malaysia M .4 1 1 $

Denmark

K 9 5 8 $

M .3 2 4 3 $

M .8 5 2 $

M .1 7 8 1 $

Mexico

M .7 8 $ India

M .8 6 7 4 $

M .9 0 0 8 $

M .5 2 9 5 $

M .3 7 4 8 $

B .9 2 $ Japan 2016 Korea South

50 Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html).

https://www.jisea.org/benchmark/led.html 1/2 Value Added Trends $ Manufacturing LED Packages

LEDBenchmark Value Added Data: Story Value Added Trends

LED6. 2014-2016 LED LED 7. Total Value FYI LED dVA, dO, VA FYI Calculations for LED package supply chainsupply chain totaldva, iva, valueAdded Imp addedacts of (tVA)retained by value LEDadded 6 component, 2014–2016 tva Global Production (%)..

Japan China South Korea Taiwan United States Malaysia 5,000 4,553 4,152 4,000 3,826

3,206 3,000 2,614 2,423 2,391 2,295 $ million

LED Package 2,000 1,863

1,410 1,360 1,304 1,090 1,047 1,075 1,000 869 821 844

0 3,000 2,813

2,500

2,000 1,810 1,866 1,594 1,472 1,500 1,349 $ million LED Chip 1,130 1,049 1,058 1,000 953 931 793 752 646 618 500

79 77 110 0

164 152 150 144 129 125 129

98 100 94 84 80 $ million 66 Sapphire Substrate 50 50

20 17 11 6 5 4 0 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Total value added (tVA) for the period in US$(2014) from manufacturing LED supply chain intermediates in key LED package economies. Darker shading indicates direct value added (dVA), lighter shading indicates indirect value added (iVA), and numerical values show total value added (tVA). Economies are listed in order of tVA from LED package production in 2016. Note that each intermediate is displayed on a different scale.

51 Manufacturing LED Packages $ Value Added Trends

Of the benchmarked economies, total value added Benchmarked economies retained varying shares of (tVA) from global production of LED packages, chips, tVA as a portion of manufacturing revenue (i.e., VA and sapphire substrate was greatest for Japan, China, retained) over the period from manufacturing LED South Korea, and Taiwan. Indirect value added (iVA) packages and intermediates. China netted the lowest was higher than direct value added (dVA) for all LED VA retained across the LED package supply chain supply chain intermediates considered, indicating that (15% - 21%).The United States netted the highest participation in the broader supply chain that supports VA retained for LED package and chip production global LED package manufacturing domestically (65%), despite relatively low production. For sapphire and internationally was more important to the substrate production, only Japan had a higher VA benchmarked economies than domestic production of retained (45%) than the United States (36%). VA these intermediates and end products. retained reflects the extent of domestic supply chains as well as prevailing wages, domestic profits, and taxes tVA from the production of LED packages in less subsidies. benchmarked economies was up slightly, from $13 billion in 2014 to $13.2 billion in 2016. In 2016, manufacturing LED packages delivered tVA of $3.8 billion for Japan, $3.2 billion for China, and $2.3 billion for South Korea. China’s tVA from LED package and chip manufacturing increased over the period, while tVA in the other economies for all supply chain intermediates remained relatively flat or declined. China’s tVA from LED chip manufacturing increased significantly over the period, resulting from a concerted effort to expand domestic manufacturing. tVA accrued to China, Japan, South Korea, and Taiwan from production of LED packages and chips over the period was each greater than that accrued to the United States. dVA from domestic production of LED packages declined or remained flat over the period for all economies except China, where dVA increased by 167% as a result of increases in domestic production revenues over the period. The greatest declines were observed in Japan (24.0%) and South Korea (19.3%). iVA from global LED package production generally followed the same trends over the period.

Both dVA and iVA for LED chip production increased in the five key LED chip manufacturing economies— with tVA up by 55.4% in China, 16.3% in the United States, 10.1% in South Korea, and 8.3% in Japan. tVA for Taiwan decreased by 6.4% over the period. From 2014 to 2016, both dVA and iVA from domestic production of sapphire substrate declined for all economies, reflecting production trends due to oversupply.

52 Value Added Trends $ Manufacturing LED Packages

LED Value Added Story

LED6. 2014-2016 LED LED 7. Total Value FYI LED dVA, dO, VA FYI Calculations for supply chain dva, iva, Added Impacts of retained LED 6 LED package supply chaintva total valueGloba l addedProduction (%) ..(tVA), domestic and non-domestic contribution, 2014–2016 Impact Country Country China Japan South Korea Taiwan United States Brazil Canada 100% China Denmark Germany 80%

VA India

a l Japan o t Malaysia 60% Mexico South Korea Taiwan LED Package United Kingdom 40% United States Co nt r ib u ti o n t T

20%

0% 100%

80% VA a l o t

60% LED Chip 40% Co nt r ib u ti o n t T

20%

0% 100%

80% VA a l o t

60%

Sapphire Substrate 40% Co nt r ib u ti o n t T

20%

0% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

For each economy (listed across the top) color-coded bars show the share of tVA accrued from domestic and non-domestic production of LED package intermediates for the top LED economies between 2014 and 2016. Domestic bars (generally the largest in each bar) represent the share of tVA (iVA plus dVA) from domestic production. Non-domestic bars represent the share of tVA (iVA) only from production in other economies (dVA only occurs in the economy where production occurs).

In general, the benchmarked economies received the production of LED packages and chips in China, Japan, greatest portion of tVA from domestic production of Malaysia and South Korea impacted all economies LED packages, chips, and sapphire substrate. However, considered through iVA.

53 Manufacturing LED Packages

LED Challenges and Opportunities: Quality and Cost

LED Applications, 2014–2016 Even within an application like general illumination, the specifications for each type of LED product varies Application 2014 2015 2016 significantly depending on the customer’s needs and General Lighting 51% 58% 63% budget; for example, color quality requires balancing spectral fidelity and gamut with well-documented LCD TVs and Monitors 15% 12% 10% preference. Beyond color quality, customers also frequently have requirements for lighting output, Cell Phones 11% 10% 9% reliability, and efficacy. Products frequently need to withstand temperature, humidity, ultraviolet radiation Notebooks and Tablets 9% 7% 6% exposure, and voltage variation varies depending on the application. The specifications that an LED need to meet Signs and Large Displays 6% 5% 5% are critical factors in determining its cost. A single LED in a commercial light has a much lower cost and much lower Automotive Lighting 4% 4% 4% specification than a single smaller LED in a smart phone that doubles as a high-lumen flashlight.

Other Displays 3% 3% 2% Application-driven differences are combined with quality differences among LED manufacturers. Although no Personal Lighting 1% 1% 1% analytical data are available on product quality across manufacturing regions, some conclusions can be drawn Percentage indicates the portion of LEDs used globally for each application (Mukish and Virey 2017). from anecdotal information. For example, consumer and media reviews might mention that one brand of LEDs is considered notably more long-lasting and of higher The high-level view of clean energy manufacturing quality, or that other LED brands are known for early supply chains provided in this benchmark report is based burnout. Where they can be verified at all, such claims on available market and trade data, which are typically can require careful laboratory tests. highly aggregated. Disaggregating these data to gain insight into the differences in quality of the products The trade-off between quality and cost can take time tracked is difficult due to data reporting limitations. The to manifest itself in the field. For example, the ceramic challenges of assessing the impacts of different types packaging used for high-power LEDs and polyphthalamide of LEDs manufactured around the world in terms of the packaging (Tuttle and McClear 2014) used in low power benchmark metrics provides one example. LEDs are not LEDs behave significantly differently with time and heat all created equal. LEDs are used in an impressive array exposure. The polyphthalamide will also cause color of products, from stadium scoreboards to televisions, shifting and other issues as the LED ages. household lightbulbs, and car headlamps. In 2016, more than 60% of LEDs were used for general lighting, Reporting on LEDs using the CEMAC benchmark as shown above. Within the lighting category, there is methodology makes highlighting differences in LED also variation in LEDs due to optimization of different quality and cost challenging. Benchmark results may parameters—saturation, preference, efficiency—for show, for instance, that an economy has only 5% of different uses. the market by physical volume, but this amount could represent 15% of revenue. The discrepancy between In the highly segmented LED industry, Chinese the price of the various types of LEDs means it is manufacturers command the greatest share of the lower difficult to ascertain quantitatively the physical volume lumen LEDs used for lighting. Western manufacturers of production and its impact on any specific regional retain market share for brighter, higher-lumen devices economy might be. that yield larger profit margins (Bradsher 2014).

54 Manufacturing LIB Cells

Lithium-Ion Battery Cells for Light-Duty Electric Vehicles The global market for automotive lithium-ion battery (LIB) cells grew rapidly as a direct result of the increasing demand for plug-in, hybrid and fully electric vehicles (EVs). Global annual demand for EV LIB cells increased threefold over the period, from 9.6 GWh in 2014 to 31.1 GWh in 2016.20 This demand was mainly driven by sales of EVs (plug-in, hybrid and fully electric vehicles), which constituted 2.6% of global light-duty vehicle (LDV) sales in 2016.21

Light-duty vehicle lithium-ion battery (LIB) cells and supply chain

Lithium-ion battery cell alignment with the Clean Energy Manufacturing Analysis Center (CEMAC) benchmark framework. Boxes highlight components included in the benchmark analysis. No raw materials were included in the analysis for lack of data. Illustration by Josh Bauer, NREL

LIB cells constitute a large portion of the cost structure for complete battery packs, and cell cost and performance drive overall pack cost and performance. This report focuses on cells, rather than packs, because car manufacturers typically design and assemble their own packs using purchased LIB cells.

Opportunities for innovation include advances in cell chemistries, formats, and manufacturing processes. Researchers focus on LIB cells used for LDVs and the intermediate materials required to make these cells, namely the cathode, anode, electrolyte, and separator.

Concerns about availability of some elements critical to LIB manufacturing (e.g., lithium and cobalt) that are largely sourced outside the major economies analyzed here, along with environmental concerns about disposal at the end of battery life, are spurring new research, policies, and regulations.

20 NREL estimate based on Curry (2017), BNEF (2017b), Zamorano (2017), Yano (2017), and Pillot (2017) 21 NREL estimate based on Richter (2017), EIA (2017;), and BNEF (2016)

55 Manufacturing LIB Cells

Notable Trends China, Japan, South Korea, United States, and Germany LIBs represent a strong growth industry for China, Japan, and the United States accrued more participating economies. Key drivers of LIB cell supply tVA from manufacturing LIB cells than the other chain trends include: benchmarked economies over the period, resulting from production to meet LIB demand for EVs. • Growing global demand for LIB cells in the Leveraging mature domestic supply chains originally manufacture of battery packs used in EVs developed to serve consumer electronics markets, • Development and expansion of domestic LIB cell China, Japan, and South Korea were the top-three manufacturing supply chains to meet all or part of producers of LIB cell intermediates, as well as the top domestic demand growth net exporters of LIB cells. While both the United States and Germany remained leaders in EV deployment, • China’s strong backing of its domestic LIB their domestic LIB supply chains could not meet industry through policy and investment to domestic EV demand, resulting in the two countries support deployment of EVs (Chung, Elgqvist, and becoming the largest net importers of LIB cells. Santhanagopalan 2016). Raw Materials in the Manufacturing Manufacturing Capacity Supply Chain Led by China, growth in demand for lithium-ion Some of the key raw materials used in manufacturing battery (LIB) cells was driven by the benchmarked LIBs include lithium, graphite, cobalt, and manganese. economies’ investment in electric vehicles (EVs), These materials are used to manufacture a range of including hybrid electric and fully electric vehicles, products, from LIBs for consumer electronics and often supported by subsidies. In anticipation vehicles to superalloys, hard metals, ceramics, and of continued increasing demand, global LIB cell polymers. As more EVs with LIBs are deployed, these manufacturing capacity soared in 2016, creating vehicles are driving demand for materials used in excess capacity across supply chain intermediates. If LIB cells. Understanding these materials’ markets is the manufacturing facilities remain underutilized, this critical to comprehending the impact of continued EV surplus capacity could continue to place downward deployment on mineral production and vice versa. pressure on LIB prices. Recycling Critical Materials Global Supply Chains As production of LIB cells for EVs continues to Across the LIB cell supply chain, indirect value added expand to meet increasing demand, and the early (iVA) was greater than direct value added (dVA), generations of these batteries begin to reach the except for U.S. production of LIB cells, indicating end of their lifespans, the opportunities to reuse and that participation in the broader supply chain that recycle components and raw materials used in these supports LIB cell manufacturing at home and abroad technologies have grown in scale and importance. The was more important to the benchmarked economies development of closed-loop systems with end-of-life than domestic production of the intermediated and recycling can diminish environmental impacts from end product. All benchmarked economies received disposed batteries and provide secure sources of high- iVA from production in China, Japan, South Korea, value materials that can be recovered and reused to and Germany. In addition, none of the benchmarked produce new batteries at lower costs. economies could meet its own domestic demand for all supply chain components, relying instead on trade More detailed information can be found on the to fill gaps. following pages.

56 Market Trends Manufacturing LIB Cells

Li-ion Battery Cell Supply Chain, 2014-2016 Market and Trade Trends Benchmark Data: Market Trends LIB1: 2014-2016 LIB LIB 2: 2014-2016 LIB 3: 2014 - 2016 LIB LIB 4: 2014-2016 LIB App/LIB 2: 2014 -2016 FYI Calc for LIB2 and Production and Production/Demand Demand, Production, Cell LIB Detailed Trade LIB3 AutomotiveDeman dLIB (%) con tricellbu.. supplyTrends GWh/year chain demandMfg Capacity ( uandpdate.. productionImport/Export/BOT T ..shares, 2014–2016

LIB Cell LIB Cathode LIB Anode Separator Electrolyte Brazil 100% Canada

7% China 13% 13% 13% 13%

9% Denmark 80% 36% 38% 38% 38% 38% 39% 40% 40% 40% 40% Germany 13% India

6% Japan 40% 40% 40% 60% 40%

5% Malaysia 14% 14% 12% 14% 14% 11% 11% 11% 11% 6% Mexico nd Sh are s 6% 40% South Korea 17% 17% 17% 17% 22% 22% 22% 22% De m a Taiwan 59% 22% 23% 23% 23% 23% 22% United Kingdom 9% 9% 9% 20% 9% United States 10% 10% 10% 10% Rest of World 22% 22% 16% 16% 16% 16% 19% 14% 19% 14% 19% 14% 19% 14%

0% 5% 100% 13% 24% 27%

80% 35% 38% 40% 49% 48% 50% 50% 66% 65% 66% 69% 69% 60% 40% 14% 11% on Sh are s 46% t i 31% 50% 40% 18% odu c 21% 17% 22% 26% 26% P r 23% 9% 14% 22% 22% 20% 20% 10% 16% 14% 28% 26% 26% 15% 14% 13% 16% 19% 14% 9% 10% 10% 7% 11% 14% 12% 8% 0% 6% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Breakdown (in %) of global demand (top) and global production (bottom) by economy for benchmarked LIB cell supply chain intermediates. Note that due to lack of country-specific demand data for the LIB intermediates, the demand shares for intermediates was assumed to equal production shares of LIB cells.

Over the period, China’s demand and production reduction in market share for EV production. shares for LIB cells increased sharply, reflecting its investment in and promotion of electric vehicles (EVs). Production shares among the top five LIB economies U.S. demand and production shares decreased, shifted over the period as China and Germany and Japan and South Korea production shares also increased production more rapidly than the other declined. Across the other supply chain intermediates, economies in efforts to build vertically integrated China and Germany demand shares increased, while domestic and regional supply chains. In 2014, 40% distribution of production shares remained relatively of global LIB production occurred in Japan, and unchanged. about 23% of automotive LIBs were made in South Korea. China contributed 13% and the United States By 2016 China had the highest share of global demand contributed 19% to global automotive LIB production (36%) for LIB cells, while the U.S. share dropped to in 2014. In 2016, China’s share of global automotive LIB 22% (down from 59% in 2014). While absolute demand production increased to 38%, with Japan contributing for LIB cells grew over the period in the United States, 17%, South Korea contributing 9%, and the United its decrease in global demand share could indicate a States contributing 16%.

57 Manufacturing LIB Cells Market Trends

Li-ion Battery Cell Supply Chain, 2014-2016 Market and Trade Trends

AutomotiveLIB1: 2014-2016 LIB LIBcell supplyLIB 2: 2014-2016 chain demandLIB 3: 2014 - 2016 and LIB productionLIB 4: 2014-2016 LIB for keyApp/LIB LIB2: 2014 -201cell6 economies,FYI Calc for LIB2 and Production and Production/Demand Demand, Production, Cell LIB Detailed Trade LIB3 2014–2016Demand (%) contribu.. Trends GWh/year Mfg Capacity (update.. Import/Export/BOT T..

China Japan United States Germany South Korea

20

LIB Cell 10 GWh/year

0

20

LIB Cathode 10 GWh/year

0

20

LIB Anode 10 GWh/year

0

20

Separator 10 GWh/year

0

20

Electrolyte 10 GWh/year

0

2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Demand (color-coded lines) and production (gray bars) trends (in GWh) for LIB supply chain intermediates in key LIB cell economies. Economies are listed in order of 2016 production levels. Note that because of lack of data, global production of automotive LIB cells was assumed to equal global demand; the remaining benchmarked economies (not shown here) and the rest of the world make up the production-demand gaps indicated in the data presented.

58 Market Trends Manufacturing LIB Cells

China had the highest global demand and production Between 2014 and 2016, the United States, Japan, of LIB cells and supply chain intermediates over the and South Korea did not produce enough separators, period. Global demand for and production of LIB cells anodes, cathodes, and electrolytes to meet their and supply chain intermediates grew rapidly, largely respective domestic demands. In contrast, China driven by China’s increasing production of EVs for expanded production over the period to surpass domestic use. Japan, the United States, and South domestic demand for three of the four intermediates. Korea were the next three largest markets for LIB cells While major producers of cells, the United States and intermediates. None of the economies considered and Germany lagged in domestic production of could meet domestic demand for all supply chain intermediates, with the exception of some U.S. intermediates. production of separators, and instead depended on imports from Asia for upstream links in the supply chain. Between 2014 and 2016, annual global demand for and production of automotive LIB cells increased by 224% from 9.6 GWh to 31.3 GWh. 22 Demand for LIB cells was concentrated in China, the United States, Japan, South Korea, Germany, and the United Kingdom, with the biggest increases observed in China and the United States, driven by increases in the domestic demand for EVs.

To meet increasing demand, global production of LIB cells and supply chain intermediates (cathodes, anodes, electrolytes, and separators) more than tripled between 2014 and 2016. Production of LIB cells was concentrated in Japan, South Korea, the United States, and China. China’s LIB cell production in 2016 was more than two times that of any other benchmarked economy. For LIB cathodes, the biggest production increases were observed in China (increasing 3.4 times, from 4.6 GWh to 15.8 GWh), Japan (increasing 2.2 times, from 2.5 GWh to 5.6 GWh), and South Korea (increasing 3 times, from 1.4 GWh to 4.3 GWh). For LIB cell separators, the biggest production increases were observed in Japan (increasing 4.3 times, from 3.0 GWh to 15.8 GWh) and China (increasing 1.5 times, from 3.3 GWh to 8.4 GWh). For LIB cell anodes and electrolytes, the biggest production increases were observed in China and Japan. With limited amounts of raw materials (e.g., cobalt and lithium), many LIB manufacturers have tried to secure their material supply chains for cathode, anode, and electrolyte materials.

22 Because of lack of data, global production of automotive LIB cells was assumed to equal global demand. See CEMAC’s Benchmarks of Global Clean Energy Manufacturing, 2014-2016: Framework and Methodologies (Sandor et al. 2021) for details.

59 Manufacturing LIB Cells Market Trends

Automotive LIB cell supply chain manufacturing capacity utilization, 2014–2016

2014 2015 2016

Capacity Utilization Capacity Utilization Capacity Utilization 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%

China LIB Cell 17% 22% 27% United States LIB Cell 39% 13% 23% Japan LIB Cell 67% 23% 28% South Korea LIB Cell 49% 12% 16% United Kingdom LIB Cell 29% 0% 45% Germany LIB Cell 76% 93% 100% 0.0K 2.0K 4.0K 6.0K 8.0K 0.0K 2.0K 4.0K 6.0K 8.0K 0.0K 2.0K 4.0K 6.0K 8.0K $ million $ million $ million

China Separator 15% 22% 36% LIB Anode 61% 76% 58% LIB Cathode 59% 71% 62% Electrolyte 92% 97% 92% United Separator 34% 47% 35% States LIB Anode 0% LIB Cathode 20% 0% 0% Electrolyte 19% 0% 0% Japan Separator 36% 71% 100% LIB Anode 54% 70% 59% LIB Cathode 50% 51% 46% Electrolyte 41% 53% 63% South Korea Separator 70% 72% 24% LIB Anode 41% 62% 51% LIB Cathode 93% 89% 86% Electrolyte 36% 38% 26% United Separator Kingdom LIB Anode LIB Cathode Electrolyte 16% 0% 0% Germany Separator 8% 0% LIB Anode LIB Cathode 0% Electrolyte 0.0K 0.2K 0.4K 0.6K 0.8K 1.0K 1.2K 1.4K 0.0K 0.2K 0.4K 0.6K 0.8K 1.0K 1.2K 1.4K 0.0K0.2K 0.4K 0.6K 0.8K 1.0K 1.2K 1.4K $ million $ million $ million

LIB Cell Separator LIB Anode LIB Cathode Electrolyte 80% 73% 65% 57% 60% 61% 43% 58% 57% 59% 54% 50% 47% 40% 41% Capacity Utilization 28% 25% 20% 20% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Bars show manufacturing capacity (lighter shading) and utilized manufacturing capacity (i.e., production, darker shading) in US$2014 for key LIB cell economies. Vertical lines and associated numerical values show capacity utilization (production as a % of manufacturing capacity). Trend lines show global capacity utilization percentage for 2014–2016 (bottom). Note that LIB cells are displayed on a different scale than the LIB intermediates. Where “0%” is noted next to a vertical line, manufacturing capacity was available but no production occurred; where no percentage is noted, no manufacturing capacity was available.

60 Market Trends Manufacturing LIB Cells

Global manufacturing capacity increased for all LIB cell intermediates over the period. In addition, there was generally an excess of manufacturing capacity relative to global demand for all LIB intermediates in 2016, indicating the potential to expand future production as demand increases. Globally, capacity utilization generally increased for LIB cell intermediates and declined for LIB cells over the period.

Global manufacturing capacity for LIB cells was estimated at 23.5 GWh in 2014, 101.5 GWh in 2015 and 110.9 GWh in 2016. Corresponding production was estimated at 9.6 GWh in 2014, 20.4 GWh in 2015, and 31.2 GWh in 2016, reflecting a global excess of manufacturing capacity. Between 2014 and 2016, China, Japan, and South Korea were home to the majority of global manufacturing capacity for LIB cells and intermediates (cathodes, anodes, electrolytes, and separators). In 2016, the United States hosted 10% of global LIB cell manufacturing capacity but was not a significant manufacturer of intermediates. While capacity utilization generally increased for the LIB cell intermediates, capacity utilization for LIB cells dipped significantly (dropping from 41% in 2014 to 20% in 2015, and then increasing to 28% in 2016) over the period, driven by rapid expansion of manufacturing capacity in anticipation of future increased demand.

61 Manufacturing LIB Cells Trade Trends

Li-ion Battery Cell Supply Chain, 2014-2016 Market and Trade Trends Benchmark Data: Trade Trends LIB1: 2014-2016 LIB LIB 2: 2014-2016 LIB 3: 2014 - 2016 LIB LIB 4: 2014-2016 LIB App/LIB 2: 2014 -2016 FYI Calc for LIB2 and Production and Production/Demand Demand, Production, Cell LIB Detailed Trade LIB3 LIB cell (forDema nalld (%) coapplications)ntribu.. Trends GWh/year trade, 2014–2016Mfg Capacity (update.. Import/Export/BOT T..

2014 2015 2016 Brazil -$306M -$198M -$168M

Canada -$53M -$72M -$93M

China $2,051M $3,170M $3,630M

Denmark -$11M -$13M -$22M

Germany -$488M -$880M -$1,067M

India -$126M -$200M -$296M

Japan $1,496M $1,384M $1,822M

Malaysia $58M -$5M $70M

Mexico $0M -$226M -$165M

South Korea $1,806M $1,610M $1,888M

Taiwan $18M $46M $45M

United Kingdom -$82M -$154M -$240M

United States -$976M -$921M -$865M

Rest of World $53M $0M $0M

-$10,000M $0M $0M $10,000M -$10,000M $0M $0M $10,000M -$10,000M $0M $0M $10,000M Imports Exports Imports Exports Imports Exports d y a l orea a n Wor l an a d mark wan ay si a h K o f a i Indi a Ch in a Braz i Ja p United Me xi co T Ca n Kingdom Ma l De n Ger m Sou t Re st United States

$2,000M

Balance of Trade $0M 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Bar chart (top) shows imports (negative values) and exports (positive values) and balance of trade (exports less imports) in US$(2014) by economy. Line chart (bottom) shows balance of trade trends for LIB cells over the period.

Trade of LIB cells among the benchmarked economies despite increased domestic production, driven by their remained relatively stable over the period, with China, increased production of EVs. Japan, and South Korea the top exporters, leveraging their mature domestic supply chains originally In 2016, the 13 benchmarked economies exported developed to serve consumer electronics markets. almost $13.7 billion in LIB cells and packs, up from $11.6 Between 2014 and 2016, the United States and billion in 2014. Imports also increased from $8.2 billion Germany remained the largest importers of LIB cells, in 2014 to $9.2 billion in 2016.

62 10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report Lithium Ion Battery Trade Flow

 (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

(/index.html)

TradeGlobal Trendssummary of Secondary (Rechargeable)Manufacturing Lithium Ion Battery LIB trade Cells flow, in 2014 USD

Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV Wind (/benchmark/wind.html)

2014 2015 2016

Imports Brazil Exports LIB cell (for all applications) trade flow, $316.3 M $10.8 M Canada $96.1 M $754.1 M $42.8 M

$3.4 B

2014-2016 United States of America $1.7 B Trade flow data suggest growing global trade of LIB

$78.3 M cells among the economies over the period, generally China

United$1 Kingdom59.9 M following expected flows given the distribution of manufacturing capacity across economies. $59.0 M

Taiwan

$41.0 M

$5.5 B

China’s well-established vertical supply chains, 10/5/2020 Joint Institute for Strategic Energy Analysis: Benchmark Report Lithium Ion Battery Trade Flow

M .1 2 8 $

B .0 1

along with increased production and manufacturing  (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)$

Malaysia

Germany

M .0 4 2 $

capacity to meet increasing domestic demand, helped

M .8 3 4 5 $

M .3 4 4 1

lower the price of LIBs (in $/kWh), making China’s $

Mexico

M .4 3 2 $

M .4 5 0 3

(/index.html) $

LIB exports less expensive than those from other Denmark

M .3 2 1 $

Global summary of Secondary (Rechargeable) Lithium Ion Battery trade flow, in 2014 USD

B .3 2

countries. $

M .5 6 2 1 $

K 7 9 1 Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV $ Wind (/benchmark/wind.html)

India

South Korea South

M .4 2 6 4 $

B .0 2 $ M .8 6 3 5

2014 2015 2016$ Japan produced more LIB cells than its domestic Japan 2014 market demand and exported them to other countries Imports Brazil Exports $204.0 M $6.2 M Canada $98.7 M in North America and Europe. Germany was a net $788.5 M Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). $27.0 M $3.3 B United States of America importer of LIB cells, due in part to increasing demand $1.7 B from German automakers, in combination with https://www.jisea.org/benchmark/battery$85.html.1 M 1/2

United Kingdom China relatively immature European LIB supply chains. The $239.5 M United States’ supply chain was similarly immature,

$137.2 M

with most U.S. cell and battery plant operators Taiwan

$91.3 M

relatively new to the industry. $6.4 B

M .2 5 1

10/5/2020 $ Joint Institute for Strategic Energy Analysis: Benchmark Report Lithium Ion Battery Trade Flow

 (mailto:[email protected])  (https://twitter.com/jisea1)  (https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis)

B .4 1 $

Malaysia

M .4 0 2 Gray chords show LIB cell trade flows in US$(2014) among $

benchmarked economies. Width of bars are scaled to total Germany

M .4 6 8 4 flows, with lighter shading showing imports and darker shading $

M .9 6 3 1 $

showing exports. Note that gray chords only capture the trade

Mexico

M .7 2 6 3 $

M .4 6 2 among the 13 benchmarked economies; trade flows with the rest of (/index.html) $

Denmark

M .7 3 1 $

B .0 2

the world are not shown. Global summary of$ Secondary (Rechargeable) Lithium Ion Battery trade flow, in 2014 USD

M .8 0 0 2 $

K 3 5 3 Batteries (/benchmark/battery.html) LED (/benchmark/led.html) PV $ Wind (/benchmark/wind.html)

South Korea South

India M .2 8 9 3 $

B .0 2 $

M .5 7 2 6

2014 2015 2016$

2015 Japan

Imports Brazil Exports $171.7 M $3.7 M Canada $116.8 M $1.1 B Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html). $24.1 M $3.0 B

United States of America $2.0 B

https://www.jisea.org/benchmark/battery.html 1/2

$110.8 M

China

United Kingdom $350.9 M

$171.2 M

Taiwan

$6.6 B

$125.9 M

B .6 1 $

M .8 8 0 1 $

Germany

Malaysia

M .3 9 3 $

M .3 4 7 5 $

M .5 7 8 1 $

M .2 6 3 $

Denmark

Mexico

M .2 2 5 3 $ M .6 4 1 $

B .3 2 $

M .7 7 9 2 $

M .3 1 $

India

South Korea South

M .4 5 8 3 $

M .2 2 3 7 $ B .6 2 $

2016 Japan

63 Back to the Benchmarks of Global Clean Energy Manufacturing (/benchmark.html).

https://www.jisea.org/benchmark/battery.html 1/2 Manufacturing LIB Cells $ Value Added Trends

LIB Value Added Story Benchmark Data: Value Added Trends

LIB 6. 2014-2016 LIB LIB 7. 2014 - 2016 LIB FYI 2014-2016LIB FYI Calc for LIB 6 tVA (top LIB tVA - shares by Production, dVA, VA Automotive LIB cellec osupplynomies) chain totalnon-dome sticvalue econom yaddedretaine (tVA)d (top LIB ec o..by value added component, 2014–2016

China Japan United States Germany South Korea

1500 1,391 1,279 1,185 1,112 1,107 1,067 1000 745 765 667 LIB Cell $ million 481 501 485 500 405 401

106 0 659 600 434 400 303 Cathode $ million 206 232 200 162 136 113 34 55 0 15 22 7 14 21

200 192

134

100 92 Anode $ million 70 64 37 7 10 14 0 3 5 2 3 5 5

200 187

133 102 100

$ million 72 72 69 Separator 37 29 34 31 33 36 0 4 3 4

200 190

147

100 83

$ million 72 Electrolyte 47 54 19 20 9 17 17 0 7 2 3 3

2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Total value added (tVA) for the period in US$(2014) from manufacturing LIB cell intermediates in key LIB cell economies. Darker shading indicates direct VA (dVA), lighter shading indicates indirect VA (iVA), and numerical values show the total VA (tVA). Note that intermediates are shown on different scales. Economies are listed in order of tVA from LIB cell manufacturing in 2016.

64 Value Added Trends $ Manufacturing LIB Cells

China, Japan, and the United States accrued more total China’s was the least (15%). VA retained reflects the value added (tVA) from manufacturing LIB cells than extent of domestic supply chains as well as prevailing the other benchmarked economies from 2014 to 2016. wages, domestic profits, and taxes less subsidies. Supported by mature domestic supply chains, China, Japan, and South Korea accrued the highest levels of tVA from manufacturing LIB intermediates. Indirect value added (iVA) was generally greater than direct value added (dVA) for the benchmarked economies across the LIB supply chain, with the exception of U.S. production of LIB cells. tVA is generally highest in economies with the highest levels of production revenue. In 2014, Japan showed the highest tVA from LIB cell production, but by 2016 China significantly increased LIB cell production and overtook Japan ($1.4 billion to $1.2 billion). As the third-largest LIB cell producer, the United States accrued $1.1 billion in tVA, up from $670 million in tVA in 2014, due to increased production and iVA from production in other economies. tVA from LIB cells showed the greatest increases in Germany, China, and the United States over the period, as production ramped up to meet growing EV demand in those economies. tVA from global cathode production also increased in China, Japan, and South Korea. iVA was higher than dVA for all LIB supply chain intermediates considered, with the exception of LIB cells for the United States, indicating that participation in the broader supply chain that supports global LIB manufacturing domestically and internationally was more important to the benchmarked economies than domestic production of the intermediates (anodes, cathodes, separators, electrolyte) and end products (LIB cells).

Benchmarked economies retained varying shares of tVA as a portion of manufacturing revenue (i.e., VA retained) over the period from manufacturing LIB cells and intermediates. VA retained was highest in economies the United States, Germany and Japan. The United States’ VA retained for LIB cell production (65%) was higher than the other economies, while

65 Manufacturing LIB Cells $ Value Added Trends LIB Value Added Story

LIB 6. 2014-2016 LIB LIB 7. 2014 - 2016 LIB FYI 2014-2016LIB FYI Calc for LIB 6 Automotive LIB celltVA supply (top LIB chain totaltVA - shares value by addedProd u(tVA)ction, dVA, VAdomestic and non-domestic contribution, 2014–2016economies) non-domestic economy retained (top LIB eco..

China Germany Japan South Korea United States Country 100% Brazil VA Canada a l

o t China Denmark 50%

LIB Cell Germany India Japan Co nt r ib u ti o n t T 0% Malaysia 100%

VA Mexico a l

o t South Korea e Taiwan

ho d 50% United Kingdom Ca t United States

Co nt r ib u ti o n t T 0% 100% VA a l o t e 50% An o d

Co nt r ib u ti o n t T 0% 100% VA a l o t or

ara t 50% Se p

Co nt r ib u ti o n t T 0% 100% VA a l o t yte 50% ectro l E l

Co nt r ib u ti o n t T 0% 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

For each economy (listed across the top) color-coded bars show the share of tVA accrued from domestic and non-domestic production of LiB cell supply chain intermediates for 2014 to 2016. Domestic bars (generally the largest in each bar) represent the share of tVA (dVA + iVA) from domestic production. Non-domestic bars represent the share of tVA (iVA only) from production in other economies (dVA only occurs in the economy where production occurs).

In general, the benchmarked economies received the The United States and Germany accrued iVA over the greatest portion of tVA from domestic production of period from the manufacturing of LIB cells in China, LIB cells over the period. Production in China, followed Japan, and South benefitting from the global supply by Japan and South Korea, had the most impact on chains that connect the economies. the other benchmarked economies through iVA across the LIB cell supply chain.

66 Manufacturing LIB Cells

Raw Materials in the Manufacturing Supply Chain

The CEMAC benchmark framework presents To more effectively explore this important supply chain raw materials as the first link in the clean energy link, the following analysis examines raw materials used manufacturing supply chain. This link is intended to to manufacture cathode sheets for EV LIB cells (Igogo capture the value added from mining and processing et al. 2019). to refining stages in manufacturing clean energy Some of the key raw materials used in manufacturing technologies. LIBs include lithium, graphite, cobalt, and manganese. While information about final manufactured products is These materials are used to manufacture a range of generally available, upstream data for raw materials are products, from LIBs for consumer electronics and often difficult to gather, track, and analyze. Finding data vehicles to superalloys, hard metals, ceramics, and at the level of detail required for robust analysis remains polymers. a challenge. Disaggregating trade and market data As more EVs with LIBs are deployed, these vehicles to estimate the contribution to a specific technology are driving demand for materials used in LIB cells. can be difficult for any industry. The challenge is even Understanding these materials’ markets is critical to greater with clean energy technologies that account for comprehending the impact of continued EV deployment a very small, albeit growing, fraction of the market. At on mineral production and vice versa. The following the same time, benchmarking raw materials data has case study applies CEMAC benchmark methodologies the potential to provide a broader view of economies’ to assess cobalt as a raw material for LIB cells between accrual of value added from clean energy technology 2014 and 2016. manufacturing, along with additional insight into potential supply chain risks and opportunities.

Cobalt supply chain supporting lib manufacturing At a high level, the cobalt supply chain encompasses mining Raw materials Processed materials (which includes both large-scale and small-scale (artisanal) mining operations), ore processing cobalt ores and to produce concentrates and concentrates metal refinery intermediates, and metal and mining and cobalt product for chemical refining to extract raw ore Li-ion batteries separation, precursor materials.23 Battery- concentration and (copper, nickel, (cobalt sulfate and grade precursors are then used cobalt) primary refining cobalt oxide by cathode active material chemical cathode powders) refinery manufacturers to produce cobalt intermediates cathode sheets for LIB cells.

The cobalt supply chain encompasses mining, ore processing to produce concentrates and intermediates, and metal and chemical refining to extract precursor materials.

23 With nickel-cobalt or copper-cobalt ore and concentrates, the main goal is to extract nickel or copper; thus, these concentrates are likely to end up at metal refineries rather than chemical refineries. Metal scraps are likely to be processed to produce metal first, while recycled batteries are likely to be directly converted into cobalt chemicals.

67 Manufacturing LIB Cells

Benchmark Data: Market Trends Cobalt Story

LIB 9. 2014 - 2016 LIB 10. 2014 - 2016 LIB 13. 2014-2016 Cobalt Production and Cobalt Use in LDV Cobalt Product Exports Cobalt reserves and mines production,Capacity - Ore and Re.. Li-ion2014–2016 batteries

Cobalt Ore United South DRC ROW Russia Australia Belgium Canada China Cuba Japan Philippines Madagascar Zambia States Korea $80B

$2B $60B

$1B $40B Reserves Production $1B $20B

$0B $0B 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 Refined Cobalt

China Finland Belgium Canada Zambia Japan Norway Madagascar Australia Russia DRC ROW

$1,000M $1,000M

$500M $500M Production

$0M $0M Manufacturing Capacity 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6 201 4 201 5 201 6

Cobalt ore production (colored bars) and reserves (brown lines) with data shown on different scales (top).Refined cobalt production (colored bars) and manufacturing capacity (brown lines) with data shown on different scales (bottom).All data are in US$(2014). Note that cobalt ore reserves are much greater than production amounts. Note that the data show that refined cobalt production output from Belgium was greater than manufacturing capacity due to lack of inclusion of Belgium’s refinery capacity that is located in China and refined cobalt that is recovered from UMICORE’s recycling operations (USGS 2018: Cobalt Institute 2018; NREL estimates). Sources: USGS 2018; NREL estimates

Cobalt is usually mined as either a by-product of copper copper and nickel, which were impacted by oversupply (67%) or nickel (32%), with just 1% of production from and economic slowdown in economies such as China mines that are dedicated to cobalt extraction. Most (Shumsky 2015; Miller 2016; Burns 2016; IMF 2016; cobalt deposits are in the Central African Copperbelt, World Bank n.d.). which includes the Democratic Republic of the Congo The majority of the DRC’s active mines are owned (DRC), Central African Republic, and Zambia. In 2016, by Chinese companies (Darton Commodities 2018) the United States Geological Survey estimated global to support China’s large domestic cobalt refining cobalt reserves at about 7 million metric tons (MT). capacity, which accounted for about 47% of global The DRC led all suppliers with a global production refined cobalt production over the period. Global share of 53% from 2014 to 2016. Overall global cobalt excess manufacturing capacity (i.e., refining capacity) production decreased by 7% from 119,000 metric tons was about 37,000 metric tons in 2016, indicating that in 2014 to 111,000 metric tons in 2016 (USGS 2018). production could be expanded to meet additional The decrease was in part due to declining prices of demand without bringing new refineries online.

68 Manufacturing LIB Cells

Cobalt Story

Cobalt demand to supportLIB 9. EV 2014 - 2016LIB cathodeLIB 10. 2014 -production, 2016 LIB 13. 2014-2016 2014–2016 Cobalt Production and Cobalt Use in LDV Cobalt Product Exports Capacity - Ore and Re.. Li-ion batteries

China Belgium Japan South Korea ROW

$550M

$500M

$450M

$400M

$350M d $300M Dema n

$250M

$200M

$150M

$100M

$50M

$0M 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Cobalt demand for economies manufacturing cathode materials for EV LIB cells. All data are in US$(2014). Source: NREL estimate

Global sales of EVs grew by 133% from 323,000 vehicles period, the proportion of cobalt used globally in EV LIB in 2014 to 753,000 vehicles in 2016 (IEA 2017), driving cells surged from 1.4% to 5% of total mine production. a sharp increase in demand for LIBs from 9.6 GWh to China remained the largest consumer of cobalt used to 31.1 GWh (Pillot 2017; BNEF 2017d). During the same manufacture EV LIBs.

69 Manufacturing LIB Cells

CobaltBenchmark Story Data: Trade Trends

LIB 9. 2014 - 2016 LIB 10. 2014 - 2016 LIB 13. 2014-2016 Cobalt Production and Cobalt Use in LDV Cobalt Product Exports Cobalt material exports,Capacity - Ore 2014–2016 and Re.. Li-ion batteries

Cobalt Ore and Concentrates Cobalt Intermediates Cobalt Oxide and Hydroxide Trading Partner Belgium 2014 $1M $81M $175M Belgium $1M $84M $172M 2015 China 2016 $2M $67M $125M Finland Germany Canada 2014 $5M $259M $0M Japan 2015 $0M $246M $0M Netherlands

2016 $0M $211M $0M Rest of World South Korea $0M $103M $225M China 2014 United Kingdom 2015 $1M $84M $173M United States

2016 $0M $53M $187M Zambia

DRC 2014 $591M $753M $54M

2015 $676M $776M $301M

2016 $295M $769M $269M

Finland 2014 $0M $222M $119M

2015 $0M $191M $84M

2016 $0M $150M $103M

Madagascar 2014 $29M $88M $0M

2015 $22M $95M $0M

2016 $0M $79M $0M

Norway 2014 $0M $101M $0M

2015 $0M $100M $0M

2016 $0M $107M $0M

United 2014 $1M $29M $32M Kingdom 2015 $1M $30M $25M

2016 $1M $27M $24M

Zambia 2014 $0M $113M $3M

2015 $3M $85M $4M

2016 $15M $99M $5M

$0M $200M $400M $600M $800M $0M $200M $400M $600M $800M $0M $200M $400M $600M $800M Exports Exports Exports

Bars show exports from country listed on left to countries as color coded for three categories of materials: cobalt ores and concentrates (HS-260500); cobalt oxides, hydroxides, commercial cobalt oxides (HS-282200), and cobalt mattes; and other intermediate products of cobalt metallurgy including unwrought cobalt and powders (HS-810520). Sources: UN-Comtrade n.d.; Trademap n.d.; NREL estimates

Globally traded cobalt materials include ores and Most economies in this period experienced a slight concentrates; intermediate products of cobalt decrease in exports and imports of cobalt, consistent metallurgy, unwrought cobalt, and powders; and oxides with the economic slowdown in emerging markets and hydroxides. The DRC was the leading exporter of and an associated decline in metal prices. The DRC cobalt materials, while China was the leading importer. experienced a substantial drop in exports relative to From 2014 to 2016, the DRC exported a total of $4.5 other economies, with exports decreasing by 31% from billion worth of cobalt materials, primarily to China and $1.8 billion in 2015 to $1.3 billion in 2016 Zambia. During the same period, China imported a total of $3.0 billion worth of cobalt materials, almost entirely from the DRC.

70 Manufacturing LIB Cells

CoBenchmark & Li VA Story Data: Value Added Trends

LIB 11. Cobalt Mining LIB 12. Refined Cobalt Calc for LIB 11 Calc for LIB 12 iVA and dVA iVA and dVA Cobalt mining total(blue value-orange line s)added(blue (tVA),-orange lines) 2014–2016 India Brazil China Japan Korea South States United Taiwan Mexico United Canada Malaysia Germany Kingdom Denmark

$14M

$80M Direct value added (dVA) $13M Indirect value added (iVA)

$12M $70M

$11M

$60M $10M

$9M $50M $8M

$7M Value Added $40M

$6M

$30M $5M

$4M $20M $3M

$2M $10M

$1M

$0M $0M 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Blue lines indicate dVA and orange lines indicate iVA for cobalt mining. All data are in US$(2014). Economies are listed in order of tVA from cobalt mining in 2016. Note that Canada is on a different scale.

Only two of the benchmarked economies—Canada and with $8.8 million in dVA accrued in 2016. the United States—accrue dVA from domestic cobalt Based on the available data24, tVA from cobalt mining mining. However, the 13 benchmarked economies within the benchmarked economies was significantly all accrue some iVA from cobalt mining in Russia, smaller than tVA from manufacture of LIB cells, Australia, Philippines, Canada, and the United States. cathodes, anodes, and electrolytes. In 2016, tVA tVA from cobalt mining in Canada declined to $91.8 accrued by the five economies for which there are data million in 2016, down from $104.6 million in 2014. (Canada, the United States, Russia, Australia, and the tVA in the United States increased from $11.0 million in Philippines) from cobalt mining was $127 million, while 2014 to $20.2 million in 2016. Almost half of the tVA in tVA from the finished product (cells) was $5.3 billion. the United States came from domestic cobalt mining,

24 This analysis is limited due to the fact that the Organization for Economic Co-operation and Development (OECD) Structural Analysis (STAN) Input- Output (I-O) database used to estimate tVA does not include the world’s largest producer—the DRC—nor Cuba, Madagascar, or Zambia.

71 Manufacturing LIB Cells

Co & Li VA Story

Cobalt refining totalLIB 11. Cobalt value Mining addedLIB 12. Refine (tVA),d Cobalt 2014–2016Calc for LIB 11 Calc for LIB 12 iVA and dVA iVA and dVA (blue-orange lines) (blue-orange lines) y a l orea a n an a d mark wan ay si a h K a i Indi a States Braz i Ja p China United United Me xi co T Ca n Kingdom Ma l De n Ger m Sou t $800M $70M $750M Direct value added (dVA) Indirect value added (iVA) $65M $700M

$650M $60M

$600M $55M

$550M $50M

$500M $45M

$450M $40M

$400M ue Add ed $35M Va l $350M $30M $300M $25M $250M $20M $200M

$15M $150M

$10M $100M

$50M $5M

$0M $0M 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016 2014 2015 2016

Blue lines indicate dVA and orange lines indicate iVA for cobalt refining.All data are in US$(2014). Economies are listed in order of tVA from cobalt mining in 2016. Note that China is on a different scale.

Three of the benchmarked economies—China, Japan, China, as the primary cobalt refiner, accrued the highest and Canada—accrued dVA from domestic cobalt tVA, with $940 billion in 2014, $1.0 billion in 2015, and refining. In addition, all 13 benchmarked economies $870 million in 2016. tVA in the United States declined accrued some iVA from cobalt refining in Australia, from $40.2 million in 2014 to $32.9 million in 2016. Belgium, Finland, Norway, Canada, China, and Japan. As the United States did not refine cobalt during the tVA attributed to cobalt refining was greater than from period, this drop was entirely due to changes in other other intermediates in the LIB raw material supply chain: countries’ production levels. $1.2 billion in 2016, down from $1.3 billion in 2014.

72 Manufacturing LIB Cells

LIB Challenges and Opportunities: Recycling Critical Materials

Global li-ion battery recycling capacity, 2016

Bars show estimated global electrochemical storage batteries (except for lead acid batteries) recycling capacities (MT) and cathode, anode, and electrolyte production capacity for virgin materials (MT) in 2016. Donut charts show the percentage of recycling capacity used for each recycling method (pyrometallurgy, hydrometallurgy, or mechanical separation). Triangles show separator production capacity (sq. meters) in 2016.

Source: Mayyas, Steward, and Mann 2019

As production of clean energy technologies expands to cathode material exceeded 180,000 metric tons in 2016 meet increasing demand, the opportunities to reuse and (Avicenne Energy 2017). Some scarce critical elements recycle components and raw materials used in these used in LIBs, namely cobalt and lithium, are mined technologies have grown in scale and importance. LIB and/or processed in only a few countries where trade cells for LDVs present one such opportunity, as the early policies or geopolitics could limit availability and create generations of these batteries begin to reach the end of instabilities in market prices, which could discourage their lifespans. further expansion of the market for EVs and other products that rely on LIBs. Demand for LIBs continues to increase in response to the growing demand for EVs. The global demand for LIB

73 Manufacturing LIB Cells

Closed-loop systems with end-of-life recycling not only diminish environmental impacts from disposed batteries, but also provide sources of high-value materials that can be recovered and reused to produce new batteries at potentially lower costs.

In 2016, as shown above, world battery recycling capacity exceeded 98,000 metric tons (Mayyas, Steward, and Mann 2019), including all types of electrochemical storage batteries, except for lead- acid batteries (the ones most commonly used in traditional internal combustion engine vehicles). Three technologies are used alone or in combination for recycling LIBs: pyrometallurgy, hydrometallurgy, and mechanical processes such as cryomilling (Ordoñez, Gago, and Girard 2016; Ellis and Mirza n.d.). Currently these technologies focus on recycling the high-value cobalt (Co), lithium (Li), and nickel (Ni) cathode materials, ignoring anodes and other pack components.

Environmental regulations regarding LIBs differ from one region to another. China hosted 32% of the world’s battery recycling capacity in 2016, and this capacity is expected to grow significantly in the near-term (Dai 2018).25 After establishing end-of-life policies and regulations for vehicles in 2001 (Wang and Chen 2013), China more recently issued ”interim” rules requiring automakers to collect and recycle the retired LIBs from electric vehicles (Wang and Chen 2013). Japan (METI 2006), Korea (ChemSafetyPro 2007), and the European Union (European Parliament 2000) also have policies for dealing with end-of-life vehicle components, including batteries.

The primary challenges for LIB recycling in the near term in many countries, including the United States, are the lack of viable collection mechanisms, low collection volumes, and insufficient economic information to inform investors on long-term costs and benefits. All these factors have resulted in a lower LIB recycling volume compared to that for recycled lead-acid and nickel-metal hydride (NiMH) batteries (Steward, Mayyas, and Mann 2018).

25 Tom Daly, “Chinese Carmaker BYD Close to Completing Battery Recycling Plant,” Reuters, March 21, 2018, https://www.reuters.com/article/uschina- byd-batteries-recycling/chinese-carmaker-byd-close-to-completing-batteryrecycling-plant-idUSKBN1GX1EZ.

74 Glossary

respect to clean energy manufacturing supply chains, and Glossary the domestic iVA indicates the strength of domestic supply chains. Clean Energy Technologies: Technologies that produce energy with fewer environmental impacts than conventional Manufacturing Capacity: Amount of product in physical technologies, or enable existing technologies to operate units that can be produced in a given period by existing more efficiently, consuming fewer natural resources to physical plant and other necessary infrastructure (e.g., deliver energy services. Clean energy technologies may megawatts of PV modules per year). Production is the include renewable energy, clean non-renewable energy, and physical amount of a product actually produced in a energy efficiency technologies for electricity generation, fuel given period. Manufacturing capacity and production are production, and sustainable transportation. measures that reflect supply. Capacity and production, in combination with market size and growth, are the basic Clean Energy Technology End Product: The finished metrics used in assessing the supply, demand, and trade product of the manufacturing process, assembled flow dynamics occurring within an industry. from subcomponents, and ready for sale to customers as a completed item. Clean energy examples include Manufacturing Supply Chain: A complex and dynamic photovoltaic (PV) modules and lithium-ion battery (LIB) supply and demand network consisting of an integrated cells. In this link of the supply chain, value added comes system of organizations, people, activities, information, from assembling subcomponents into a marketable product and resources involved in moving a product or service that customers value. from supplier to customer. Supply chain activities involve the procurement of transformation, and logistics of, raw Clean Energy Technology Manufacturing: Manufacturing of materials, and components into a finished product that is clean energy end products (renewable energy, sustainable delivered to the end customer. transportation, and energy efficiency technologies) and their associated supply chains, as well as activities to Manufacturing Value Chain: The value created in each step improve manufacturing of all products by increasing of the supply chain though the key activities that companies energy productivity and using low-cost domestic fuels and perform to bring a product from conception to end use. feedstocks in the production process. Value chain activities can produce goods or services, include a single company or span multiple companies, and Direct Value Added (dVA): Value added from the output occur within a single geographical location or are spread of the sector in question, accrued solely from the domestic across economies. While the supply chain tracks the flows manufacturers of a product. For example, if solar PV module of raw materials and intermediate products to customers manufacturing generated $100 million in production revenue (upstream to downstream), the value chain tracks the with 70% of that associated with intermediate inputs, dVA demand and cash flows from customers to companies would consist of the remaining $30 million. (downstream to upstream). Indirect Value Added (iVA): Value added that is supported Estimate of the demand for a specific product by the intermediate expenditures made by the sector in Market Size: or service within an economy, typically expressed in units question. The indirect VA (iVA), sometimes referred to as of product volume (e.g., megawatts of PV modules) and in “the economic ripple effect,” comes from the greater supply terms of monetary value (e.g., US$). The latter expression chain that provides domestic inputs used by manufacturers of market size accounts for both demand volumes and (domestic iVA), and from goods and services exported selling prices. Market size serves as a core metric of demand to support manufacturing that takes place in other development and growth over time, and it is a key measure economies (non-domestic iVA). This is a comprehensive of the relative importance of an industry within an economy figure that captures all supply chain activity (domestic and globally. and international) needed to support the output of the sector in question within the economy in question. The Processed Materials: Materials that have been transformed non-domestic (intercountry) iVA indicates the globalization or refined from basic raw materials as an intermediate step and interconnectedness of benchmarked economies with in the manufacturing process. Processed materials include

75 Glossary

steel, glass, and cement. In this link of the supply chain, value Value Added Retained: A measure of an industry’s added comes from processing raw materials into precursors contribution to GDP per unit of production revenue. Value that can be more easily transported, stored, and used for added retained is calculated by dividing manufacturing VA downstream subcomponent fabrication. by production revenue. High wages and larger economies tend to retain higher levels of VA, as more inputs can be Raw Materials: Basic materials mined, extracted, or sourced domestically, and workers are paid higher wages. harvested from the earth. Also referred to as “unprocessed material,” examples include raw biomass and iron ore. In this Wind Generator Sets (Gensets): Assembled nacelles link of the supply chain, value added comes from extracting, shipped with blades. harvesting, and preparing raw materials for international marketing in substantial volumes.

Subcomponents: Unique constituent parts or elements that contribute to a finished product. Clean energy technology examples include generation sets for wind turbines and crystalline silicon wafers for crystalline silicon PV modules. Note that what is considered a component by the manufacturer may be considered the finished product by its supplier. In this link of the supply chain, value added comes from fabricating processed materials into subcomponents that can then be assembled (with other subcomponents) into end products.

Total Value Added (tVA): Total value added is the sum of indirect and direct value added.

Trade: The buying and selling of goods and services between economies. Trade measures include the balance of trade in terms of the amount of goods that one economy sells to other economies (exports) minus the amount of goods that same economy buys from other economies (imports).

Trade Flows: As applied in this report, identifies the trade between specific economies for specific intermediates and the end product in a supply chain.

Value Added (VA): The contribution from an industry or government sector to overall GDP. VA consists of labor payments, gross operating surplus, and taxes, and can be a measure of GDP. Labor payments include all payments to workers, including benefits. Gross operating surplus is a property-type income that includes payments for capital (including depreciation) and payments to investors and includes profits. Taxes are net payments to or from the government with subsidies potentially resulting in negative tax amounts.

76 Acronyms

List of Acronyms and Abbreviations

ANL Argonne National Laboratory PV photovoltaic

ASP average selling price ROW the rest of the world

BNEF Bloomberg New Energy Finance tVA total value added

BOS balance of system USITC U.S. International Trade Commission

BOT balance of trade VA value added

CEMAC Clean Energy Manufacturing Analysis Center c-Si crystalline silicon

DOE U.S. Department of Energy

DRC Democratic Republic of the Congo dVA direct value added

GDP gross domestic product

GW gigawatt

GWh gigawatt hour

I-O input-output iVA indirect value added kWh kilowatt hour

LCOE levelized cost of electricity

LDV light-duty vehicle

LED light-emitting diode

LIB lithium-ion battery

M million

MT metric tons

MW megawatts

NiMH nickel-metal hydride

NREL National Renewable Energy Laboratory

OECD Organisation for Economic Cooperation and Development poly-Si polysilicon

77 References

References Ahmad Mayyas, Darlene Steward, Margaret Mann. “The Case 2016). Accessed February 2, 2019. https://agmetalminer. for Recycling: Overview and Challenges in the Material Supply com/2016/09/14/what-is-the-copper-market-telling-us- Chain for Automotive Li-Ion Batteries. Article No. e00087.” oversupply-or-demand-weakness/. Sustainable Materials and Technologies, Vol. 19 (2019): 13 pp. Campa, José Manuel. 2000. Exchange Rates and Trade: How https://doi.org/10.1016/j.susmat.2018.e00087. Important is Hysteresis in Trade? IESE Research Division, Avicenne Energy. Christopher Pillot, The Rechargeable Battery University of Barcelona. Research Paper No. 427. Market and Main Trends 2016-2025. International Battery https://core.ac.uk/download/pdf/6626516.pdf. Seminar and Exhibit. Fort Lauderdale, FL, March 2017 20. Chase, Jenny, Pietro Radoia, Xiaoting Wang, and Hugh Bloomberg L.P. 2016. “PV Operating Margins.” Bloomberg Bromley. 2016. Q2 2016 Global PV Market Outlook: Making Terminal. Accessed Aug 1 ,2017. Hay While the Sun Shines. Market Research Report. New York: Bloomberg New Energy Finance. BNEF (Bloomberg New Energy Finance). 2018a. “U.S. Business Energy Investment Tax Credit.” Bloomberg New Energy ChemSafetyPRO. 2007. “Korea RoHS/ELV/WEEE.” National Finance Desktop Portal. Accessed August 7, 2019. Assembly of Korea. April 2, 2007. Accessed May 25, 2018. https://www.bnef.com/Policy/126/. https://www.chemsafetypro.com/Topics/Korea/Korea_ RoHS_WEEE.html. ———. 2018b. “Germany Offshore Wind Feed-in Tariff.” Bloomberg New Energy Finance Desktop Portal. Chung, Donald, Emma Elgqvist, and Shriram Accessed August 7, 2019. https://www.bnef.com/Policy/112/. Santhanagopalan. 2016. Automotive Lithium-Ion Cell Manufacturing: Regional Cost Structures and Supply Chain ———. 2017a. “China National Wind Fixed Feed-in Considerations. Golden, CO: Clean Energy Manufacturing Tariff.” Bloomberg New Energy Finance Desktop Portal. Analysis Center. NREL/TP-6A20-66086. Accessed August 7, 2019. https://www.bnef.com/policy/201. https://doi.org/10.2172/1247459. ———. 2017b. “Germany Onshore Wind Feed-in Tariff and Cision PR Newswire. 2013. “Electric Vehicle Batteries: Lithium Premium.” Bloomberg New Energy Finance Desktop Portal. Ion Batteries for Hybrid, Plug-in Hybrid, and Battery Electric Accessed August 7, 2019. https://www.bnef.com/Policy/110/. Light Duty Vehicles, Market Analysis and Forecasts.” Janu 14, ———. 2017c. “Q1 2017 Global PV Market Outlook.” Bloomberg 2013. https://www.prnewswire.com/news-releases/electric- New Energy Finance Desktop Portal. Accessed August 7, vehicle-batteries-lithium-ion-batteries-for-hybrid-plug-in- 2019. https://www.bnef.com/core/insights/15928/view. hybrid-and-battery-electric-light-duty-vehicles-market- analysis-and-forecasts-186760071.html. ———. 2017d. “Q1 2017 Global Electrified Transport Market Outlook.” Bloomberg New Energy Finance Desktop Portal. Cobalt Institute. 2018. Cobalt News Magazine. Issue 2: April Accessed September 29, 2017. 2018. https://www.cobaltinstitute.org/assets/files/cobalt- news/2018%20Cobalt%20News/Cobalt-News-Ap2018.pdf ———. 2016. Aleksandra O’Donovan. “Q1 2016 Global Electrified Transport Market Outlook.” Bloomberg New Energy Curry, Claire. 2017. Lithium-Ion Battery Costs and Market: Finance Desktop Portal. Accessed February 19, 2018. Squeezed Margins See Technology Improvements and https://www.bnef.com/core/insights/13776/view. New Business Models. Bloomberg New Energy Finance (BNEF). July 5, 2017. https://data.bloomberglp.com/bnef/ ———. 2015. “Q4 2015 Global PV Market Outlook.” Bloomberg sites/14/2017/07/BNEF-Lithium-ion-battery-costs-and- New Energy Finance Desktop Portal. Accessed August 7, 2019. market.pdf. https://www.bnef.com/core/insights/13225/view. Dai, Qiang. 2018. “China Perspective on Lithium-Ion Battery Bradsher, Keith. 2014. “As LED Industry Evolves, China Elbows Recycling.” Presented at the International Workshop on Ahead.” New York Times (June 17, 2014). https://www.nytimes. Lithium-Ion Battery Recycling, May 30–31, 2018, Golden, CO. com/2014/06/18/business/as-led-industry-evolves-china- Unpublished. elbows-its-way-to-the-fore.html. Darton Commodities Ltd. 2018. Cobalt Market Review Burns, Stuart. 2016. “What is the Copper Market Telling Us? 2017/2018. March 2018. Guildford, UK. Oversupply? Or Weakness?” MetalMiner (September 14,

78 References

EIA (U.S. Energy Information Administration). 2017. Annual GWEC (Global Wind Energy Council). 2017. Global Wind Energy Outlook 2017 with Projections to 2050. January 5, Report. Annual Market Update 2016. Brussels: Global Wind 2017. https://www.eia.gov/outlooks/aeo/pdf/0383(2017).pdf. Energy Council.

Ellis, Timothy W., and Abbas H. Mirza. 2017. Battery Recycling: IEA (International Energy Agency). 2017. Global EV Outlook Defining the Market and Identifying the Technology Required 2017: Two Million and Counting. Paris: IEA. to Keep High Value Materials in the Economy and Out https://doi.org/10.1787/9789264278882-en. of the Waste Dump. National Institute of Standards and Igogo, Tsisilile, Debra Sandor, Ahmad Mayyas, and Jill Technology (NIST). https://www.nist.gov/system/files/ Engel-Cox. 2019. Supply Chain of Raw Materials Used in the documents/2017/04/28/245_battery_recycling_defining_ Manufacturing of Light-Duty Vehicle Lithium-Ion Batteries. the_market.pdf Golden, CO: Clean Energy Manufacturing Analysis Center. En.Lighten. n.d. “Global Lighting Map.” Accessed May 20, 2016. NREL/TP-6A20-73374. https://doi.org/10.2172/1560124. http://www.enlighten-initiative.org. IMF (International Monetary Fund). 2016. World Economic European Parliament (The European Parliament and the Outlook: Subdued Demand: Symptoms and Remedies. Council of the European Union). 2000. “Directive 2000/53/ October 2016. Washington, D.C.: International Monetary EC of the European Parliament and of the Council of 18 Fund. https://www.imf.org/en/Publications/WEO/ September 2000 on End-of Life Vehicles: Commission Issues/2016/12/31/Subdued-Demand-Symptoms-and- Statements.” Official Journal of the European Communities. Remedies. http://data.europa.eu/eli/dir/2000/53/oj. ———. 2015. Spillovers from Dollar Appreciation. Prepared by EWEA (European Wind Energy Association). 2009. The Florence Jaumotte. July 2015. Washington, D.C.: International Economics of Wind Energy. Søren Krohn, editor. Brussels: Monetary Fund. https://www.imf.org/external/pubs/ft/ European Wind Energy Association. http://www.ewea.org/ pdp/2015/pdp1502.pdf. fileadmin/files/library/publications/reports/Economics_of_ ———. 2020. Renewable Power Generation Costs in 2019. Abu Wind_Energy.pdf. Dhabi: International Renewable Energy Agency. https://irena. Federal Reserve Bank of St. Louis. 2018. “Trends in Capacity org/-/media/Files/IRENA/Agency/Publication/2020/Jun/ Utilization Around the World.” The FRED Blog. April 9, 2018. IRENA_Power_Generation_Costs_2019.pdf. https://fredblog.stlouisfed.org/2018/04/trends-in-capacity- IRENA (International Renewable Energy Agency). n.d. utilization-around-the-world/. “Statistics Time Series.” Accessed August 31, 2020. ———. n.d. “Annual Rates.” Accessed 2018. https://fred. https://www.irena.org/Statistics/View-Data-by-Topic/ stlouisfed.org/categories/32219. Capacity-and-Generation/Statistics-Time-Series.

Fraunhofer ISE. 2018. Annual Report 2017/18. Christina ITRPV. 2018. International Technology Roadmap for Lotz, editor. Freiburg: Fraunhofer Institute for Solar Energy Photovoltaic: Results 2017 Including Maturity Report Systems ISE. https://www.ise.fraunhofer.de/content/dam/ 2018. Ninth ed. September 2018. https://pv.vdma.org/ ise/en/documents/annual_reports/fraunhofer-ise-annual- documents/105945/26776337/ITRPV%20Ninth%20 report-2017-2018.pdf. Edition%202018%20including%20maturity%20report%20 20180904_1536055215523.pdf/a907157c-a241-eec0-310d- Fu, Ran, David Feldman, Robert Margolis, Mike Woodhouse, fd76f1685b2a. and Kristen Ardani. 2017. U.S. Solar Photovoltaic System Cost Benchmark: Q1 2017. Golden, CO: National Renewable Energy Kavlak, Goksin, James McNerney, and Jessika E. Trancik. 2018. Laboratory. https://doi.org/10.2172/1395932. “Evaluating the Causes of Cost Reduction in Photovoltaic Modules.” Energy Policy 123: 700–710. Fullenkamp, Patrick H, and Holody, Diane S. 2014. U.S. Wind https://doi.org/10.1016/j.enpol.2018.08.015. Energy Manufacturing and Supply Chain: A Competitiveness Analysis. 2014. Prepared for the U.S. Department of Energy. Leigh, Daniel, Weicheng Lian, Marcos Poplawski-Ribeiro, Cleveland: GLWN, Global Wind Network. DE-EE-0006102. Rachel Szymanski, Viktor Tsyrennikov, and Hong Yang. 2017. https://doi.org/doi:10.2172/1156678. Exchange Rates and Trade: A Disconnect? IMF Working Paper WP/17/58. March 15, 2017. https://www.imf.org/en/ GTM Research. 2017. U.S. PV Racking 2017: Landscape, Pricing Publications/WP/Issues/2017/03/15/Exchange-Rates-and- and Forecasts. Boston: Greentech Media. April 26, 2017. Trade-A-Disconnect-44746.

79 References

Lowe, Marcy, Saori Tokuoka, Tali Trigg, and Gary Gereffe. 2010. Richter, Felix. 2017. “Electric Cars Have Lots of Room to Grow.” Lithium-Ion Batteries for Electric Vehicles: The U.S. Value Chain. Electric Mobility. June 8, 2017. Statista. https://www.statista. Center on Globalization, Governance and Competitiveness. com/chart/9742/registrations-of-electric-cars-worldwide/. Duke University. October 5, 2010. https://unstats.un.org/ Rose, Andrew K. 1991. “The Role of Exchange Rates unsd/trade/s_geneva2011/refdocs/RDs/Lithium-Ion%20 in a Popular Model of International Trade: Does the Batteries%20(Gereffi%20-%20May%202010).pdf. ‘Marshall-Lerner’ Condition Hold?” Journal of International MAKE (MAKE Consulting). 2015a. Global Wind Turbine Trends Economics 30(3-4): 301–316. https://doi.org/10.1016/0022- 2015. December 23, 2015. Aarhus, Denmark: MAKE Consulting. 1996(91)90024-Z.

Mayyas, Ahmad, Darlene Steward, and Margaret Mann. 2018. Sanderson, Susan Walsh, and Kenneth L. Simons. 2014. “Light “Impact of Li-ion Battery Recycling on the Supply Chain.” Emitting Diodes and the Lighting Revolution: The Emergence Advanced Automotive Battery Conference. San Diego, CA. of a Solid-State Lighting Industry.” Research Policy 43(10): June 4–8, 2018. 1730–1746. http://dx.doi.org/doi:10.1016/j.respol.2014.07.011.

METI (Japanese Ministry of Economy, Trade and Industry). Sandor, Debra, David Keyser, Samantha Reese, Ahmad 2006. “End-of-Life Vehicle Recycling Law.” Last updated May Mayyas, Ashwin Ramdas, Scott Caron, James McCall, and 22, 2006. Accessed April 13, 2018. http://www.meti.go.jp/ Jill Engel-Cox. 2021. Benchmarks of Global Clean Energy policy/recycle/main/english/law/end.html. Manufacturing, 2014-2016: Framework and Methodologies. Golden, CO: Clean Energy Manufacturing Analysis Center. Miller, John W. 2016. “Supermines Add to Supply Glut of NREL/TP-6A50-78038. Metals.” Wall Street Journal (January 4, 2016). Accessed August 12, 2019. https://www.wsj.com/articles/ Schubert, E. Fred. 2003. Light-Emitting Diodes. Cambridge, supermines-add-to-supply-glut-of-metals-1451952511. UK: Cambridge University Press. ISBN 0-8194-3956-8.

Mukish, Pars, and Eric Virey. 2017. LED Packaging Market, Shumsky, Tatyana. 2015. “Copper’s Rout Deepens as China’s Technology and Industry Landscape 2017. Lyon, France: Yole Inflation Decelerates.”Wall Street Journal. Updated November Développement. 10, 2015. Accessed February 20, 2019. https://www.wsj.com/ articles/copper-extends-six-year-low-on-weak-chinese- ———. 2014. LED Packaging Technology and Market Trends inflation-data-1447169948?mod=article_inline. 2014. Lyon, France: Yole Développement. Su, Yu-Shan. 2014. “Competing in the Global LED Industry: The Navigant Consulting. 2014. Energy Savings Forecast of Case of Taiwan.” International Journal of Photoenergy (2014): Solid-State Lighting in General Illumination Applications. 1–11. http://dx.doi.org/doi:10.1155/2014/735983. Prepared for the U.S. Department of Energy. Washington, D.C.: U.S. Department of Energy. DOE/EE-1133. Trademap. n.d. “Trade Map” (lithium and cobalt trade data). https://doi.org/10.2172/1220291. Accessed February 1, 2019. https://www.trademap.org.

Navigant Research. 2015. LED Lighting: Global Outlook Tuttle, Ralph, and Mark McClear. 2014. “Understand the Commercial, Residential, Industrial, Outdoor, and Automotive True Cost of LED Choices in SSL Systems.” LEDs Magazine Markets: Global Market Analysis and Forecasts. (February 2014). https://www.ledsmagazine.com/leds-ssl- design/packaged-leds/article/16695263/understand-the- Ordoñez, J., E. J. Gago, and A. Girard. 2016. “Processes true-cost-of-led-choices-in-ssl-systems-magazine. and Technologies for the Recycling and Recovery of Spent Lithium-Ion Batteries.” Renewable and Sustainable UN-Comtrade. n.d. “UN Comtrade Database.” Cobalt and Energy Reviews 60: 195–205. lithium trade data accessed February 1, 2019. https://doi.org/10.1016/j.rser.2015.12.363. https://comtrade.un.org.

Pillot, Christopher. 2017. The Rechargeable Battery Market USGS (U.S. Department of the Interior U.S. Geological Survey). and Main Trends 2016–2025. International Battery Seminar 2018. Mineral Commodity Summaries 2018. Reston, VA: U.S. and Exhibit. Fort Lauderdale, Florida. March 20, 2017. Geological Survey. https://s3-us-west-2.amazonaws.com/ Avicenne Energy. prd-wret/assets/palladium/production/mineral-pubs/mcs/ mcs2018.pdf. Quitzow, Rainer. 2015. “Dynamics of a Policy-Driven Market: The Co-Evolution of Technological Innovation Systems for Valley Recycling. 2015. “Metal Market is Still in Oversupply.” Solar Photovoltaics in China and Germany.” Environmental Posted August 4, 2015. Accessed February 20, 2019. Innovation and Societal Transitions 17: 126–148. http://dx.doi.org/10.1016/j.eist.2014.12.002.

80 References

Wang, Lu, and Ming Chen. 2013. “Policies and Perspective on End-of-Life Vehicles in China.” Journal of Cleaner Production 44: 168–176. https://doi.org/10.1016/j.jclepro.2012.11.036.

Wang, Zhongying, Haiyan Qin, and Joanna I. Lewis. 2012. “China’s Wind Power Industry: Policy Support, Technological Achievements, and Emerging Challenges.” Energy Policy 51: 80–88. https://doi.org/10.1016/j.enpol.2012.06.067.

Weight, David. 2018. “Challenges for a Technology Enabling Metal.” Advanced Automotive Battery Conference. San Diego, CA. June 2018.

Wiser, Ryan, and Mark Bolinger. 2019. 2018 Wind Technologies Market Report. U.S. Department of Energy Office of Energy Efficiency and Renewable Energy. DOE/GO-102019-5191. https://doi.org/10.2172/1559881.

———. 2017. 2016 Wind Technologies Market Report. Washington, D.C.: U.S. Department of Energy. DOE/GO-10216- 4885. https://doi.org/10.2172/1375677.

———. 2016. 2015 Wind Technologies Market Report. Washington, D.C.: U.S. Department of Energy. DOE/GO-10216- 4885. https://doi.org/10.2172/1312474.

World Bank. n.d. “GDP Growth: Annual %.” GDP database accessed August 8, 2019. https://data.worldbank.org/indicator/NY.GDP.MKTP. KD.ZG?end=2017&locations=CN&start=2013.

Wright, Maury. 2016. “Popping the Charts: Strategies Unlimited Unveils SSL Market Data.” LEDs Magazine, April 28, 2016. https://www.ledsmagazine.com/leds-ssl-design/ modular-drivers/article/16696076/popping-the-charts- strategies-unlimited-unveils-ssl-market-data-magazine.

Yano (Yano Research Institute Ltd.). 2017. Global In-Vehicle LiB Market: Key Research Findings 2017. October 20, 2017. https://www.yanoresearch.com/press/pdf/1750.pdf.

Zamorano, Alejandro. 2017. Global EV Trends and Forecast. Bloomberg New Energy Finance. April 18, 2017. https://efiling.energy.ca.gov/GetDocument.aspx?tn=217132.

81 Appendix

Appendix Contents

Crosscutting ...... 84

Additional Benchmark Data: Trade Trends—Impact of Currency Rate Fluctuations ...... 84

Wind Turbines ...... 87

Manufacturing Capacity and Production Origins and Explanations ...... 87

Additional Benchmark Data: Trade Trends—Expanded View ...... 89

Crystalline Silicon Photovoltaic Modules ...... 91

Manufacturing Capacity and Production Origins and Explanations ...... 91

Additional Benchmark Data: Trade Trends—Expanded View ...... 92

Light-Emitting Diode Packages ...... 95

Manufacturing Capacity and Production Origins and Explanations ...... 95

Additional Benchmark Data: Trade Trends—Expanded View ...... 97

Lithium-Ion Battery Cells ...... 98

Manufacturing Capacity and Production Origins and Explanations ...... 98

Additional Benchmark Data: LIB Cell Trade Trends—Expanded View ...... 100

82 Appendix

Appendix List of Figures and Tables

Exchange Rates for Benchmarked Economies Relative to the U.S. Dollar, 2012–2017 ...... 84

Exchange rate and balance of trade, United States with China, 2014–2016 ...... 86

Key Wind Supply Chain Manufacturers by Location, 2016 ...... 87

Wind turbine generator set (nacelles and blades) trade, 2014–2016 ...... 89

Key PV Supply Chain Manufacturers by Location, 2016 ...... 91

PV module trade, 2014–2016 ...... 92

Key LED Supply Chain Manufacturers by Location, 2016 ...... 95

LED package trade, 2014–2016 ...... 97

Key LIB Supply Chain Manufacturers by Location, 2016 ...... 98

LIB cell trade, 2014–2016 ...... 100

U.S. imports of LIB cells for use in EVs, 2015–2016 ...... 101

83 Appendix

Appendix Crosscutting Additional Benchmark Data: Trade Trends—Impact of Currency Rate Fluctuations Exchange Rates for Benchmarked Economies Relative to the U.S. Dollar, 2012–2017

Economy Currency 2012 2013 2014 2015 2016 2017

Brazil Brazilian real 1.95 2.16 2.35 3.34 3.48 3.19

Canada Canadian 1.00 1.03 1.10 1.28 1.32 1.30 dollar

China Chinese yuan 6.31 6.15 6.16 6.28 6.64 6.76

Denmark Euro 1.29 1.33 1.33 1.11 1.11 1.13

Germany Euro 0.78 0.75 0.75 0.90 0.90 0.88

India Indian rupee 53.37 58.51 61.00 64.11 67.16 65.07

Japan Japanese yen 79.82 97.60 105.74 121.05 108.66 112.10

Malaysia Malaysian 3.09 3.15 3.27 3.90 4.14 4.30 ringgit

Mexico Mexican peso 13.15 12.76 13.30 15.87 18.67 18.88

Korea South Korean 1126.16 1094.67 1052.29 1130.96 1159.34 1129.04 won

Taiwan New Taiwan 29.56 29.68 30.30 31.74 32.23 30.40 dollar

UK British pound 0.63 0.64 0.61 0.65 0.74 0.78

US United States 1.00 1.00 1.00 1.00 1.00 1.00 dollar

Annual average exchange rates expressed in units of local currency (to the left) per US dollar (nominal). Source: Federal Reserve Bank of St. Louis n.d. and its imports become less expensive, leading to a decrease Over the period, currency rate fluctuations were relatively in exports and increase in imports in the economy with the large, moving up and down by double digit percentage stronger currency. Some studies have shown that empirical points in some cases. For example, according to a study data are consistent with this theory for the economy writ sponsored by the International Monetary Fund, the dollar large (Rose 1991). For example, Leigh et al. (2017) and appreciated by more than 10%, while the euro and yen Campa (2000) find that a 10% depreciation in a currency depreciated by more than 10% and 30%, respectively, leads to a 1.5% increase in net exports; this depreciation between 2012 and 2015 (IMF 2015). Exchange rates for India, effect on trade varies from economy to economy because China, Taiwan, and South Korea showed the most stability the relative strength of the other currencies vary. There are over the period. economists who disagree with these findings.

Macroeconomic theory indicates that as a currency While economies in aggregate reacted to currency appreciates in an economy, its exports become more fluctuations in these studies and several others in a expensive relative to an economy with a weaker currency statistically significant way, when disaggregated by activity

84 Appendix

(e.g., manufacturing, sales, professional services), this relationship is not necessarily maintained (Leigh et al. 2017). In addition, for products with extensive global supply chains, the cost of production can be impacted by the exchange rates of economies (i.e., those providing supply chain intermediates) that are not direct trading partners, reducing the influence of exchange rate changes on trade flows (Leigh et al. 2017). Trade agreements, tariffs, and preferences of importers (such as preferring a specific brand or economy of origin) can impact demand, product quality or availability, and government subsidies. Any number of nonmonetary factors can also impact imports and exports of specific commodities.

For these reasons, estimating the impact of currency rate variability on imports and exports of the four clean energy technologies is challenging. In addition, in total the four technologies represent just a small fraction of total trade in the benchmarked economies and each clean energy technology represents a different industry within the manufacturing sector, supported by its own global supply chain. The figure below illustrates the difficulty of attributing trade trends for a specific product to changes in currency rates.

85 Appendix

Exchange rate and balance of trade, United States with China, 2014–2016

$500,000 6.7

$400,000 6.6

$300,000 6.5 $200,000

6.4 $100,000

$0 6.3

2014 2015 2016 -$100,000 6.2 Yuan per USD (Nominal) Balance of Trade (US$(2014)) -$200,000 6.1 -$300,000

6.0 -$400,000

-$500,000 5.9

Net U.S. LED Exports to China in thousands

Net U.S. Polysilicon Exports to China in thousands Total U.S. Balance of Trade with China in millions

Exchange Rate (yuan per dollar)

Annual average exchange rate (lime green line) in yuan per U.S. dollar (nominal). U.S. Total BOT with China (green line) in US$(2014) millions. U.S. LED package BOT with China (red line) in US$(2014) thousands. U.S. Polysilicon (poly-Si) BOT with China (yellow line) in US$(2014) thousands.

The figure compares the balance of trade between exports from the United States declined, yet LED the United States and China for two different net exports increased, suggesting that currency rate commodities—LED chips and packages, and fluctuation is not the only driver affecting trade. For polysilicon—and exchange rate trends between 2014 example, in this period, polysilicon trade was also and 2016. As the U.S. dollar became more expensive impacted by Chinese tariffs on polysilicon imports for Chinese importers, the total U.S. balance of trade from the United States. declined. Over the period, polysilicon (poly-Si) net

86 Manufacturing Wind Turbines Appendix

Wind Turbines Manufacturing Capacity and Production Origins and Explanations Key Wind Supply Chain Manufacturers by Location, 2016

Manufacturing Nacelles Blades Towers Generators Location

Enercon, GE, Nordex Enercon, LM Enercon, Torrebras/Windar ABB, Enercon, GE, WEG Brazil Group, SGRE, Vestas, WEG

Canada LM, SGRE CS Wind, Marmen

CCWE, CSIC Aeolon, Dawntine, Jilin AVIC Hongbo Windpower ABB, CRRC Yongji Haizhuang, DEC, Chongtong Chengfei Equipment, CRRC Tongli Electric, CRRC Zhuzhou Envision, GE, Goldwind, New Material, LM, Steel, Dajin Heavy Industry Electric, Dongfeng Mingyang, SANY, Sinoma, Sino-wind, Corporation, Fuchuan Electric, Nanjing Turbine SEwind, SGRE, Sinovel, Sunrui, TPI, ZFLZ, Yifan New Energy, Gansu & Electric Machinery China TYHI, United Power, Zhuzhou Times New Jiugang, Qingdao Tianneng Group, Shanghai Vestas, Windey, XEMC Material Technology Electric Power Engineering Electric, XEMC Xiangtan Machinery, Qingdao Electric-DFIG, XEMC Wuxiao, Sinohydro Bureau Xiangtan Electric-PMG 4, TITAN, TSP

MHI Vestas, SGRE, LM, MHI Vestas, SGRE, TITAN, Welcon Denmark Vestas Vestas

Enercon, GE, Nordex Carbon Rotec, Enercon, Ambau, Enercon, Max Bogl Enercon, Loher GmbH, Germany Group, Senvion, SGRE Nordex Group, Senvion, Wind AG VEM, Vensys, Vestas SGRE, Vestas

GE, Global Wind Power, INOX, LM, SGRE, Suzlon, Barakath Engineering, ABB, GE, Regen INOX, Leitwind, Nordex Vestas, WWI Fedders Lloyd, INOX, Powertech, Suzlon, WWI India Group, REGEN, SGRE, Nordex Group, Suzlon, Suzlon, Vestas, WWI Windar

Japan Hitachi Hitachi

Mexico TPI Trinity, Windar Renovables Potencia Industrial

Doosan, Unison Dongkuk S&C, Speco, South Korea Unison

United SGRE MHI Vestas, SGRE Kingdom

Broadwind, Gestamp, United States GE, SGRE, Vestas LM, SGRE, TPI, Vestas Ingeteam, SGRE, Vestas Marmen, Trinity, Vestas

Modern wind turbine manufacturing originated in Europe Western Europe slowed, and new growth has emerged in the 1980s and 1990s, and it continued to have a strong elsewhere in the world, specifically North America and Asia. presence in Germany and to a lesser extent in Denmark More recently however, the offshore sector has begun to and Spain. Generally, European manufacturing capacity gain market share in Europe and Asia and has grown its levels stabilized and, in some cases, eroded as demand in manufacturing footprint.

87 Appendix Manufacturing Wind Turbines

Beginning in the early to mid-2000s, the United States, for which they had a comparative advantage. For example, China, India, and Brazil began increasing wind turbine Mexico had relatively lower labor costs and close proximity manufacturing capabilities. In the United States, to U.S. markets (Make 2015b). manufacturing capacity was developed or repurposed Through 2016, global wind power manufacturing was driven from complementary industries as an increasing number of by a combination of historical demand, projected future U.S. states adopted renewable energy portfolio standards demand, and existing complementary production and (RPS), and steady extensions of a federal production fabrication industries. Europe and Germany in particular had tax credit (PTC) supported robust growth, typically 5–10 robust wind power demand in the past, and a sophisticated gigawatts (GW) per year with a peak of 13 GW installed in manufacturing sector that allowed Europe to be the 2012 (Wiser and Bolinger 2016). The first components to primary source of wind turbine components through the be manufactured in the United States were technologically early 2000s. Given their existing infrastructure and skill simple but large and relatively costly to transport, including sets, some European manufactures were able to remain a towers and blades. Nacelle assembly capacity was followed continued source of supply for local European demand, to by production of some subcomponents within the nacelles serve economies with somewhat variable demand (e.g., (e.g., bearings, gearboxes, electrical components), which the United States) that might not always be met with came online in anticipation of sustained North American domestic production, and to serve markets otherwise too wind energy growth. By the end of the period, U.S. small to justify local manufacturing capacity (e.g., Africa, the production included 2 megawatt (MW) and 3 MW wind Middle East, and Central and South America). In contrast, turbines with capabilities for blades, towers, generators, manufacturing capacity in the United States, China, and and gearboxes (Fullenkamp and Holody 2014). Accordingly, Brazil was established primarily due to high transport costs U.S. domestic content estimates included 80%–85% for for large wind turbine components, and the relatively large towers, 50%–70% for blades and hubs, and more than 85% quantities of current and anticipated domestic or regional for nacelle assembly (Wiser and Bolinger 2016). At the same demand. Moreover, in each of these economies, there was time, much of the nacelle internals were still imported and some ability to leverage existing manufacturing synergies domestic content for wind turbine equipment as a whole within existing large and diverse industrial sectors. was estimated at approximately 40% in 2012 (Wiser and Bolinger 2016). Along with imports for nacelle internals generally, a persistent gap in the U.S. supply chain was the large structural castings used in the nacelle. U.S.-based manufacturers tended to import these castings from Asia and South America (Fullenkamp and Holody 2014).

Over the period examined, China was the largest manufacturer of wind power equipment in the world in terms of production capacity and output, supported predominately by its domestic demand (MAKE 2015a). A number of policies in the mid-2000s supported the establishment of a local Chinese wind power supply chain (Wang, Qin, and Lewis 2012).

Brazil has also observed the development of a sizable wind market built upon reverse auctions. In part as a function of increased demand, as well as strict domestic content requirements for wind equipment, blade, tower, hub, and nacelle assembly facilities were developed in Brazil. India also maintained gigawatt scale manufacturing facilities serving primarily domestic demand (Make 2015b). Canada, Mexico, and an array of other European and Asian economies also maintained some degree of manufacturing capacity; however, these more isolated pockets of manufacturing tended to focus on specific components

88 Manufacturing Wind Turbines Appendix

Wind Turbine Supply Chain, 2014-2016 Market and Trade Trends

Wind 1. 2014-2016 Wind 2. Wind 3. 2014 - 2016 Wind 4. 2014-2016 App/Wind 2. WInd - FYI calc for wind 2 FYI calc for wind 4 AdditionalWind Production and Benchmark2014-2016Production Data:Wind De Trademand, Trends—ExpandedWind Genset Gensets Detailed View BOT Demand (% contributi.. Demand Trends Production, Mfg Capa.. Import/Export/BOT T.. Wind turbine generator set (nacelles and blades) trade, 2014–2016

2014 2015 2016 Brazil Brazil $0M $0M $0M Canada $0M $0M $0M China $0M $0M $0M Denmark -$22M $0M $0M Germany $0M $0M $0M India -$29M $0M $0M Japan $0M $0M $0M Malaysia $0M $0M $0M Mexico $0M $0M $0M South Korea $0M $0M $0M Taiwan $0M $0M $0M United Kingdom $0M $0M $0M United States -$207M -$26M -$33M China Brazil $0M $0M $0M Canada $1M $2M $1M China $0M $0M $0M Denmark -$8M $0M $0M Germany -$1M $1M -$1M India $0M $1M $3M Japan $0M -$1M $8M Malaysia $0M -$0M $0M Mexico $74M $45M $196M South Korea $0M $0M $7M Taiwan $0M $0M $0M United Kingdom $17M $22M $14M United States $0M $27M $36M Denmark Brazil $22M $0M $0M Canada $36M $98M $0M China $8M $0M $0M Denmark $0M $0M $0M Germany $471M $422M $204M India $0M $0M $0M Japan $12M $23M $2M Malaysia $0M $0M $0M Mexico $111M $0M $118M South Korea $25M $0M $35M Taiwan $0M $0M $13M United Kingdom $403M $148M $164M United States $50M $90M $15M Germany Brazil $0M $0M $0M Canada $245M $217M $82M China -$1M -$1M $1M Denmark -$471M -$422M -$204M Germany $0M $0M $0M India $0M -$1M $0M Japan $40M $31M $49M Malaysia $0M $0M $0M Mexico $0M $1M $21M South Korea $0M $0M $8M Taiwan $0M $15M $19M United Kingdom $39M $30M $93M United States $0M $23M -$1M United Brazil $207M $26M $33M Canada $319M $90M $13M States China $0M -$27M -$36M Denmark -$50M -$90M -$15M Germany $0M -$23M $1M India $0M -$1M $0M Japan $1M $2M $4M Malaysia $0M $0M $0M Mexico $38M $23M $9M South Korea $12M $13M $0M Taiwan $0M $0M $0M United Kingdom $14M $12M $2M United States $0M $0M $0M

Imports Exports Imports Exports Imports Exports

Bars show trade (exports positive and imports negative) and numerical value indicates balance of trade (BOT, exports minus imports) for wind turbine generator sets (nacelles and blades) by economy. All data are in US$(2014).

89 Appendix Manufacturing Wind Turbines

Denmark, Germany, and, increasingly, China were the top exporters of wind turbine generator sets (nacelle and blades) over the period. The top three destinations for exports from Denmark were Germany, the United Kingdom, and Mexico. Germany exported primarily to Canada and the United Kingdom. The majority of Chinese exports went to Mexico.

Between 2014 and 2016, a few economies contributed to the trade of wind turbine generator sets. Specifically, Denmark, Germany, and China were critical exporters; Mexico, the United Kingdom, and the United States were the largest importing economies. Trading partners for the five key wind turbine economies remained relatively consistent over the period.

In 2016, the top three destinations for exports of generator sets from Denmark were Germany ($200 million), the United Kingdom ($160 million), and Mexico ($120 million). Germany exported $93 million, $82 million, and $49 million to the United Kingdom, Canada, and Japan, respectively. China exported $196 million of generator sets to Brazil in 2016.

U.S. exports to Canada and Brazil dropped dramatically over the period from $319 million to $16 million and from $207 million to $33 million, respectively. Germany’s and Denmark’s exports to Canada also declined over the period.

90 Manufacturing c-Si PV Modules Appendix

Crystalline Silicon Photovoltaic Modules Manufacturing Capacity and Production Origins and Explanations Key PV Supply Chain Manufacturers by Location, 2016

Manufacturing PV Module PV Cell PV Wafer Polysilicon Location

Trina, Jinko, Canadian Trina, JA Solar, Yingli, GCL, LDK, ReneSolar, China GCL, REC Solar, JA Solar, Yingli Jinko, Canadian Solar Yingli, Jinko

SolarWorld, PV Crystalox Germany SolarWorld SolarWorld Wacker Solar

United States SolarWorld, Suniva SolarWorld, Suniva Panasonic Hemlock, REC, Wacker

LG, Hyundai, Shinsung South Korea LG, Hyundai Nexolon, Woongjin Energy OCI, Hanwha Solar Energy

Malaysia Hanwha, Panasonic SunPower AUO, Comtec

Philippines SunPowerq SunPower

Gigastorage, Green Energy Motech, Gintech, Taiwan Neo Solar Technology, Sino-American Neo Solar Silicon Products a. SunPower announced in August 2016 its intent to close its Philippines module assembly capacity and relocate a portion of it to Mexico (http://www.pv-tech.org/news/sunpower-streamlining-project-development-focus-and-closing-module-assembly)

Some vertical integration existed during the period across The buildup of manufacturing capacity through 2011 was the PV manufacturing industry between wafer, cell, and driven by expectations of continued strong demand growth, module production, with integrated manufacturers citing and especially in China by local and provincial subsidies and lower “in-house” production costs as compared to those investment supports (Quitzow 2015). achieved when sourcing materials and components from The resulting overcapacity drove large price reductions third parties (Chase et al. 2016). Several companies shown between 2008 and 2012, exacerbated by a slowdown in owned manufacturing assets in economies other than their global demand between 2011 and 2012. However, with robust headquarters locations (e.g., Wacker and SolarWorld were demand returning and pricing stabilizing, PV manufacturer headquartered in Germany, but owned manufacturing capital expenditures (related to capacity additions) facilities in the United States; Panasonic is headquartered rebounded again in 2014 and 2015.26 in Japan but owned manufacturing facilities in Malaysia and the United States).

26 Bloomberg L.P. 2016. “PV Operating Margins”. Bloomberg Terminal. Accessed Aug 1 ,2017.

91 Appendix Manufacturing c-Si PV Modules

Additional Benchmark Data: Trade Trends—Expanded View PV module trade, 2014–2016

PV Module 2014 2015 2016 Canada $44M $40M $14M China China Germany $275M $270M $226M India $59M $190M $334M Japan $4,540M $3,586M $2,684M Malaysia -$27M $28M $4M Mexico $8M $8M $25M South Korea $54M $37M $59M Taiwan -$1M -$2M -$2M United Kingdom $653M $587M $196M United States $1,579M $1,660M $1,457M Canada Germany China -$275M -$270M -$226M Germany $0M $0M $0M India -$8M -$4M -$2M Japan -$20M -$22M -$22M Malaysia -$261M -$244M -$105M Mexico -$4M -$3M -$1M South Korea -$37M -$23M -$78M Taiwan $0M $0M $3M United Kingdom $95M $82M $22M United States $72M $366M $95M Canada Japan China -$4,540M -$3,586M -$2,684M Germany $20M $22M $22M India -$10M $9M $19M Japan $0M $0M $0M Malaysia -$375M -$318M -$164M Mexico $0M $0M $0M South Korea -$285M -$68M $58M Taiwan -$1M $6M $0M United Kingdom -$61M $14M $24M United States -$1M $76M $93M Canada Malaysia China $27M $48M -$4M Germany $261M -$2M $105M India $7M $0M $45M Japan $375M -$17M $164M Malaysia $0M $0M $0M Mexico $0M $0M $0M South Korea -$7M -$48M $7M Taiwan $13M $7M $1M United Kingdom $126M -$1M $6M United States $0M $0M $0M Canada South China -$54M -$37M -$59M Korea Germany $37M $23M $78M India $9M $1M $1M Japan $285M $68M -$58M Malaysia $7M -$1M -$7M Mexico $0M $0M $0M South Korea $0M $0M $0M Taiwan $0M $1M -$1M United Kingdom $1M $0M $0M United States $107M $388M $1,265M Canada $14M $3M $2M Taiwan China $1M $2M $2M Germany $0M $0M -$3M India $0M $0M $0M Japan $1M -$6M $0M Malaysia -$13M $0M -$1M Mexico $0M $0M -$1M South Korea $0M -$1M $1M Taiwan $0M $0M $0M United Kingdom $0M $0M $0M United States $738M $305M $200M Canada -$12M -$47M -$82M United China -$1,579M -$1,657M -$1,457M States Germany -$72M -$366M -$95M India -$27M -$12M $11M Japan $1M -$85M -$93M Malaysia $0M $16M $0M Mexico -$505M -$902M -$833M South Korea -$107M -$388M -$1,265M Taiwan -$738M -$305M -$200M United Kingdom $28M $0M $17M United States $0M $0M $0M Imports Exports Imports Exports Imports Exports

Bars show trade (exports positive and imports negative) and numerical value indicates balance of trade (BOT, exports minus imports) for PV modules by economy. All data are in US$(2014).

In the benchmarked economies, the balance of trade (exports trade were observed for Japan and the United States; minus imports) remained relatively stable for PV modules, China’s PV module exports to the United States and Japan with the exception of Japan (imports from China declined) declined over the period. At the same time, U.S. imports and the United States (imports from South Korea increased). from South Korea and the rest of the world increased.

China maintained its position as the largest PV module exporter over the period. The biggest changes in PV module

92 Manufacturing c-Si PV Modules Appendix

PV cell trade, 2014–2016

2014 2015 2016 Canada $158M $30M $30M China China Germany -$13M -$11M -$4M India $373M $1,203M $2,093M Japan -$481M -$101M -$174M Malaysia -$68M $195M $61M Mexico $22M $20M $1M South Korea $358M $253M $391M Taiwan -$1,832M -$1,125M -$896M United Kingdom $144M $130M $43M United States Canada -$2M $7M $5M Germany China $13M $11M $4M Germany $0M $0M $0M India $1M $3M $2M Japan -$8M -$6M -$5M Malaysia -$10M -$8M -$1M Mexico $2M $12M $2M South Korea $2M $11M $1M Taiwan -$333M -$400M -$381M United Kingdom $330M $346M $161M United States Canada Japan China $481M $101M $174M Germany $8M $6M $5M India $4M $21M $30M Japan $0M $0M $0M Malaysia -$19M $15M $4M Mexico -$25M $106M $60M South Korea $108M $332M $375M Taiwan -$798M -$404M -$237M United Kingdom $0M United States -$23M -$6M Canada $0M $8M $17M Malaysia China -$90M $22M -$61M Germany -$7M $16M $1M India $0M $141M $174M Japan -$6M $23M -$4M Malaysia $0M $0M $0M Mexico $0M $311M $270M South Korea -$4M $194M $134M Taiwan -$5M $13M $11M United Kingdom $0M $9M United States $23M Canada South China -$358M -$253M -$391M Germany -$2M -$11M -$1M Korea India $3M $0M $0M Japan -$108M -$332M -$375M Malaysia -$19M -$178M -$134M Mexico $19M $91M $0M South Korea $0M $0M $0M Taiwan -$100M -$105M -$65M United Kingdom $20M $14M $10M United States Canada $122M $180M $62M Taiwan China $1,832M $1,125M $896M Germany $333M $400M $381M India $41M $46M $50M Japan $798M $404M $237M Malaysia -$48M $112M -$11M Mexico $27M $15M $16M South Korea $100M $105M $65M Taiwan $0M $0M $0M United Kingdom $74M $143M United States $60M Canada $6M $24M $21M United China $18M $10M $1M Germany $2M $16M -$9M States India $0M $1M $0M Japan $23M -$3M $6M Malaysia -$54M $4M -$23M Mexico $2M $0M $0M South Korea $0M $0M $0M Taiwan -$7M -$28M -$60M United Kingdom $1M $1M $1M United States Imports Exports Imports Exports Imports Exports

Bars show trade (exports positive and imports negative) and numerical value indicates balance of trade (BOT, exports minus imports) for PV cells by economy. All data are in US$(2014).

In the benchmarked economies, the balance of trade between 2014 and 2016. The largest changes in PV cell (exports minus imports) remained relatively stable for PV trade were observed for China and the Taiwan; China’s PV cells, with the exception of China (exports to India increased cells exports to Taiwan decreased, while exports to India sharply) and Taiwan (exports to China and Japan declined). increased. Taiwan exports to China and Japan declined over the period. China maintained its position as the largest PV cell exporter

93 Appendix Manufacturing c-Si PV Modules

Polysilicon trade, 2014–2016

2014 2015 2016 Canada -$3M -$1M China China $0M Germany -$698M -$614M -$543M India $87M $81M $36M Japan $9M $32M -$4M Malaysia -$1M -$74M -$127M Mexico $0M $0M $0M South Korea -$735M -$795M -$1,047M Taiwan -$235M -$235M -$245M United Kingdom United States -$389M -$238M -$181M Canada -$1M -$2M $0M Germany China $698M $614M $543M Germany $0M $0M $0M India $0M $0M $0M Japan $178M $128M $132M Malaysia $0M $0M $2M Mexico $0M $0M $0M South Korea $43M $32M $64M Taiwan $172M $117M $135M United Kingdom -$12M United States -$223M $50M Canada $0M Japan China -$9M -$32M Germany -$178M -$128M -$132M India $0M $0M $0M Japan $0M $0M $0M Malaysia -$1M -$11M $0M Mexico $0M $0M $0M South Korea $37M $25M $31M Taiwan $100M $26M $55M United Kingdom United States -$830M -$632M -$944M Canada $0M $0M $0M Malaysia China $1M $74M $127M Germany $0M $0M -$2M India $0M $0M $0M Japan $1M $11M $0M Malaysia $0M $0M $0M Mexico $0M $0M $0M South Korea -$4M $0M -$2M Taiwan -$26M -$53M -$57M United Kingdom United States -$21M -$9M -$8M Canada South China $735M $795M $1,047M Germany -$43M -$32M -$64M Korea India $0M $0M $0M Japan -$37M -$25M -$31M Malaysia $4M $0M $2M Mexico $0M $0M $0M South Korea $0M $0M $0M Taiwan $325M $238M $181M United Kingdom United States -$133M -$142M -$110M Canada $0M $0M $62M China $235M $235M $884M Taiwan Germany -$172M -$117M $251M India $0M $0M $50M Japan -$100M -$26M $19M Malaysia $26M $53M $47M Mexico $0M $0M $16M South Korea -$325M -$238M -$117M Taiwan $0M $0M $0M United Kingdom United States -$230M -$221M -$244M Canada $0M $0M $0M United China $389M $238M $181M Germany $223M -$50M -$2M States India $0M $0M $0M Japan $830M $632M $944M Malaysia $21M $9M $8M Mexico $0M $1M $1M South Korea $133M $142M $110M Taiwan $230M $221M $300M United Kingdom $6M $6M $6M United States Imports Exports Imports Exports Imports Exports

Bars show trade (exports positive and imports negative) and numerical value indicates balance of trade (BOT, exports minus imports) for polysilicon by economy. All data are in US$(2014).

Over the period, the balance of trade (exports minus relatively steady value of polysilicon exports despite a dip imports) for polysilicon shifted among the benchmarked in 2015. Between 2014 and 2016, the United States, South economies, led by China, South Korea, Taiwan and the Korea, and Germany exported polysilicon to Taiwan to avoid United States. Chinese tariffs (although still exporting some polysilicon directly to China) and Taiwan increased exports to China. In 2016, China continued to be the largest net importer The United States increased exports to Japan to offset some of polysilicon. The United States was the second largest of the loss of exports to China over the period. exporter of polysilicon (after Taiwan), maintaining a

94 Manufacturing LED Packages Appendix

Light-Emitting Diode Packages Manufacturing Capacity and Production Origins and Explanations Key LED Supply Chain Manufacturers by Location, 2016

Manufacturing Packagea Chipb Substrate2 Location

China ChangFang, HongliZhihui Electech ETI, Epistar, HC Aurora Sapphire, Crystaland, Crystal (Honglitronic), Jufei Opto, Semitek, Lextar, Nantong, Applied Technology, Crystal Optech, ECBO, Mulinsen (MLS), NationStar, San’an, Tongfang Fujian Crystal, GAPSS, J-Crystal, JeShine, Refond Opto Nanjing J-Crystal Photoelectric, NJC, SinoNitride, SinoNitride (Sinopat), TDG Core, Unisem New Material Technology- Youzhong

Germany Osram OS

Japan Citizen, Nichia, Sharp, Stanley, Nichia Kyocera, Namiki Toyoda Gosei

South Korea LG Innotek, Lumens, Samsung, LG, Samsung, Seoul Vyosis Hansol Technics, Iljin Display, LGS, Unisem, Seoul Semiconductor Sapphire Technology, SSLM

Taiwan AOT, Everlight, Harvatek, Epistar, Lextar AcepluxOptotech, AimCore, Crystal Kingbright, Lextar, Lite-On, Applied Technology, Crystalwise, Unity Opto Lucemitek, Phecda, Procrystal, Rigidtech, Tera Xtal, TXC Quartz

United States CREE, Lumileds Cree Rubicon

a. Manufacturing location indicates company headquarters, not equipment location. b. Manufacturing location indicates equipment location. List does not include vertically integrated companies.

Light-emitting diodes (LEDs) were first used as an electronic more than 50% of market share (Su 2014). In 2010, lower- component in 1968 (Schubert 2003). Early LEDs were used priced products helped Taiwanese, South Korean, and as indicators in electronic devices and in displays. As LEDs Chinese firms gain market share, and Japan for the first time improved with longer lifetimes, smaller size, and lower garnered less than 50%. energy consumption, uses expanded to include general South Korean advances were led by Samsung’s and LG’s lighting, consumer electronics, displays, signs, automotive, development of LCD televisions with LED backlighting. and other uses (Mukish and Virey 2014; Wright 2016). South Korean companies further leveraged their brand The growth of the LED market was driven by increased advantage and, with the government support, adopted electricity costs coupled with the worldwide adoption of vertically integrated operations, leading to further market standards and regulations for energy efficiency, including share gains. the phase-out of incandescent lighting (En.Lighten n.d.). Taiwan’s industry originally grew from international Furthermore, Navigant Research (2015) reported that investment that established the first packaging plants. “LED prices have declined to a point where this type of Taiwan formed a collaborative in the late 1990s which lighting is becoming the economical choice in almost every focused on integrating the entire LED value stream into application.” three main industrial clusters that created a pool of expertise While LEDs were originally invented in the United States, in the LED and opto-electronic fields. As a consequence, Shuji Nakamura from Japan’s Nichia Corporation made Taiwan companies reduced product development time and enormous scientific advances in LEDs in the early 1990s and costs and slowly gained market share in both LED packages catapulted Japan into a leading market position. Prior to and sapphire substrate. 2010, Japan maintained this technical leadership, capturing

95 Appendix Manufacturing LED Packages

China’s LED industry grew significantly between 2005 and 2007. LEDs Magazine identified only one Chinese firm in 2005, with the number jumping to 51 by 2007 (Sanderson and Simons 2014). By 2009, there were 100 wafer companies, but by 2016 industry consolidation reduced the total number to about 28 with meaningful levels of revenue (Mukish and Virey 2017). Government subsidies for production equipment, land, leasing, and taxation aided domestic firms (Schubert 2003; Mukish and Virey 2014).

LED package production in Japan and South Korea constituted more than 55% of total global package output in 2014, with China and Taiwan together contributing another 25%. By 2016, China was the highest-volume package producer with a 30% share, with Japan, South Korea, and Taiwan together contributing 54%. Sapphire substrate production was concentrated in Taiwan and China, accounting 35% and 29% of global production in 2016.

96 Manufacturing LED Packages Appendix

Additional Benchmark Data: Trade Trends—Expanded View LED package trade, 2014–2016

2014 2015 2016 Brazil China Canada $1M $4M $2M China $0M $0M $0M Denmark $13M $14M $7M Germany $139M $195M $140M India $38M $94M $140M Japan -$1,923M -$1,680M -$1,443M Malaysia -$837M -$794M -$867M Mexico $1M $9M $8M South Korea $211M -$272M $75M Taiwan -$1,595M -$2,901M -$1,319M United Kingdom $72M $66M $48M United States -$402M -$381M $187M Brazil $3M $3M $4M Japan Canada $5M $5M $12M China $1,923M $1,680M $1,443M Denmark $0M $0M $0M Germany $51M $41M $71M India $12M $12M $12M Japan $0M $0M $0M Malaysia -$156M -$150M -$125M Mexico $22M $20M $24M South Korea $16M $14M $76M Taiwan $184M $113M $65M United Kingdom $16M $17M $17M United States $361M $377M $381M Brazil $0M $0M $0M South Canada $0M $0M $0M Korea China -$211M $272M -$75M Denmark $1M $1M $0M Germany $11M $26M $25M India $3M $6M $7M Japan -$16M -$14M -$76M Malaysia $0M $0M $0M Mexico $0M $0M $0M South Korea $0M $0M $0M Taiwan -$233M -$143M -$68M United Kingdom United States $34M $32M $36M Brazil $2M $1M $2M Taiwan Canada $1M $1M $1M China $1,595M $2,901M $1,319M Denmark $1M $0M $0M Germany -$2M $13M $6M India $6M $4M $2M Japan -$184M -$113M -$65M Malaysia $86M $52M $48M Mexico $0M $6M $4M South Korea $233M $143M $68M Taiwan $0M $0M $0M United Kingdom United States $37M $36M $36M Brazil $15M $9M $9M United Canada $51M $49M $43M States China $402M $381M -$187M Denmark $0M $0M $0M Germany $57M $43M $30M India $1M $2M $5M Japan -$361M -$377M -$381M Malaysia -$269M -$251M -$219M Mexico $346M $434M $427M South Korea -$34M -$32M -$36M Taiwan -$37M -$36M -$36M United Kingdom $9M $12M $13M United States

Imports Exports Imports Exports Imports Exports

Bars show trade (exports positive and imports negative) and numerical value indicates balance of trade (BOT, exports minus imports) for LED packages by economy. All data are in US$(2014).

From 2014 to 2016, China was the largest net importer of Taiwan. In 2014, South Korea, Japan, China, and Taiwan were LED packages and Japan, Malaysia, and Taiwan were the key exporters of LED packages. By 2016, China had become largest net exporters (primarily to China). In the benchmarked the largest LED package trading partner with the other key economies, the balance of trade (exports minus imports) LED package economies. Over the period, China primarily remained relatively stable, with a few exceptions including imported LED packages and chips from Taiwan, Japan, and China, Taiwan, the United States, and South Korea. Malaysia. Top destinations for China’s LED packages and chips were the United States, South Korea, and Taiwan. Top U.S. From 2014 to 2016, the largest changes in trade flow of LED import partners in 2016 included China, Japan, and Malaysia. chips and packages were observed in China, Japan, and

97 Appendix Manufacturing LIB Cells

Lithium-Ion Battery Cells Manufacturing Capacity and Production Origins and Explanations Key LIB Supply Chain Manufacturers by Location, 2016

Manufacturing Cell Cathode Anode Separator Electrolyte Location

China BAK, BYD, Jinhe, Reshine, ShanShan, Fengfan, Jinhui Cap-chem, Guotai- CATL, Guoxuan, ShanShan Shenzhen Huarong, ShanShan Op-timumNano Zhnagjiang

Japan AESC, Nichia, Nihon Hitachi, Mitsubishi AsahiKASEI, Toray, Mitsubishi Chemical, Panasonic, Kagaku Sangyo, Chemical, Nippon Sumito-mo, Ube Mitsui, Tomiyama, Pri-mearth Sumi-tomo, Toda Carbon Indus-tries, W-Scope Ube Industries, Kogyo,

South Korea LG Chem, Ecopro, L&F LGC, Posco, Tonen Panax-Etec, Sam-sung SDI Umicore Samsung SDI Soulbrain

United States LG Chem, Celgard Panasonic, Tesla

Lithium-ion batteries (LIBs) were developed in the 1990s to Japan’s LIB cluster grew from sustained investments in power consumer electronics. Electric vehicle (EV) original LIB technology by consumer electronics companies in equipment manufacturers (OEMs) have since adopted LIBs the 1990s. The Japanese government bolstered private as a technology of choice for the rechargeable batteries sector investments with research and development (R&D) used in their vehicles. funding and low-cost capital to establish manufacturing plants. Japan made these investments despite the long In 2016 electrified vehicles (including pure electric, plug-in commercialization cycle of Li-ion technologies and the low hybrid, and hybrid drive vehicles) constituted nearly 2.6% of returns on the LIB business because the technology enabled global light-duty vehicle (LDV) sales.27 competitive advantages in portable consumer electronics In 2016, LIB cell manufacturing capacity (serving all end- end applications (Chung, Elgqvist, and Santhanagopalan market applications) was primarily located in China, 2016). Korea and China followed Japan’s lead in investing in Japan, South Korea, and the United States. Together, LIB cell and pack production for consumer electronics. these economies constituted 90% of 2016 global LIB cell South Korea’s LIB cell cluster is a result of government and production capacity for all end-use applications. Notably, industry efforts, started in the 2000s, to build up this portion clusters of key intermediate product manufacturing facilities of the supply chain (Chung, Elgqvist, and Santhanagopalan were also well established in China, Japan, and South Korea. 2016; Cision PR Newswire 2013). China, too, has fortified Such clusters may contribute to regional supply chain its LIB cluster development through various government advantages and cost benefits not available to cell research and development, tax, and investment incentives manufacturers located outside such clusters. Some degree (Lowe et al. 2010; Chung, Elgqvist, and Santhanagopalan of vertical integration exists across Asian processed 2016). While South Korean and Chinese cell manufacturers materials and cell production, which may also contribute initially relied heavily on Japanese suppliers, their national to lower input costs for certain manufacturers. The United efforts to build LIB clusters resulted in less dependence States, in contrast, hosted a relatively immature supply on Japanese suppliers and may have contributed to chain, and most U.S. cell and battery plant operators were advantageous pricing on key materials for fully scaled and relatively new to the industry. Nearly all U.S. LIB capacity colocated cell producers (Cision PR Newswire 2013). was targeted at serving the automotive market.

27 NREL estimate based on Richter (2017), EIA (2017), and BNEF 2016

98 Manufacturing LIB Cells Appendix

Historically, the United States has not been a leader in LIB manufacturing. The United States hosted 16% of global LIB cell production capacity for EVs in 2016; this share is expected to increase with the expansion of Tesla’s Gigafactory, a joint venture between Tesla and Panasonic with an announced production capacity upon completion of 35 GWh.28

China, Japan, and South Korea developed vertically integrated supply chains for LIB cells and intermediate products (cathodes, anodes, electrolytes, and separators). However, the U.S. supply chain is more focused on cell and pack manufacturing.

28 Information was accessed November 2018 on the Tesla 2018 Tesla Gigafactory website (https://www.tesla.com/gigafactory).

99 Appendix Manufacturing c-Si PV Modules

Additional Benchmark Data: LIB Cell Trade Trends—Expanded View LIB cell trade, 2014–2016

2014 2015 2016 Brazil China $199M $124M $107M Canada $45M $44M $53M China $0M $0M $0M Denmark $6M $8M $11M Germany $208M $249M $314M India $65M $141M $237M Japan -$287M -$191M -$251M Malaysia -$318M -$416M -$334M Mexico $118M $156M $131M South Korea -$950M -$802M -$724M Taiwan -$10M $2M $25M United Kingdom $52M $77M $91M United States $711M $775M $872M Brazil $0M $1M $0M Germany Canada -$3M $1M $6M China -$208M -$249M -$314M Denmark $3M $3M $3M Germany $0M $0M $0M India $0M $0M $0M Japan -$200M -$307M -$294M Malaysia -$65M -$82M -$97M Mexico $0M $0M -$3M South Korea -$236M -$271M -$446M Taiwan -$7M -$7M -$9M United Kingdom $7M $15M $20M United States -$64M -$93M -$61M Brazil $10M $12M $1M Japan Canada $10M $13M $14M China $287M $191M $251M Denmark $1M $1M $1M Germany $200M $307M $294M India $9M $2M $10M Japan $0M $0M $0M Malaysia -$1M -$3M -$8M Mexico $65M $34M $57M South Korea $47M $20M -$64M Taiwan -$2M $5M $14M United Kingdom $3M $5M $31M United States $580M $431M $366M Brazil $65M $17M $21M South Korea Canada $2M $2M $2M China $950M $802M $724M Denmark $0M $1M $0M Germany $236M $271M $446M India $4M $3M $4M Japan -$47M -$20M $64M Malaysia -$11M -$8M -$3M Mexico $14M $13M $26M South Korea $0M $0M $0M Taiwan $2M $8M $17M United Kingdom $8M $4M $3M United States $137M $57M $184M Brazil $5M $3M $2M United Canada $14M $14M $20M States China -$711M -$775M -$872M Denmark $0M $0M $1M Germany $64M $93M $61M India $40M $28M $27M Japan -$580M -$431M -$366M Malaysia -$18M -$21M -$13M Mexico $90M $130M $90M South Korea -$137M -$57M -$184M Taiwan -$61M -$60M -$67M United Kingdom $5M $55M $108M United States $0M $0M $0M Imports Exports Imports Exports Imports Exports

Bars show trade (exports positive and imports negative) and numerical value indicates balance of trade (BOT, exports minus imports) for LED packages by economy. All data are in US$(2014).

Between 2014 and 2016, the distribution of trade among the imports in 2016 came primarily from South Korea, Japan, five key LIB economies remained relatively stable over the and Malaysia. The top three destinations for exports from period, with China, Japan, and South Korea the top exporters Japan and South Korea in 2016 were China, Germany, and and the United States and Germany the largest importers. the United States. The United States imported LIB cells primarily from China, Japan, and South Korea in 2016. The top three destinations for exports from China in 2016 were the United States, Japan, and Germany. China’s

100 Manufacturing c-Si PV Modules Appendix

U.S. imports of LIB cells for use in EVs, 2015–2016

LIB Cell - LDV LIB Cell - All Applications

Japan 2015 $38M $530M

2016 $64M $497M

China 2015 $24M $841M

2016 $43M $957M

South Korea 2015 $20M $109M

2016 $7M $216M

United 2015 $4M $17M Kingdom

2016 $0M $15M

Mexico 2015 $0M $7M

2016 $3M $16M

Taiwan 2015 $1M $61M

2016 $1M $68M

Germany 2015 $1M $22M

2016 $1M $38M

Canada 2015 $0M $6M

2016 $1M $5M

Malaysia 2015 $0M $23M

2016 $0M $16M

Imports Imports

Bars show U.S. imports of LIB cells for EVs (left) and all applications (right) in US$(2014). Note that data for All Applications are shown on a different scale.

U.S. LIB cell imports for EVs accounted for a small but (EVs specifically). Overall, U.S. imports of LIB cells increased increasing share of LIB imports for all applications—5.3% in from $1.7 billion in 2015 to $2.0 billion in 2016, while imports 2015 and 8.4% in 2016. of LIB cells for EVs increased from $90.5 million to $167.3 million. The majority of U.S. LIB cell imports for EVs came In 2015, the United States added a new Harmonized System from Japan and China. (HS) code29 to track imports of LIB cells for use in LDVs

29 Code 850760-0010 : Lithium-ion storage batteries of a kind used as the primary source of electrical power for electrically powered vehicles of subheading 8703.90

101 Prepared by: Debbie Sandor David Keyser Samantha Reese Ahmad Mayyas Ashwin Ramdas Scott Caron James McCall

Established in 2015 by the U.S. Department of Energy’s Clean Energy Manufacturing Initiative, CEMAC engages the DOE national lab complex, DOE offices, U.S. federal agencies, universities, and industry to promote economic growth in the transition to a clean energy Operated by the Joint Institute for Strategic Energy Analysis economy. CEMAC is operated by the Joint Institute for Strategic Energy Analysis at the https://www.jisea.org/manufacturing.html DOE’s National Renewable Energy Laboratory.

@JISEA1 NREL/TP-6A50-78037 January 2021 https://www.linkedin.com/company/joint-institute-for-strategic-energy-analysis/