The Pennsylvania State University

The Graduate School

John and Willie Leone Family Department of Energy and Mineral Engineering

A FEASIBILITY STUDY IN ADOPTION TO AUGMENT

SUSTAINABLE ELECTRICITY GRID EXPANSION IN DEVELOPING COUNTRIES

A Thesis in

Energy and Mineral Engineering with Option in Energy Management and Policy

by

Anukalp Narasimharaju

 2014 Anukalp Narasimharaju

Submitted in Partial Fulfillment of the Requirements for the Degree of

Master of Science

December 2014

The thesis of Anukalp Narasimharaju was reviewed and approved* by the following:

Antonio Nieto Associate Professor of Mining Engineering Thesis Advisor

Serguei Lvov Professor of Energy and Mineral Engineering & Materials Science and Engineering, Director of Electrochemical Technologies Program

Zhen Lei Assistant Professor of Energy and Environmental Economics

Luis F. Ayala H. Associate Professor of Petroleum and Natural Gas Engineering Associate Department Head for Graduate Education

*Signatures are on file in the Graduate School.

ii

ABSTRACT

The quality of affects the efficiency of Energy Consumption. With developing countries struggling with power deficiency, energy storage coupled with renewable sources of energy can act as a short to medium term solution by providing power and kick-starting economic growth from stagnation. It can bring down dependence on thermal and nuclear power in the long run, which are both expensive to develop and run. Battery Energy Storage is also being developed using different materials and technologies serving different purposes. This thesis reviews several different types of batteries along with its advantages and disadvantages. Further, this thesis focuses on the electrochemistry of Redox Flow Batteries, most recent developments, its applications and a case study. The long lifetime of a redox flow battery adds tremendous value in maintaining and building energy capacity for a micro grid or for an expanding community. There are several drivers for energy storage adoption in developing countries. A synopsis of drivers behind flow battery adoption along with the power deficit specific to India is examined. New energy technologies have high costs of development, which are passed on to the consumers. However, unaffordable technologies are not a solution until they become affordable. A breakdown of costs in manufacturing flow batteries are investigated. An evaluation of manufacturing flow batteries both in India and the U.S has been prepared. Further, opportunities for cost reduction are identified. A detailed economic analysis of the cost of adoption and operational costs has been explored while examining the feasibility and adoption of technology for a tourist hotel in the desert region of Udaipur, Rajasthan in India. The results indicate that manufacturing flow batteries in India not only has a higher initial cost than Lead Acid batteries but also has similar pay back periods. The cost of flow batteries must be driven down below $300/kWh to be effectively deployed on a larger scale, while making profits for companies and serving communities with fast and reliable access to energy.

iii

TABLE OF CONTENTS

List of Figures ...... vi

List of Tables ...... vii

Acknowledgements ...... ix

Chapter 1 Introduction ...... 1

Chapter 2 Energy Storage ...... 4

Technology Overview ...... 4 Battery Energy Storage ...... 7 Lead Acid Batteries ...... 7 Lithium Batteries ...... 9 Sodium Sulfur Batteries ...... 10 Nickel-Metal-Hydride Batteries ...... 11 Redox Flow Batteries ...... 12 Summary of Advantages and Disadvantages ...... 15

Chapter 3 Flow battery Technology...... 17

Technical Review of an All Battery ...... 17 Complete Cell System Characteristics ...... 20 Technical Review of the Vanadium Bromine Redox Flow Battery ...... 23 Properties of a Battery ...... 28 Conductivity ...... 28 Vanadium Ion Diffusion ...... 28 Water Content ...... 28 Dimensional Stability ...... 28 ...... 29 Temperature Effect ...... 29

Chapter 4 Electricity Market in India ...... 32

Drivers for Flow Battery adoption ...... 32 Cost Structure of a Redox Flow Battery ...... 39 Assumptions of the Cost Model ...... 40 Design of Cells and Stacks ...... 41 and Membrane Configuration and Costs ...... 43 Configuration and Costs ...... 45 Tank Configuration and Costs ...... 46 Pumps Configuration and Costs ...... 46 Pipes Configuration and Costs ...... 47 Power Electronics Configuration and Costs...... 48 Container Configuration and Costs ...... 49 Labor Costs ...... 50 Transportation Costs ...... 51 Final Costs...... 52 iv

Summary ...... 54

Chapter 5 Flow Battery – Real World Applications ...... 60

Rajasthan’s Climate and Power Deficit Situation ...... 60 Case Study – No Power Outages ...... 64 Methodology ...... 64 Electric Loads ...... 66 Simulation ...... 67 Economic Considerations ...... 69 Results – No Power Outages ...... 71 Case Study – Power Outages Modeled into the Grid ...... 72 Methodology ...... 72 Simulation and Results – Flow Battery Configuration ...... 74 Simulation and Results – Vanadium Redox Lead Acid (VRLA) Battery Configuration ...... 75

Chapter 6 Discussion and Conclusion ...... 78

Appendix A ...... 80 Appendix of Results ...... 80 Model Data and Graphs – Flow Battery with Outages ...... 80 Model Data and Graphs – Valve Regulated Lead Acid (VRLA) Battery with Outages ...... 90 References ...... 103

v

LIST OF FIGURES

Figure 2-1: Pumped Hydropower reservoir in PA(Engineers 2012) ...... 5

Figure 2-2: For the Charge Process, the Hydrogen atom dissociates itself and is absorbed by MH alloy. During Discharge, a hydrogen atom dissociates from MH alloy and joins NiOH ...... 12

Figure 2-3: Configuration of a Redox Flow Battery (Shigematsu 2011) ...... 13

Figure 2-4: Cross Section of a Cell Stack (Shigematsu 2011) ...... 14

Figure 3-1: Layout of an All Vanadium Redox Flow Battery (Skyllas-Kazacos, Milne et al. 2008) ...... 18

Figure 4-1: Peak Shaving Effect by the usage of Redox Flow Batteries (Preetesh U. Munshi 2009) ...... 35

Figure 4-2: Cell Stacks Arrangement with PVC tubing and connections(Zhao, Zhang et al. 2006) ...... 43

Figure 4-3: Arrangement of 14 cells in a stack(Zhao, Zhang et al. 2006) ...... 44

Figure 4-4: Stack of batteries arranged on Metal Shelving, connected with PVC pipes(Zhao, Zhang et al. 2006) ...... 49

Figure 4-5: Container Dimensions to set up Battery ...... 50

Figure 4-6: Distance between Manufacturing Unit and Customer (Google Maps, 2014) ...... 51

Figure 4-7 : Distance between Manufacturing Unit and Shipping port ...... 52

Figure 5-1: Daily average temperatures, lows and highs(WeatherSpark 2013) ...... 61

Figure 5-5: Monthly Solar Irradiation at Udaipur, Rajasthan (NREL 2014) ...... 64

Figure 5-7: HOMER configuration using Vanadium Redox Flow Battery ...... 67

Figure 5-8: HOMER configuration using VRLA battery stack ...... 68

vi

LIST OF TABLES

Table 2-1: Advantages and Disadvantages of Battery Energy Systems ...... 15

Table 3-1: Membrane Properties ...... 29

Table 4-1: Specifications of a Cell Stack(Zhao, Zhang et al. 2006) ...... 42

Table 4-2: Component Cost by Percentage...... 54

Table 4-3: Individual Costs for Cell Components ...... 55

Table 4-4: Chemical and Electrical Characteristics of Cell ...... 55

Table 4-5: Plant Configuration ...... 56

Table 4-6: Pump Parameters ...... 56

Table 4-7: Tank Parameters ...... 57

Table 4-8: Power Electronics Parameters ...... 57

Table 4-9: Stack Parameters ...... 57

Table 4-10: Comparison of Assembly costs, India and USA ...... 58

Table 4-11: Comparison of Drivers of Assembly Cost by Percentage, India and USA ...... 58

Table 5-1: Average Solar Radiation from Clearness Index at Udaipur, Rajasthan(NREL 2014) ...... 63

Table 5-2: Load during Normal Season (Widatalla and Zinko) ...... 67

Table 5-3: Load during Tourist Season (Widatalla and Zinko , Wikipedia 2014) ...... 67

Table 5-4: Variations in Equipment Type considered with Flow Batteries ...... 69

Table 5-5: Variations in Equipment Type considered with VRLA Batteries ...... 69

Table 5-6: Flow Battery Simulation Results ...... 71

Table 5-7: VRLA Battery Simulation Results ...... 72

Table 5-8: Component Parameters ...... 74

Table 5-9: Economic Analysis ...... 74

Table 5-10: Financial Gains ...... 74

Table 5-11: Component Specification ...... 75

vii

Table 5-12: Economic Analysis ...... 76

Table 5-13: Financial Gains ...... 76

viii

ACKNOWLEDGEMENTS

I am most grateful to my adviser Dr. Antonio Nieto for his precise guidance and utmost assistance. I was always welcomed with an enthusiastic and approachable personality that helped me complete this work within challenging deadlines. He gave me immense freedom in my research and never created boundaries. I will be forever thankful for his exceptional leadership and skillful management of my work.

I sincerely thank Dr. Serguei Lvov and Dr. Zhen Lei for being on my committee and for providing valuable inputs and suggestions to improve my work. I would like to acknowledge and thank Dr. Sanchit Khurana who helped me grasp the electrochemistry of flow batteries and provided his expertise to aid my understanding.

I would like to thank my parents who have given me this extraordinary opportunity to be at Penn State and have provided their support and strength to the fullest, while making sacrifices for which I will always be indebted.

I have been able to shape my career in energy because of the funding provided by Denise Bechdel of the PA Small Business Development Center. I thank her for the opportunity to work as her Intern and for her enduring support. I will always remember my learning in the first year of appointment. I would also like to thank Shawn, Diane, Cece and Claire at Penn State Public Broadcasting for continuing my funding and aiding in the completion of my program.

I express my gratitude to Srivatsa Krishna and Tim Hennessey for believing in me and advising me in shaping my future.

Further, I would like to thank Sweta Padma for her timeless companionship and heartfelt affection every day. My journey at Penn State is replete with memories of my most special friends. I thank Jaggi for always being available and making life beautiful.

ix

Chapter 1

Introduction

India faces an acute shortage of electricity. Several major cities and villages across the country face intermittent power cuts throughout the day every single day. This results in economic losses across all industries. There is also a severe lack of investment in power back up systems.

Poor management of coal infrastructure, weak financial health of power companies and commercial losses by theft and defective meters are the major reasons that contribute to the situation. Currently, there is an enormous lack of affordable and quality power, which is hindering India’s growth. There is a massive strain on the current production facilities and recently a big failure of the Northern and

Eastern grids halted two fifths of the country.

Flow batteries can compensate for losses acting as a method of power back-up and can significantly curb the impact of power shortage. Energy storage has the potential to be an integral part of the grid by providing support to more efficient use of assets and seamlessly integrating into the grid with existing power plants and transmission lines. It is also an effective way of using energy arbitrage, wherein individual power producers can store excess energy during off-peak hours and sell this power at a higher price. Intermittent nature of renewable energy resources is also a compelling reason to use battery energy storage.

Imergy Power Systems is an energy company headquartered in Silicon Valley with operations in Gurgaon, India. Its mission is to provide low cost ownership of stationary energy

1

solutions. The firm provides a family of redox flow batteries, which significantly improve the cost and reliability of electricity.

It is proposed to examine the feasibility of installing flow batteries in different market segments and how to reshape the energy landscape in India. These batteries can be installed in remote villages with weak electricity connectivity. India is supporting programs to implement advanced methods of energy storage to bridge the gap between available generation and customer loads during peak hours. India is also pursuing energy storage as a secure source of power for

300,000 telecom towers. India can be used as a launchpad for global deployment of these battery technologies. It is proposed to also look at cost efficiencies and measure the socio economic impact of launching these in Indian villages or other sectors where this can be supported.

Losses in the electricity supply chain are the highest in the transmission and distribution sector.

Currently, local distribution companies face maximum losses. It is a viable option to target the distribution companies for adopting these products, since their power losses are higher than others and they can take maximum advantage of this technology. This technology will give transmission companies the capacity to provide power to their customers for almost 4 more hours in a day. It can also be bundled with Solar PV to provide energy storage. The analysis focuses on the financial payback and usability in the long run after flow batteries have been adopted in remote towns and villages or any place that is acutely power deficit.

In a broader perspective, this thesis addresses the following issues underlying the feasible adoption of redox flow batteries in India as a model state for future cross-deployment in other developing nations:

2

• Can Redox Flow batteries provide a solution to the lack of high quality electricity

in developing nations like India?

• Can businesses with small-scale stationary systems such as telecom towers,

railway signaling and ATM machines immediately benefit from an installed redox

flow battery replacing generators?

• Can flow batteries be connected to residential PV systems to reduce reliance on

unreliable grid supply feasibly?

3

Chapter 2

Energy Storage

The Energy Storage section gives a key description of the different technologies available.

Different types of battery technologies along with their respective advantages and disadvantages are discussed. Battery Energy Storage (BES) is currently being used in several different applications, as stand-alone devices powering individual units such as telecom towers and also as synchronized units integrated into a smart grid electricity network. This section explores different types of batteries.

Technology Overview

Pumped Hydro is the practice of using natural forms of rivers or lakes to fill a reservoir at a higher elevation than a dam at a lower elevation. A hydroelectric power plant is constructed along the banks and is used for producing electricity and as a pumped storage plant. A hydroelectric turbine pumps water from the stored water behind the dam up to the reservoir at the higher elevation. During peak hours in the day, the water from the top reservoir is released to the bottom reservoir producing electricity and transmitted to the grid for a higher peak price. At night, low cost electricity is used to pump the same water to the upper reservoir. (2012)

4

Figure 2-1: Pumped Hydropower reservoir in PA(Engineers 2012)

Pumped Hydro is able to store the maximum amount of electricity for the longest duration of time (up to half a year). It can also have the lowest cost of operation and production per kWh of energy produced. However, pumped hydro has one of the highest construction costs at $5,600/kW.

A conventional hydroelectric plant costs $3,100/kW to build. The facilities are also prone to damaging the environment and natural habitat due to the frequent change in water levels. (Munshi,

Pichel et al. 2009, 2012)

Thermal Energy Storage is mostly used where electricity is consumed locally. It uses low cost power over night to chill a coolant that is used to cool the house during peak time hot afternoons. It lowers the need for peak time supply. Similarly, salt is used to store heat in the day through concentrating solar plants. It is a highly complicated technology and is expensive to implement (Munshi, Pichel et al. 2009)

Fuel Cells consist of an electrolyte and have continuing electrochemical reactions. It has the capability of providing back up power for more than 7 hours. These cells have a poor efficiency

5

rate of about 35% and are not used when power back up is required for small durations only.

However, it is a good method to store energy in remote areas. (Munshi, Pichel et al. 2009)

Flywheels consist of a rotating wheel made of a highly-dense material shaped into a rotational cylinder. Strong magnets help in levitating bearings to minimize wear and tear. It stores energy for long periods of time and requires almost no maintenance for nearly two decades. Some of the advantages of flywheels are that they are extremely reactive to very quick charge and discharge cycles when required. A flywheel can be charged in less than an hour. They can be used in power plants to balance out spikes or sudden drops in power at power plants. However, it can discharge energy for only quick bursts such as less than 30 seconds and is used to provide energy when a generator has dropped in its power producing capacity for brief moments. Flywheels are extremely expensive and the technology is not wide spread. They need large spaces to operate and are typically used in few power plants. (Munshi, Pichel et al. 2009)

Superconductors offer little to no resistance to energy flow. They are used in grids where high intensity power is required almost instantaneously over few seconds. Superconductors are highly efficient energy storage devices. The current carrying capacity is significantly higher than other wires: however, it is also less cost efficient and technology adoption is slow. (Munshi, Pichel et al. 2009)

Compressed Air Energy Storage uses high pressure in the range of 40 to 70 bars to compress air and store underground in empty caverns. This air is later used to combust using a gas turbine. However, this method is not very efficient (~40-50%). Underground caverns are able to store vast amounts of compressed air and therefore can be adopted in the future once efficiency improvements are developed. (Ibrahim, Ilinca et al. 2008)

6

Batteries are the most well known and common, yet advanced energy storage device with a wide range of applications. They are used to power electronics and are versatile, robust, proven and inexpensive. Battery Energy Storage (BES) is a technology used for stationary applications and comes in different types based on the most apt application. (Munshi, Pichel et al. 2009)

Battery Energy Storage

Battery Energy Storage (BES) is the most commonly used form of energy storage technologies. It is also in advanced stages of development. These are used in portable electronics, grid storage, and electric vehicles. These systems are available in various shapes, sizes and capacities. Each battery system is built for specific use. The following sections describe some of the battery energy storage systems available.

Lead Acid Batteries

Lead-Acid batteries are used in Uninterrupted Power Supply systems and as batteries in vehicles. They can be connected in parallel and provide power to homes as standby during power outages. Lead Acid batteries contain dilute (H2SO4) as the electrolyte. Both the negative and positive are made from lead sulfate (PbSO4). During the charging state, the positive electrode takes up the form of lead oxide (PbO2) and the negative electrode takes up the form of Lead in its elemental form (Pb). The electrolyte used is 33.5% w/w sulfuric acid. Lead acid batteries are typically used in automobiles. The complete reaction taking place in this battery is described as:

Pb(s) + PbO2(s) + 2H2SO4(aq) → 2PbSO4(s) + 2H2O(l) …(1)

7

The negative plate reaction is given by:

− + − Pb(s) + HSO4 (aq) → PbSO4(s) + H (aq) + 2e …(2)

The positive plate reaction is given by:

− + − PbO2(s) + HSO4 (aq) + 3H (aq) + 2e → PbSO4(s) + 2H2O(l) …(3)

The Valve regulated lead acid battery is largely ‘maintenance free’ and comes sealed. It can be fitted into small enclosures, and also has a wide operating temperature. It has the ability to absorb vibrations. During the life cycle of the battery, incomplete charging causes sulfuric acid to travel through the water-acid mixture and collect at the base of the unit. Due to the lack of acid on the top of the battery i.e. closer to the electrode plates, the performance of the battery decreases over time. This also leads to corrosion and a reduction in battery capacity. Complete charging solves this problem, which mixes the electrolyte completely and leaves an even distribution of the chemicals in the battery.(Catherino, Feres et al. 2004)

Lead Acid Batteries are most commonly used in automobiles to start the engine. They are also used in emergency equipment for power back up. Larger power requirements include home- offices, where the battery is used in its wet-cell form. It has large power discharge capacity for long hours during power shut down situations. It is used as an Uninterrupted Power Supply (UPS) source. It is also flexible across small electric vehicles such as golf carts, wheelchairs and e- scooters. Military applications include the use of these batteries in nuclear submarines for power back up. (Hagerty , Misra 2007)

8

Lithium Ion Batteries

A Lithium Ion (Li-Ion) Battery is extremely popular due to its ability to not only be recharged but also due to extreme portability in hand held devices. They can be designed in various shapes such as cylindrical, flat bodied, and in long sandwich forms to be used in electric vehicles.

These batteries have high-energy storage capabilities in a small space and discharge slowly when not in use. Lithium Ion batteries are much lighter than lead-acid batteries, are able to provide the same amount of voltage and can be replaced in electric vehicles without much modification. The negative electrode is made from carbon. The positive electrode is made from a metal oxide. Popular materials for the negative electrode are graphite, hard carbon and silicon. These are paired with

Lithium Manganese Oxide or Lithium Iron Phosphate. The electrolyte used is an organic solvent with lithium salt. (Catalog 2003) The negative electrode half reaction is written in its general form as:

+ - xLi + xe + xC6→ x LiC6 …(4)

The positive electrode half reaction is written in its general form as:

+ - LiCoO2→ Li(1-x)CoO2 + xLi + xe …(5)

Current flows from the negative electrode to the positive electrode during the discharge cycle. Charging causes the to move in the reverse direction by applying a high voltage. The ions are thus stored in the negative electrode at full charge. Li-Ion batteries have a tendency to become electroplated at the negative electrode if the battery is charged in conditions below 0 degrees centigrade irreversibly. Other applications for Li-Ion batteries include common household indoor and outdoor power tools. These batteries also have the ability to be connected in parallel and be used in high powered electric vehicles. Special circuitry must be designed in these 9

applications with advanced materials for heat dissipation and temperature management.

(Thackeray, Thomas et al. 2000, Silberberg 2006)

Sodium Sulfur Batteries

Sodium sulfur batteries are very high capacity batteries. They work extremely well in large scale applications. They are mainly used in grids due to its very high operating temperature of 300 degrees centigrade. The electrodes are in liquid state at this temperature but the electrolyte remains in the solid state. These batteries last approximately 2500 cycles. That can be translated into a time period of 15 years when charged and discharged completely. If charged and discharged to only

65% of capacity, they can last for nearly 6000 cycles. Some of the characteristics of this battery include a rapid response time. It can discharge its complete charge in 1 ms when required. The energy density of this battery can be up to five times that of a lead acid battery. It requires minimal maintenance and can work remotely. Additionally, it can function in any environment and causes zero emissions or vibrations. Another feature of this battery is that 98% of the components can be recycled. (Hagerty)

During the charging cycle, positive Sodium ions are passed, which combine with sodium forming sodium polysulfide (Bito 2005).During the discharge cycle, the electrode allows positive sodium ions to flow through and at the same time electrons flow through the circuit of the device.

The total chemical reaction is given by:

2Na+4S = Na2S4 …(6)

10

Small scale distributed power generation systems like photovoltaic systems usually require only a few MW of energy storage. For these applications, NaS batteries are well suited.

The batteries are able to effectively serve as a robust solution to energy management by having good load balancing and peak shaving abilities. (Nourai 2002, Toledo, Oliveira Filho et al. 2010)

Nickel-Metal-Hydride Batteries

Nickel Metal Hydride batteries or Ni/MH batteries have several advantages over lead-acid and older Nickel Cadmium (Ni/Cd) batteries. The severe toxicity to the environment caused by cadmium disposal has lead to faster development of the Ni/MH device. Thus, Ni/MH batteries are directly replaceable in electronic devices that supported Ni/Cd batteries. Ni/MH batteries are based on the concept of an alkaline storage device. It has a 1.2V rating. These batteries are connected in series and offer more energy per unit weight and volume than lead acid batteries or older Ni/Cd batteries. These batteries have excellent energy density and high power density. The negative electrodes consisting of the Metal Hydrides compound is the main driver behind some of the advantages over lead-acid and Ni/Cd systems. Some improvements are needed in the battery due to a moderate to high rate of self-discharge. Additionally, a longer cycle life is needed for it to be adopted in applications such as electric vehicles and in other bipolar designs.

11

Figure 2-2: For the Charge Process, the Hydrogen atom dissociates itself and is absorbed by MH alloy. During Discharge, a hydrogen atom dissociates from MH alloy and joins NiOH

The hydride forming alloy, which is the active material determines the durability, capacity

and kinetics. Most research is being done in proposing the design and modeling of Metal hydride

materials for electrodes. (Feng, Geng et al. 2001)

Redox Flow Batteries

A redox flow battery works on the principle of oxidation and reduction between two active

chemicals. Two external tanks hold the and these are passed through the cell, which

also holds the electrodes. The positive and negative sides have their own respective electrolytes

12

stored in separate tanks to prevent self-discharge. The active chemical contains metal ions that are restrictively soluble in a solution. The ability to respond to any changes in frequent short cycle

Figure 2-3: Configuration of a Redox Flow Battery (Shigematsu 2011)

grid fluctuations is nearly instantaneous with a response time of only a few milliseconds. This is particularly useful in applications regarding renewable energy generation where energy generation is not constant. The electrolyte need not be changed at all. It is only recycled within the battery.

The only change occurring within the cell system is the change in ion valance. External power is connected to circulate the electrolyte through the cells. This is done by pumps. Flow batteries are fairly large in size because they do not have high energy density. The charge and discharge cycle is long during its life. The Vanadium-Vanadium system or V-V system uses the same metal ions in both negative and positive electrodes. If the material is mixed through the membrane, the battery capacity remains the same. (Shigematsu 2011) .The following reactions show the concept of functioning in a V-V system:

Positive Electrode:

2+ 2+ + -- VO (tetravalent) + H20 ↔ VO + 2H + e …(7)

13

The direction from left to right represents the charging cycle. Tetravalent Vanadium ions get oxidized to Pentavalent Vanadium ions at this juncture. The Hydrogen ions that are produced at the positive electrode are transferred to the negative electrode through the membrane. This assures that the electrolyte remains electrically neutral.

Negative Electrode:

3+ - 2+ - V (Trivalent) + e ↔ V (bivalent) …(8)

Trivalent Vanadium ions get reduced to bivalent Vanadium ions.

During the discharge cycle, the energy stored in the system is given out by the reverse process. The Electromotive force (EMF) achieved in this configuration is 1.26V. Different configurations are able to produce more EMF. These cells are connected in series as cell stacks to produce high voltages for real world applications.

Figure 2-4: Cross Section of a Cell Stack (Shigematsu 2011)

14

Summary of Advantages and Disadvantages

A summary of advantages and disadvantages of various Battery Energy Storage systems is represented in a table below:

Table 2-1: Advantages and Disadvantages of Battery Energy Systems

Lead -Acid Flow Batteries Sodium Sulfur Lithium-ion Nickel-Metal – Batteries (NaS) Batteries Hydride (NiMH) . Battery . The electrolyte . Upto four times as . High energy . Voltages similar to technology is and electro active efficient as lead density compared standard alkaline well developed materials are acid battery owing to other batteries and reliable stored separately to high energy technologies . Multiple recharges . Capable of therefore the density . Low self- possible thereby producing high uncharged . Production costs discharge rate making frequent voltages solution can be are low as sodium . Low battery . Low internal replaced like and sulfur are maintenance- no replacements impedance refueling a car cheap periodic obsolete

thereby leading to . Long lifetime . Useful discharge . Easy storage and the lowest . Electrolytes used applications in required transportation discharge rate have a low self solar and wind . Battery shape and . Inexpensive among discharge rate farm energy size can be . Absence of

Advantages rechargeable . Installation time storage customized cadmium makes it systems period is short (8 according to relatively . Robust; low mo for multi MW requirements and environmentally maintenance systems and 3 mo space constraints friendly; contains requirements; for smaller mild toxins long shelf life if systems) stored without . No toxic or electrolyte polluting gas . Construction is emissions; silent simple; low operation manufacturing costs

15

. Energy density is . Low energy . Poses accidental . Technology is not . Degraded relatively low density hazards brought on fully mature — performance at high leading to . Manufacturing is by violent reaction changes in metal temperatures degraded expensive as it is b/w sodium and and chemical . High maintenance performance over a complex system sulfur combinations requirements- must time to build . Only one known affect battery test be completely . Must be stored in . Toxic electrolyte application till results discharged to a charged state leaks possible date- load leveling prevent crystal once the by electrical . System formation electrolyte has utilities in Japan complexity leads been introduced . Limited shelf life- to high to avoid 2 to 5 years manufacturing deterioration of . Requires bulky costs

Disadvantages the active insulation . Battery chemicals. performance . Electrolytes and management lead used are requires complex toxic to circuitry environment . Each type of . Danger of battery requires a overheating specific type of during charging charger which is and not suitable inconvenient for fast charging

16

Chapter 3

Flow battery Technology

The Redox Flow Battery is one of the most efficient battery technologies available today to store, convert and deliver electrical energy (A. Landgrebe 1989, Vafiadis and Skyllas-Kazacos

2006). It is a regeneration device that converts the chemical energy stored in soluble reactants into

DC electricity. A reversible electrochemical reaction stores and releases the energy. Two redox couple solutions are present with a difference in their potentials. An ion exchange membrane separates the two solutions physically. This type of battery is much superior to other technologies due to many characteristics. The physical components and electrochemistry of the battery are discussed below.

Technical Review of an All Vanadium Battery

The all vanadium redox flow cell contains redox couples of Vanadium(II)/Vanadium(III) and Vanadium(V)/Vanadium(IV) in the negative and positive half cells respectively. Sulfuric acid is used as the supporting electrolyte. The dual electrolyte system achieves the separation of the redox couple using a cation exchange membrane (H+). The transfer of a proton takes place through this membrane. Simultaneously, an electron is transferred through the external circuit of a connected load. During the discharge cycle, the reactions occur in the forward direction. During the charge cycle, the reactions occur in the reverse direction. The reactions occurring at the positive electrode are:

+ + - + 2+ - VO2 (aq) + H (m) + e + H3O (aq) → VO (aq) + 2 H2O(l) …(9)

17

The reactions occurring at the negative electrode are:

2+ + +3 + - V (aq) + H3O (aq) → V (aq) + H (m) + H2O(l) + e …(10)

+ 2+ As can be seen, the VO2 is converted to VO at the positive electrode during the charge

cycle. Also, V2+ is converted to V+3.(Vafiadis and Skyllas-Kazacos 2006, Kear, Shah et al. 2012)

Electrical Charging

Source

V

2

+

VO

(

Negative Electrode Half Reaction Reaction Half Electrode Negative

aq

Positive Electrode Half Reaction Half Electrode Positive

2

) + )

+

Cation Exchange Membrane Exchange Cation

+ +

H

H

3

VO

O

Negative Electrolyte Negative

Positive Electrolyte Positive

+

+

(

H

(

m

aq

2

2

+

O

) +

) )

+ + (

Negative V → Positive

l

)

2 2

e

Electrolyte + Electrolyte

-

H

e

+

+ +

- 3

Tank 2 Tank

O

(

H

aq

(

3

l

) + )

O

)

+

H

(

aq

+

(

m

) )

) + + ) → →

Pump Pump

Load External Circuit

Figure 3-1: Layout of an All Vanadium Redox Flow Battery (Skyllas-Kazacos, Milne et al. 2008)

18

Positive Electrode

Electrode

Usually, carbon electrodes have deemed successful adoption in the V-V system. Monofiber electrodes made from bundles of 3000 carbon cloth fibers of 7-10 micrometers have been found to be excellent. These are well defined with electrode reaction rates similar to Graphite (Hiroko

Kaneko 1991). More recent research has continued the trend of testing materials, which are graphite based. Carbon nanotubes which have good electrical conductivity have not shown desired reversibility features (Zhu, Zhang et al. 2008). The chemical stability of the electrodes with the solution is also taken into consideration. Carbon felt electrodes are being used most commonly for commercial applications (F. Mohammadia). Carbon felt can be used in pure form or in doped form.

It is able to provide stability to the electrolytes and materials. Polypropylene is a carbon polymer design that has also been successfully tested for this purpose(Yue, Li et al. 2010).

Solution

New techniques of preparing VOSO4 for use in the positive electrolyte solution are being researched. Recently, Panzhihua Steel Corporation set up a procedure to prepare this using vanadium bearing slag (Li L 2003, Huang 2008) . The positive electrolyte solution contains

+ +2 vanadium with redox numbers of 5 and 4 and has VO2 and VO ions. Certain key components like KCl, which is a precipitation inhibitor has been used to eliminate precipitation of solid vanadium in the positive electrolyte solution (Maria Skyllas‐Kazacos 1999).

Negative Electrode

Electrode 19

The negative electrode is commonly manufactured with carbon felt or carbon based electrodes. This material is regarded as a good fit for maintaining the required chemical stability and kinetics for optimum performance of the all vanadium redox flow cell.

Solution

A 2-3 mol/l solution of H2SO4 and 1 mol/l Vanadium species ensures that the reactions take

+ +2 place as desired. The positive electrode solution contains VO2 and VO electrolyte. The negative electrolyte solution does not have the active ion associated with these, V2+ and V3+. The negative electrodes are less studied due to the greater significance of positive electrolyte in affecting different factors (Ch. Fabjan a 2001).

Complete Cell System Characteristics

Performance Characteristics

The complete reaction is given by the equation:

+ 2+ + 2+ +3 VO2 (aq) + V (aq) + 2 H3O (aq) ↔ VO (aq) + V (aq) + 3 H2O(l) …(11)

The power output or the performance and the capacity of the redox flow battery is determined by the stack size of the number of cells and the volume of the active material (electrolyte). Carbon and plastics have low cost and are usable components in designing the cell. Vanadium concentrations of 1.5 to 5.4 M are used along with 3.6 to 4.3 M of H2SO4. This is determined by using different

+ +2 2+ 3+ electrode materials like carbon or platinum. The values for VO2 / VO and V / V are reversible

(Ch. Fabjan a 2001, B. FANG 2003, Ye Qin 2010). They are designed with ratings from 1kW to

10 MW. They are able to discharge power for up to 16 hours and are able to carry out load leveling and can balance out other factors such as intermittent renewable energy production. The battery

20

lasts for approximately 10,000 cycles and can operate at a discharge efficiency of approximately

78% (Huang 2008).

Some of the biggest challenges to over come in the performance of vanadium batteries are thermal stability related to the membrane, precipitation of vanadium and viscosity. This limits the working temperature between -5 and 40 degrees centigrade. All Vanadium batteries can be used to support emergency equipment in hospitals, industrial trucks, railroad signalling, telecom towers, renewable energy load leveling and applications involving peak shaving. Batteries can last for up to 10 years and are available in configurations from 1 kW to 10 MW. (Rychcik and Skyllas-

Kazacos 1988, Zhao, Zhang et al. 2006)

Membrane

A membrane is best suited to be used for these applications when it allows the transfer of a proton. It must however resist the crossover of vanadium and water. Additionally, it must be

+ 5+ maximally resistant towards the chemical degradation from VO2 and V .(Theresa Sukkar 2003,

Dongyang Chen 2010). Some early tests had revealed that other than perfluorinated membranes, most other membranes like Selemion CMV showed poor stability in vanadium solutions. A modification of the Daramic Company manufactured microporous separator, Daramic Membrane was used in Vanadium Redox Batteries due to their high stability. These membranes cannot be used without modification because of high permeability leading to low coulombic efficiency. Nafion solution is used to produce The Daramic Composite membrane. In tests, the Daramic/Nafion2 membrane with a backweb thickness of 0.25 mm had a low water uptake of 31.2%/wet membrane.

When the separator was used with vertical ribs, the lowest water uptake of 28.8%/wet membrane was noticed. Nafion contributes to the low water uptake. The Daramic/Nafion composite membrane

21

limits self discharge and gives high current efficiency (Mohammadi and Skyllas-Kazacos 1995,

Hwang and Ohya 1996, Tian, Yan et al. 2004).

Recent Developments and Challenges

Vanadium Redox Batteries are currently limited by their energy density, which is not as high as Li-Ion batteries (higher than 150 WhL-1). The Energy Density of Vanadium Batteries is between 20-30 WhL-1.

Recent research has focused on increasing the operating temperature range along with keeping costs down. The cost of vanadium and the membranes used is a limiting factor in keeping costs low. (Aaron et al.) have managed to increase the power density of the battery by more than five fold by modifying the architecture of the batteries. This was achieved by using carbon paper electrodes and a zero gap flow architecture. Additionally, charge transport distances were reduced.

Multiple electrode material layers were used in negative and positive electrodes, which improved the output and decreased kinetic, mass and ohmic transport losses in the system (Skyllas‐Kazacos

2010, Li, Kim et al. 2011, Aaron, Liu et al. 2012) .

Commercial Batteries

ZBB Energy Corporation has a range of advanced energy storage battery modules that are expandable from 50kWH to 500kWh. One unique advantage of these batteries is its operating range between -30 to 50 degrees centigrade along with an increased energy density capacity of 5 times of commercially available Vanadium Redox Battery systems.

22

Technical Review of the Vanadium Bromine Redox Flow Battery

The Vanadium-Bromine redox flow battery was developed due to the low energy density of the All

Vanadium Battery (25-35 Whkg-1). The All Vanadium battery has a maximum concentration of vanadium of approximately 2 mol dm-3, which limits the solubility of V(II) and V(III) ions in

H2SO4. The energy density is determined by factors such as the concentration of the redox ions in solution, the transfer of number of electrons during discharge (per mol of active redox ions) and cell potential. The Vanadium-Bromine redox flow battery however, is capable of reaching the concentration of active ions to 3-4 mol dm-3. This allows the Vanadium-Bromine battery to reach energy densities of up to 50 Whkg-1 (Skyllas-Kazacos 2003, Skyllas-Kazacos, Mousa et al. 2003,

Ponce de León, Frías-Ferrer et al. 2006).

Positive Electrode

The electrodes are designed with conductive plastic sheets bonded to carbon or graphite felt. The

- - Vanadium-Bromide (V/Br) redox flow cell uses the redox couple Cl / BrCl2 at the positive electrode. The reactions occurring at the positive electrode are shown below. The reaction occurring during the charging cycle is:

− − − − 2Br + Cl → ClBr2 + 2e (charge) …(12)

The reaction occurring during the discharge cycle is:

− − 2− − 2Br + Cl ← ClBr2 + 2e (discharge) …(13)

23

Negative Electrode

The V/Br redox flow cell uses the redox couple VBr2 /VBr3 at the negative electrode. The reactions occurring at the negative electrode are shown below. The reaction occurring during the charging cycle is:

− − VBr3 + e → VBr2 + Br (charge) …(14)

The reaction occurring during the discharge cycle is:

− − VBr3 + e ← VBr2 + Br (discharge) …(15)

The figure of a Vanadium Bromine battery is shown:

Electrical Charging

Source

Negative Electrode Half Reaction Half Electrode Negative

Positive Electrode Half Reaction Half Electrode Positive

Cation Exchange Membrane Exchange Cation

1

Negative Electrolyte Negative

Positive Electrolyte Positive

/

V

2

Br

2

+(

2

aq (

Negative aq Positive

) ) ) + )

Electrolyte ⇄ Electrolyte

V

Tank e Tank

-

3

⇄

+(

aq

Br

) + )

-

(

aq

e

) -

Pump Pump

Load External Circuit

24

Figure 3-2: Graphical Representation of a Vanadium-Bromine Flow Battery (Skyllas-

Kazacos, Milne et al. 2008)

Alternate Negative Electrode

(Skyllas-Kazacos, 2003) has researched an alternate negative half-cell that contains the redox

2+ 3+ couple given by V /V . The electrolyte is designed using VCl2 or VCl3 dissolved with HCl (aq).

To prevent or minimize any cross contamination problems, it has been recommended that the

- - VCl2/VCl3 couple be used rather than the Br /Br3 couple.

The half reaction for this electrode can be represented by the equation below. The reaction occurs from left to right during the charging phase. The reaction occurring from right to left is during the discharge phase.

V3+(aq) + e-⇄ V2+(aq) …(16)

The VCl2/VCl3 redox couple also has good reversibility.

Membrane & Electrolyte

A Nafion 112 ion exchange membrane is used. An electrolyte consisted of 3 mol dm-3 V(IV) bromide solution in 3-4 mol dm-3 HBr or HBr/HCl on each membrane side (Ponce de León, Frías-

Ferrer et al. 2006).

25

Full Cell Characteristics

The complete reaction of the Vanadium Bromide Redox Flow Battery is given by the chemical equation below. The reaction occurring from left to right is during the charging cycle whereas the reaction occurring from right to left is during the discharge cycle.

2+ - 3+ (1/2) Br2 (aq) + V (aq) ⇄ Br (aq) + V (aq) …(17)

The Nernst equation

Current voltage prediction models have inaccuracies of 131 to 140 mV or approximately 10% of the total voltage. This difference is typically added to the data to account for discrepancies that might result from using experimental data to the calculated open circuit voltage (OCV).

Additionally, these errors are caused due to only few parameters considered in the model. The

Nernst equation addresses this issue by including the electrochemical mechanisms and accurately describes the potential difference between the electrolyte and electrode at no net reaction inside the cell.

The Nernst Equation is described as follows:

푅푇 푐표푥.훿표푥 퐸푓푢푙푙 = 퐸°0 − ln [ ] 푛퐹 푐푟푒푑. 훿푟푒푑

where Efull is the potential difference, 퐸°0 is the standard reduction potential, R is the universal gas constant, T is the absolute temperature, n is the number of equivalents transferred per mole of

26

reduced or oxidized species and F is Faraday’s constant. The term c describes the ionic concentration and delta represents activity coefficient of species. Additionally, red and ox describes the reduced and oxidized species respectively. The activity coefficient (훿) is typically 1 for redox flow batteries due to negligible interactions among ions. This is due to the usage of liquid electrolytes and good circulation (Sukkar and Skyllas-Kazacos 2003, Hamann, Hamnett et al. 2007,

Knehr and Kumbur 2011).

The two half reactions that are used to calculate the cell potentials are:

- - Positive electrode: (1/2) Br2 (aq) + e ⇄ Br (aq) …(18)

Negative electrode: V3+(aq) + e-⇄ V2+(aq) …(19)

Membranes

Ion Exchange membranes are defined as the most critical part of the redox flow battery. It defines its economic value to a large extent and even the battery’s performance. The membrane prevents the oxidation and reduction reactants from mixing and undergoing a direct chemical reaction. It is a separator that provides a path for ionic conduction between the two electrolytes. Membranes are designed to be extremely selective towards conducting the charge-carrying ion. The Hydrogen Ion is the main charge carrying cell in the Vanadium Bromide cell. Additionally, the flow of vanadium and polybromide ions is restricted. Nafion 112, Nafion 117 is a commonly used cation exchange membrane. The thickness of a membrane can range from 0.03 to 0.62 mm (Oei 1985, Chieng,

Kazacos et al. 1992, Vafiadis and Skyllas-Kazacos 2006).

27

Properties of a Battery

Conductivity

The ionic conductivity of the membranes is measured using impedance spectroscopy. The frequency range is from 1 to 106 Hz with AC amplitude of 0.2V. Additionally, the resistance of the conductivity cell is measured with and without the membrane. (Vafiadis and Skyllas-Kazacos

2006).

Vanadium Ion Diffusion

A static diffusion test determines the rate of diffusion of vanadium ions across the membrane.

(Vafiadis and Skyllas-Kazacos 2006).

Water Content

To calculate the water content of a membrane, it is immersed in distilled water for 24 hours. It is then removed, patted dry and placed on filter paper. The weight is measured until the surface of the membrane is dry. It is then further dried at 60 degrees centigrade by a vacuum. The water content of the membrane is measured by taking the percent weight change before and after vacuum drying(Vafiadis and Skyllas-Kazacos 2006).

Dimensional Stability

Samples were immersed in a solution of distilled water, a solution containing 1 M V3+ and 1 M V4+ in 6.4 M HBr, 2 M HCl and in 6.4 M HBr, 2 M HCl for 15 days to determine the dimensional

28

stability. Physical changes of length, width and thickness were determined using a digital caliper

(Vafiadis and Skyllas-Kazacos 2006).

Energy Density

The energy density of a redox flow battery is proportional to the concentration of the redox ions in solution, to the cell potential and to the number of electrons transferred during discharge per mole of active redox ions (Skyllas-Kazacos 2003).

Temperature Effect

Voltage efficiency decreases with decreasing temperature and increasing current due to lower reaction rates and ohmic losses. On the other hand, current efficiency increases with increasing current and decreases with increasing temperature due to faster diffusion rates of vanadium and poly bromide ions across the membranes at higher temperatures (Skyllas‐Kazacos 2010)

Recent Developments

Most membranes being tested are unsuitable for use in vanadium bromine flow batteries due to low chemical stability or very high resistance. One membrane in particular, the ABT3 seems to have shown better adaptability towards use in a Vanadium Bromide flow cell. Some of its properties compared with the Gore M04494 membrane are as follows:

Table 3-1: Membrane Properties

Membrane IEC(mmol g-1) Resistivity(Ωcm2) Water content V4+diffusivity (%) (10-7cm2 min-1) 29

Gore M04494 1.00 0.41 0.96 2.16 ABT3 6.01 3.24 0.11 4.25

The highly oxidizing nature of Br3- ions causes rapid deterioration of most polymeric membranes and thus only limited types of membranes can be used for long life. Another problem that could occur with the vanadium bromine flow cell is the possible emission of bromine gas during cell charging. This mainly occurs due to multiple ionic equilibria of bromide ions in aqueous solutions.

To overcome this issue, various bromine complexing agents are added. The complexing agents bind with bromine producing an orange layer, which settles down at the bottom and can be easily separated (Skyllas-Kazacos, Milne et al. 2008, Skyllas‐Kazacos, Kazacos et al. 2010).

Another important issue with energy storage technologies is the cost of raw materials. In 2008, there were large variations in the prices of vanadium pentoxide. This led to fluctuations in vanadium redox flow battery pricing and thereby caused investor hesitation. The largest reserves of vanadium are in China and are usually sourced from fly-ash, spent catalysts and waste slags from steel production. Vanadium Pentoxide prices have stabilized in recent years. A mechanism to lower battery prices is to use vanadium pentoxide with high impurities. However, extensive research is required to identify the appropriate impurity level for specific materials that can be used successfully in commercially produced batteries to minimize fouling and precipitation(Huang, Li et al. 2008, Skyllas-Kazacos, Chakrabarti et al. 2011). Imergy Power, has recently been able to produce flow batteries from recycled vanadium from mining slag, oil field sludge, fly ash and other environmental waste thereby lowering its cost from $500 per kWh to under $300 per kWh for its flow batteries. Additionally, this has led to the doubling of the energy density (Still Jul 9, 2014).

30

Commercial Products

V-Fuel Pty Ltd was founded by the inventor of the Vanadium Redox Batteries Professor Maria

Skyllas-Kazacos and Mr Michael Kazacos. V-Fuel has an exclusive world-wide license for the

Vanadium bromine Technology and was founded in the University of New South Wales, Australia.

Batteries are available with power outputs ranging from 5 to 50 kW (2014).

31

Chapter 4

Electricity Market in India

India has suffered enormously due to power deficiency. Some rural areas of the entire Indian subcontinent remain not just without power, but also without an electricity connection. Nationwide,

33% of Indian households are off the grid without any access to electricity and 6% of the urban population also remains without power. That amounts to nearly 300 million people living without power. The per capita power consumption stands at 778 kWh compared with that of the US per capita rate of consumption at 11,919 kWh (FICCI 2012). For substantial progress of the industry, agriculture or service sector, meeting energy requirements is the fundamental criterion. If power requirement is adequatey met, small-scale industries can progress resulting in employment opportunities followed by rise in income and purchasing power, thereby leading to the development of the nation.

Drivers for Flow Battery adoption

Rapid Growth in Installed Capacity of Renewables

Approximately 12% of India’s 207,000 MW installed electricity capacity comes from renewable sources. The National Solar Mission aims to install 20,000 MW of grid connected by

2022. India also ranks 5th in the world in installed wind power capacity. This will enhance the need for flow batteries due to the intermittent nature of renewable energy resources, namely solar and wind. It provides a good case for using Battery Energy Storage (BES) to augment the renewable energy supply. 32

Carefully planned BES installations can cost as low as $500 per kilowatt hour. The economic cost of setting up BES is becoming more affordable. Imergy Power Systems is already producing batteries made from recycled vanadium, which has abundant availability in the form of waste slags. This has the potential to create cheaper vanadium markets and tie-ups with steel factories can bring down the cost of raw material significantly. A special formulation is used to treat the vanadium that exists at a lower purity level (98.5 percent) and this can be used as the electrolyte (John 2014).

Although there was a lack of incentives until now to pursue flow battery adoption in the renewable energy sector, the government has taken several initiatives to accelerate early adoption of new technologies. Fiscal and financial incentives include concessions of 80% accelerated depreciation, concessional custom duty on renewable technologies, excise duty exemption, sales tax exemption and income tax exemption for 10 years on profits. Additionally, the Remote Village Electrification

(RVE) program has provided 1,537 remote villages and hamlets solar powered home lighting in

2010-2011. Today, more than 8,100 villages and hamlets have been benefited by solar powered installations. These remote hamlets and villages can benefit from government subsidized installation of flow batteries. A guaranteed electricity for nearly 4 to 5 hours a day on battery power alone, can jump-start the productivity of individuals living in these remote settlements.

Energy Arbitrage

Energy Arbitrage is the “process of storing excess energy generated during off peak hours and storing it before selling the low-priced electricity at a much higher price during peak demand hours

(Preetesh U. Munshi 2009).” This concept can be used in remote Indian villages with a slight

33

variation. Villages typically receive power around midnight. A village can use electricity from the grid at night to irrigate fields and store energy in the battery. The stored energy could be sold later during the day to a neighboring village that is off grid. Over the years, if solar PV is connected to the flow battery and the battery capacity is increased, an opportunity to sell back excess power to other neighboring settlements is a viable and cost-effective option.

Time of day Pricing

The dependence on the main grid during peak time demand is what causes excessive load shedding in cities. Utilities can manage demands by adopting the time of day pricing which is not yet present in India due to old metering systems installed in majority of the consumer entities of the country.

Flow battery adoption can help begin an initiative to move away from all day flat rate pricing towards time of day pricing thereby making the Indian electricity market more nimble and efficient.

It is only fair that a user pays more during peak time demand and manages personal consumption to prevent wastage of power. Additionally, power producers do not need to sell power at a loss and seek subsidies from the government or pass on huge timely price increases to the customer. As a demand for power is better managed, load shedding of power will reduce and electricity generation companies can benefit from higher incomes. This is called peak shaving.

34

Figure 4-1: Peak Shaving Effect by (thePreetesh usage U.of MunshiRedox Flow 2009 Batteries) (Preetesh U. Munshi 2009)

As can be seen from the graph above, demand for electricity rises from 6AM to 12PM. The evening hours between 4PM to 7PM see a higher rate of demand in households. This peak demand is supplied by power plants that are specially set up to handle peak loads and function only for certain hours in a day. During this time, flow batteries can be used to discharge into the grid to consumers at a lower cost of power. Additionally, after midnight, when the price of electricity is relatively lower, the flow batteries can be charged from the base load generating power plant. Energy Storage can also improve the quality of power supplied due to minimal or no variations in output. It enables regulation of generation to match demand.

Energy Independence

A deal has recently been struck between India and Indonesia wherein the latter country to supply

105.8 million tonnes of coal to India during the summer months of April to October 2014. This is a 20 percent increase in imported coal. India ranks 3rd as the largest importer of coal despite having the world’s 5th largest reserves (Dave 2014). Coal India Limited, a state owned coal company is 35

expected to fall short of 155 million tonnes of coal this fiscal year (2013-2014). Additionally for the years 2016-2017, the projected coal shortage is 350 million tonnes. The coal imported from

Indonesia is priced at $60 a tonne (Eric Yep 2014). The total expenditure on coal for the year 2016-

2017 is projected to be:

350,000,000 tonnes of coal * $60 = $21,000,000,000

Despite an expenditure of $21 billion dollars of imported coal in India, the country still faces an acute shortage of supplied power. Peak hour deficit during 2010-2011 was 9.8 percent. The demand stood at 8,61,691 GWh whereas the supply available was only 7,88,355 GWh. Over the last 30 years, the demand has grown steadily by 3.5% annually and is expected to grow at higher rates in the coming few years. The major reasons for these situations are:

 Poor transportation infrastructure has not allowed even available coal to be supplied to

plants leading to shortage of fuel

 Power Distribution Companies (DISCOMs) have poor financial health.

 DISCOMs have high aggregate transmission and distribution losses of 31%

 Plant load factor is low

India could take the initiative to spend more money on becoming energy independent.

Strategically, energy security is important to India and by investing in energy storage immediately and taking its time to slowly develop its reach of renewable or thermal power to other off-grid remote settlements can save a large amount of capital and can see rapid development in areas that desperately need it. Energy Storage can act as a short to medium term solution to providing power and transmitting it across short distances.

36

Emergency Power and Smart Grid Integration

There have been two instances of notable importance in terms of power cuts in the country. A power crisis that occurred on January 2nd 2001 saw the failure of the northern power grid. This halted most of the Northern states and this black out costed industries a loss of more than $500 million. This incident occurred again on July 30th 2012 when the Eastern and North Eastern grids failed. This brought the entire nation to a standstill and effectively halted the transportation sector and industry. The services industry too was brought to a standstill affecting millions of lives (FICCI

2012).

By incorporating flow batteries, transmission and distribution assets can be integrated with each other and can effectively cooperate by having two-way communication to improve the performance of the grid. In case of major shut downs, the flow batteries can be automatically used to discharge power into the grid to the end user.

There are several smart grid communication technologies available, however the challenge is to use a standardized protocol that will be able to interact and communicate with systems across the network. As new power stations are inaugurated in India, it is a great opportunity to use new assets that are compatible with a smart grid concept.

An example of such an integration of technology is present in England. Scottish and Southern

Energy has installed flow batteries, that have an output of 100 kW. This system is mainly used for supplying power in case of an emergency load-loss to the substation (MacLeman 2009). About

42% of the Indian utilities are State run. This means there is a tremendous opportunity to deploy smart grid technology on a large scale. There are vast amounts of internal resources that could be

37

utilized towards providing incentives to vendors to install features towards making the smart grid.

More number of sites can benefit from smart grid equipment. Utilities that are smaller do not have the funds or the resources to benefit from large scale deployment of customized technology for smart grid operation.

Apart from new power plants that are beginning service in India, there is a tremendous opportunity to introduce smart grid components in upgradation of power plant equipment through schemes.

Usually, large thermal power plants are installed far away from consumption centers like towns or cities. These systems allow a top – down and a unidirectional power flow from the power plants to smaller load centers. Smaller and more localized power plants may not allow the flow of power bi- directionally to allow the functioning of the smart grid.

Smart Grid technologies enable real time data exchange and communication from generation through consumers. Transmission and distribution losses can be minimized along the grid with the incorporation of flow batteries and replacing generators. The need for human intervention can be minimized or eliminated completely when the power plant needs to be protected due to sudden changes in the working characteristics. This decreases the likelihood of a complete grid shut down.

The integration of flow batteries also makes the entire network more robust and flexible. The demand side should diversify localized electricity generation technologies by incorporating PV or wind. This will help them achieve economic targets and reduce environmental emissions.

Distribution losses are much greater than transmission losses. About 2% of overall losses can be attributed to just transmission of power. Operational constraints also force power generation plants to operate at lower efficiencies. (Hamidi, Smith et al. 2010).

38

Environmentally Friendly

Vanadium Flow batteries do not let out any carbon emissions. They burn clean and only require electricity from the plant to charge. These can replace current generators that are both noisy and carbon emitting. Imergy Power Systems recently designed vanadium flow batteries from recycled vanadium of 98.5% purity. A single percentage brings the cost down by a great degree. Slag from steel plants can be used to recover vanadium. It’s a clean source of fuel.

Cost Structure of a Redox Flow Battery

The cost of flow batteries is currently on an average $500 per kWh. Some companies have experimentally developed batteries that cost $300 per kWh. Some advanced processes in obtaining vanadium are being developed in flow battery company technology labs. The cost is likely to come down to approximately $220 per kWh in the next few years. The cost is the key component, which will determine the adoption rate of the battery. There is a critical relationship between the energy capacity and the cost of the energy storage platform.

Every single component adds weight and cost. An attempt is made to evaluate the cost structure of different components to further understand how optimization processes can bring cost of individual components down thereby affecting the price package of the battery pack.

Simple economies of scale and tough competition in the race to develop cheaper flow batteries may decrease overall production costs over time. Individual component costs will change over time.

However, strategic sourcing from various destinations may bring down costs.

39

A comprehensive list of all battery components for an All Vanadium battery is generated. Different aspects of the battery were evaluated to be included in the cost model. The cost model was designed to issue results, which were compared with published work for an evaluation of trends in the relationship between costs per kilowatt hour and energy capacities.

Assumptions of the Cost Model

Some model costs were determined from published papers such as by Jossen and Sauer. A few other material costs were determined by availability in India or the US. Since Imergy Power

Systems has a manufacturing unit in India, it would be appropriate to consider an Indian assembled product. A few materials for which reliable price points in India were unavailable, a currency conversion of pricing was done for the product that is available in the US.

Added to the equation would be an assumed overhead cost along with a labor cost equivalent to that of an Indian working professional. Material costs for the exact electrolyte wasn’t found. The cost for vanadium was assumed from a mass wholesale retailer. The membrane used in this cost model is a Nafion 115 membrane. Costs that were not found were either not considered or were used from published work. High unit and volume costs were directly calculated from the cost of a single unit of material. A case for wholesale prices of materials from different sources was not considered in the cost model.

The time to assemble a system and install it is assumed to be 8 months. The overhead costs associated with the energy storage system along with miscellaneous costs are assumed to be 12% of total component cost. The average labor cost in India for a well-experienced assembly professional is assumed to be $2 per hour. This is not a conservative estimate by Indian standards

40

and is a realistic figure assuming that the company employs well-trained technical personnel with experience in the assembly and manufacturing of medium to large mechanical devices. If the same battery system were to be assembled in the US, it would cost the company approximately $30 per hour in labor wages. This cost is a large saving when economies of scale are considered.

Transportation of these battery units will need care and expertise in shipping and handling.

It will be important to identify shippers that will professionally and safely transport this equipment.

Three different shipping cases are considered. One scenario identifies shipping costs within India by road for a distance of 349 miles from a manufacturing unit in Gurgaon to the Thar Desert,

Rajasthan, a hot bed of construction for solar power facilities funded by the National Solar Mission.

A second scenario investigates transportation costs of a battery pack across the world to the US by sea. A third case is then evaluated to assess costs associated with transporting these battery systems to Frankfurt, Germany by air. These cost estimates are made by evaluating shipping costs of equipment with similar size and weight specifications.

Real costs associated with commercial battery production are rare to find and is proprietary information of companies. These costs are also dependent on the location of production and the timeline of development. (Nieto, Bai et al. 2014)

Design of Cells and Stacks

It is desired to find costs of building and installing a battery of power capacity 10 kW. The energy density of the battery is 40 kWh. The battery is capable of delivering power of 10 kW for a period of 4 hours (Zhao, Zhang et al. 2006). A VRB cell stack was designed and fabricated by Skyllas-

Kazacos et al., that produced an average power output of 1.1 kW at a current of 60 mA. The current

41

density during discharge was 60 mA cm-2. The energy efficiency of this battery was 77.7%(Skyllas-

Kazacos, Kasherman et al. 1991).

Table 4-1: Specifications of a Cell Stack(Zhao, Zhang et al. 2006)

To produce the required power, it was determined that a total of 112 cells will be required.

To optimize stacking, 8 stacks of 14 cells each are used. Cells are stacked in 2 top and bottom shelves alongside 4 on each shelf. The connection between series and parallel can be done either way as it can affect the power only negligibly.

42

Electrode and Membrane Configuration and Costs

All vanadium batteries operate mostly on carbon felt electrodes or carbon based electrodes. Sodium

Figure 4-2: Cell Stacks Arrangement with PVC tubing and connections(Zhao, Zhang et al. 2006)

polysulfide/bromine redox flow batteries also use similar carbon felt electrodes or active carbon

electrodes. However, direct comparison studies between the two shows superior characteristics of

carbon felt electrodes for use in redox flow batteries

43

.

Figure 4-3: Arrangement of 14 cells in a stack(Zhao, Zhang et al. 2006)

There are 8 stacks of which each stack is using 875 cm2 of electrode surface area. The total surface area used by all the electrodes is 7000 cm2 or 0.7 m2. The exact cost of Carbon Electrodes could not be determined, but is estimated to be approximately $90 per square meter. For the cell configuration, it brings the total cost to $63.

Membrane costs were obtained from DuPont. The membrane area per stack is assumed to be 1 m2.

There are 8 stacks present and approximately 8 m2 of membrane material is required. The costs obtained for a high performance Nafion membrane from Ion Power, Inc is found to be approximately $1114 for 1.5 m2. The costs of a membrane for 8 m2 of the material is found to be

$5,941.

44

Current collectors cost approximately $51 per m2. For 8 m2 of material, the total cost required would be approximately $408.

The total cost comprising of electrode material, membrane material and current collectors for building all 8 stacks is found to be $6412.

Electrolyte Configuration and Costs

VOSO4 and Sulfuric acid are the main components in the electrolyte solutions for a vanadium battery system. The aqueous electrolyte concentration that is considered will be a concentration of

1.5M VOSO4 and 3M H2SO4.

The exact suppliers of vanadium were unavailable for access to exact price data. The price of vanadium obtained from conversion of waste slag or recycled material was also unavailable.

Vanadium costs were obtained from USGS, 2013 which was $14.33/kg. The cost of sulfuric acid is $0.07/kg from Sigma Aldrich (Spellman, Stiles et al. 2013).

The quantity of electrolyte needed is about 14.8 liters per cell. For a total of 112 cells, the system uses 1657.6 liters. The quantity of VOSO4 needed is 621.6 kg. This brings the total cost of vanadium to be $8,907. The quantity of sulfuric acid needed is 1036 kg. The cost of sulfuric acid would thus be $73.

The total cost of the electrolyte would be the combined cost of sulfuric acid and vanadium solutions for a 10 kW all vanadium battery, which amounts to $8,980.

45

Tank Configuration and Costs

Tanks are available in sizes ranging from a maximum capacity of 75,000 liters to a minimum capacity of about 37 liters. PVC tanks were chosen since they are not easily prone to corrosion. It is also least expensive compared to steel tanks.

The amount of solution that is used is approximately 1,700 kgs. By approximation, 1 kg of vanadium is approximately 1 liter of solution. The tank thus needs to hold 1700 liters of solution to accommodate both liquid electrolytes in a working battery. Tanks with a capacity of about 1700 liters (450 gallons) cost approximately $918.

The total cost of two tanks with the ability to hold these solutions would be $1,836.

Pumps Configuration and Costs

The pumps are used to circulate the electrolytes around the system. DC brushless pumps will be used. A total of two pumps are required, attached to each tank of electrolyte. The cost of a 1 hp pump from Global Industrial is $450. The total cost of two pumps would amount to $900.

Flow meters are used to monitor the flow of electrolytes. It also assists in regulating electrolyte flow through the cells. Each stack of cells requires two flow meters to monitor the input and output flow of electrolyte through the cell stacks. A total of 16 flow meters are required for 8 stacks at a cost of $200 each. The total cost of flow meters for the entire energy storage system is $3,200.

The total cost of pumps and flowmeters is found to be $4,100. 46

Pipes Configuration and Costs

The types of pipes that are considered for flow battery operation are Polyvinylchloride or PVC pipes. These can be bought in bulk and are fairly inexpensive. It is cheaper to use PVC than to use polypropylene pipes or High Density Poly Ethylene pipes. PVC pipes are cheaper by more than half the cost of HDPE and polypropylene pipes.

PVC pipes available in various size and shapes offering great flexibility and customizability of design. Once the battery is altered for different power configuration and cell sizes, these pipes can be taken apart and interconnected again according to new specifications. The main driver of the cost for these pipes would be the length and diameter of tubing and the number of elbow joints and gasket connections required.

The cost of the pipes is found to be $5 per meter for pipes of 3 inches in diameter. Each elbow joint is $0.46. The cost of PVC tubing is found to be $0.90 per meter. A special tubing that is capable of withstanding temperatures of up to 150 F can be used for flow batteries installed in desert regions.

Special polymer gaskets prevent ruptures, leaks and cracks from forming at critical edges. These can be obtained at a cost of approximately $46.

It was estimated that each cell stack requires 4 meters of PVC tubing, 4 meters of PVC piping, 2 gasket connections and 4 elbow joints (Larsson 2009). The total cost of the components is calculated to be $117.44.

47

Power Electronics Configuration and Costs

Flow batteries produce their power in Direct Current. In order to make this power usable to run lights, fans and home appliances, the power needs to be converted into Alternating Current.

Inverters perform this function and are available in the market from a kilowatt to megawatt scale.

Some models suggest that inverters cost approximately 25-30% of the entire cost of building a flow battery. This is a significant cost addition. However, this is usually true for Megawatt scaled systems where inverters with high capacities are bound to be very expensive.

A good quality inverter of 10 kW capacity costs between $2,000 and $6,000.

A control system is required to monitor and regulate the output voltage and current. Its also needed to display system characteristics, like charge capacity, mode of running and number of hours of running. Another added cost is the power electronics of the system. This would include the cabling, circuit breakers, and electrical connections. These costs are estimated to be approximately $80 per kW.

A lightweight model of 10 kW capacity was chosen to be used for evaluation in this model at a cost of $5,000. The total costs of power electronics and control systems is found to be $800. The total cost of control systems, power electronics and the inverted are priced at $5,800.

48

Container Configuration and Costs

The entire battery energy storage system can be built into a 20 foot container or less. This would help in ease of mobility. It can also be shipped out by sea readily and can also be transported by road ways and railway. Metal shelving can be obtained or built according to exact specifications to stack cells and arrange them in exact configurations. Tanks can be fixed in place by hinges or in metallic boxes. The total cost of metallic shelving is assumed to be $300.

A 20 ft container can be obtained for $1,500 or less. The total costs for metal shelving and container is found to be approximately $1,800.

Figure 4-4: Stack of batteries arranged on Metal Shelving, connected with PVC pipes(Zhao, Zhang et al. 2006)

49

Labor Costs

With higher demands, economies of scale could be obtained by scaling up manufacturing and assembly lines which would bring down costs of assembly. A 10 kW system is a smaller system and can be manufactured relatively quickly when compared to Megawatt scale systems.

Since the assembly is being done in India, the labor costs are fixed at $2 per hour. Being a

Figure 4-5: Container Dimensions to set up Battery relatively new technology, every order is custom made with 3 technicians. It is assumed that the entire battery system can be visualized, planned, designed, manufactured, built and transported in

60 days with around 40 hour week schedule. Further, it is assumed that it can be installed, tested and run in another 30 days by the same technicians.

The number of total working hours inclusive of all labor components amounts to 2,160 man hours.

At a cost of $2 per hour, the total labor cost on the system is $4,320.

50

Transportation Costs

Transporting the battery from its factory in Gurgaon, India to the Thar Desert, Rajasthan, India by road would be performed by a freight truck. Trucking prices were found by the Indian Foundation of Transport Research and Training (2014) . Truck freight services can be rented at a cost of $0.37 per km for a 9 tonne payload truck. The one-way distance of 561 km can be covered at a cost of

$208.

Figure 4-6: Distance between Manufacturing Unit and Customer (Google Maps, 2014)

For a shipment to the United States, the container can be transported to the port of Mumbai by road and then shipped out overseas by container ship. The same shipment of a 20ft container leaving

Gurgaon, India by road to Mumbai, covering a distance of 1,371 kms would cost $508. The

51

following journey by ship from the port of Mumbai to the port of Los Angeles would cost

Figure 4-7 : Distance between Manufacturing Unit and Shipping port

$2,966. The total cost of shipping the battery from India to the US would cost approximately

$3,500.

Final Costs

The total costs of the entire system when manufactured in India is calculated as $45,666. A comparative cost was developed for manufacturing the same product in the US. A steep hike in labor charge increases the cost to $111,379. The cost of manufacturing the same battery in the US is 2.43 times more than to manufacture it in India.

52

The top 10 contributors to cost are displayed in a pi-chart below. The electrolyte is the largest driver to cost. It contributes almost 20% to the entire cost of the battery energy system. It is followed closely by the control systems and power electronics. The membrane is one of the top 3 highest costing components of the flow battery system.

Figure 4-8: Top 10 Drivers of Cost in All Vanadium Flow Battery Assembly

53

The following table illustrates the top 10 contributors of cost in components.

Table 4-2: Component Cost by Percentage

Top 10 Drivers of Cost - Components Electrolyte 19.83% Control Systems and Power Electronics 17.67% Membrane 13.12% Inverter 11.04% Overhead 10.81% Labor 9.54% Pumps 9.06% Tanks 4.05% Container 3.98% Current Collectors 0.90%

Components that add most cost are also the most critical components with stringent specifications.

Costs that could be lowered are that of pumps, tanks, container and collector. Also, the costs for

PVC tubing, pipes, metal shelving and overhead costs could be lowered. All costs evaluated are on the higher side owing to realistic scenarios.

Summary

The following table represents costs from components that make up the cell. The costs for electrolyte, membrane and electrodes are summarized below.

54

Table 4-3: Individual Costs for Cell Components

Items Cost Vanadium Electrolyte Cost $8,907 Sulfuric Acid Electrolyte Cost $73 Total Electrolyte Costs $8,980 Total Membrane Costs $5,941 Current Collector Costs $408 Carbon Felt Electrodes $63

The following table represents battery chemistry and power densities. The stoichiometry and concentration of electrolytes are summarized below.

Table 4-4: Chemical and Electrical Characteristics of Cell

Cell Properties

Stoichiometry Positive Electrode / Charge Cycle V4+- e → V5+

Stoichiometry Positive Electrode / Discharge Cycle V5+ → V4+- e

Stoichiometry Negative Electrode / Charge Cycle V3+ + e → V2+

Stoichiometry Negative Electrode / Discharge Cycle V2+ → V3+ + e

Operational Temperature 73 F / 23 C

Concentration of Vanadium 1.5M

Concentration of Sulfuric Acid 3M

Power Rating 10 kW

Energy Capacity 40 kWh

Current Density 60 mA cm-2

55

The following table represents plant characteristics. It describes the cell, stack and plant configuration.

Table 4-5: Plant Configuration

Plant Characteristics Number of Stacks 8 Total number of cells 112 Container Cost $1,500 Piping and Mechanical Cost $117 Size of Stacks 1 m2 Container length 20 ft Container breadth 8 ft Container width 8 ft Cycles per year 365

The following tables describe the properties of the pumps and tanks used in the design and configuration of the battery. The tables also summarize costs of power electronics and design configurations of each stack.

Table 4-6: Pump Parameters

Cost of Pumps Pump Rating 1 hp Efficiency Rating 75% Cost of Pumps $450 Number of Pumps 2 Total Cost of Pumps $900

56

Table 4-7: Tank Parameters

Tank Characteristics Number of Tanks 2 Tank Size 1,700 liters Cost of each Tank $918 Total Cost of Tank $1,836

Table 4-8: Power Electronics Parameters

Power Electronics Inverter Rating 10 kW Inverter Cost $5,000 PC/Control System Cost $800

Table 4-9: Stack Parameters

Stack Characteristics Number of Cells per stack 14 Membrane Area per stack 875 cm2 Total Membrane Area 0.7 m2

The following table summarizes the cost break down of manufacturing the battery energy system in India versus manufacturing the systems in USA. It was assumed that the components that were used in both regions would be the same. Individual component costs available in India was not researched or considered. It is estimated that by using components locally available, it is possible to lower the costs even more by a significantly large margin with savings of 10% or more.

The only difference lies in labor costs and transportation costs. There is a substantial increase in labor costs when the battery system is manufactured in the U.S. Additionally, transporting the battery to Thar Desert in Rajasthan in India costs $3,500 as opposed to a lower transportation cost of just $208 when manufactured locally. 57

Table 4-10: Comparison of Assembly costs, India and USA

Break down of costs of Manufacturing Categories India USA Transportation $ 208 $ 3,500 Labor $ 4,320 $ 64,800 Container $ 1,800 $ 1,800

Control Systems and Power Electronics $ 8,000 $ 8,000 Inverter $ 5,000 $ 5,000 Pipes $ 117 $ 117 Pumps $ 4,100 $ 4,100 Tanks $ 1,836 $ 1,836 $ Electrolyte $ 8,980 8,980 Electrode $ 63 $ 63 Membrane $ 5,941 $ 5,941 Current Collectors $ 408 $ 408 Overhead $ 4,893 $ 12,545 Total Costs $ 45,666 $ 117,091

Table 4-11: Comparison of Drivers of Assembly Cost by Percentage, India and USA

Breakdown of Costs of Manufacturing - Drivers by Percentage Categories India USA Labor 9.46% 58.18% Overhead 10.71% 9.09% Electrolyte 19.66% 8.06% Control Systems and Power Electronics 17.52% 7.18% Membrane 13.01% 5.33% Inverter 10.95% 4.49% Pumps 8.98% 3.68%

58

Tanks 4.02% 1.65% Container 3.94% 1.62% Current Collectors 0.89% 0.37% Transportation 0.46% 0.19% Pipes 0.26% 0.11% Electrode 0.14% 0.06%

59

Chapter 5

Flow Battery – Real World Applications

India has struggled with meeting power demand requirements, even after relying on a combination of thermal, hydro and nuclear power generation methodologies. Rural electrification is of utmost importance and is the biggest challenge for the Government of India. A few hybrid systems can work to ease the power shortage situation in remote villages and towns. India has a significantly large and exploitable solar resource with 300 days of average sunshine. The Thar Desert in

Rajasthan is being developed for solar power projects. Apart from the National Solar Mission where

2 GW of solar installations are being made in this region, State governments are taking the initiative to implement and install their own solar based power projects in MW scale to supply power to local villages. This section explores two case studies that are developed in order to substantiate findings of results of benefits of using a flow battery in grid connected systems to replace the use of generators.

Rajasthan’s Climate and Power Deficit Situation

Rajasthan, a State in western India is popular among tourists and has several hotels of various sizes spread across the State. Local attractions mainly include Thar desert and lake palace hotels along with an experience of handicraft shopping and cultural experiences. The region has mostly desert climate with acute power shortage in cities and towns. The region typically receives a large amount of sunlight and its temperatures range from 48 F to above 105F.

60

For the case study, an eco-friendly hotel in Udaipur is chosen. Udaipur is located at 24.58 degrees

North and 73.68 degrees East. It has an elevation of 2000 ft with a total population of approximately

451,735. The following graphs show the daily average temperatures, warm days, and spread of the sun’s cover over an average year (WeatherSpark 2013).

Figure 5-1: Daily average temperatures, lows and highs(WeatherSpark 2013)

Figure 5-2: Time period of Temperatures in various zones(WeatherSpark 2013)

61

The summer starts early around April 10th and lasts until June 29th. The warmest month is May.

The region receives ample sunlight in the winter months as well. December 4th to February 12th is the winter season. January is the coolest month. Additionally, the sun shines for a long duration of time during the days, even in winter.

Figure 5-3: Everyday hours of Sunlight and Twilight(WeatherSpark 2013)

Figure 5-4: Everyday Sunrise and Sunset hours(WeatherSpark 2013)

62

The annual average daily solar radiation on a horizontal surface at this region of Udaipur with the coordinates of Latitude 24 degrees North, 35 Minutes and Longitude 73 degrees North, 41

Minutes is 5.59 kWh/m2/d. The information is also available with the database on the Surface

Meteorology and Solar Energy website (NASA 2014).

Table 5-1: Average Solar Radiation from Clearness Index at Udaipur, Rajasthan(NREL 2014)

63

Case Study – No Power Outages

A five star eco friendly hotel was chosen for the case study. Hotel Fatehgarh Udaipur already uses solar panels for water heating. It has also taken an initiative for rainwater harvesting. The hotel has

48 rooms and has ample amount of roof space. A similar sized hotel in Sudan with a similar climate uses 88,100 kWh/yr of electricity (Widatalla and Zinko). This number was used for the case study of the Fatehgard hotel Udaipur aswell. The aim of the project was to combine solar photovoltaics, a generator and a battery system to minimize annual expenditure on electricity. The batteries considered were modern Valve Regulated Lead Acid (VRLA) batteries and an All Vanadium

Flow Battery.

Figure 5-5: Monthly Solar Irradiation at Udaipur, Rajasthan (NREL 2014)

Methodology

An energy simulation software HOMER (Hybrid Optimization Model for Electric Renewables) was used for simulating the energy usage patterns of the hotel. It was developed by the National

64

Renewable Energy Laboratory (NREL). HOMER takes into account 8,760 hours in a year and makes energy balance calculations by simulating the operation of a system. It determines the supply and demand of electricity profile using a combination of energy sources that is decided by the user.

When considering flow batteries and PV, it automatically decides for every hour when to operate generators and when to charge-discharge batteries.

These energy balance calculations are performed for every system configuration that the user wants to consider. It calculates feasibility by providing Net Present Value and Levelized Costs of Energy by estimating the costs of installing and operating the system over the lifetime of the components.

The costs include operation and maintenance costs, fuel consumption, initial capital required and interest(NREL 2005). A large number of simulations were run to obtain results for this case study.

As can be seen in Figure 6, almost 36,500 simulations were run for a few systems to improve the accuracy of the situation.

Figure 5-6: Number of Simulations run for each configuration

A sensitivity analysis can also be performed to model the cost-effectiveness of the system over time as resource availability and economic conditions change. A range of resource availability and component costs can be fed as an input variable to identify factors that may impact the design or operation of the system. HOMER can calculate and provide several solutions from which an optimum system can be chosen.

65

The inputs that were chosen for this study are solar PhotoVoltaics, a generator, a DC/AC converter and a battery system. The system is connected to the grid and functions during power cuts as power back up system. Component specifications for PV panels, generators, batteries and converter were entered as an input for the simulation. Additionally, solar radiation density data for the hotel’s particular region using latitude and longitude co-ordinates was used for the model. Economical costs such as replacement capital, operation and maintenance costs of different equipment were also used as an input for financial evaluations. Most importantly, this model does not have any power outages built into the grid. The price of electricity supplied by the grid to the hotel is $0.08 per kWh.

Electric Loads

Hotel loads vary through the day. The loads will be more during the beginning of the day and during the tourist season. The type of load will define the kilowatt ratings required of the solar panels, flow batteries, converters and generators. The hotel will consume the least amount of power through the night. The electric load is modeled from the data written in the paper by

(Widatalla and Zinko). This is because the original data from the hotel was unavailable. It is safe to use the same data as the both businesses lie in similar climate regions and at a similar distance from the equator where the solar radiation density across the horizontal surface is also identical.

Both businesses are of the same size and offer indistinguishable services to their customers.

Additionally, the annual expenditure of the hotel on energy is $70,400. This includes a base case of the cost of running diesel generators burning 175 liters of diesel a day at $1 per liter to make up for power cuts.

66

The loads for the normal season as well as the tourist season are described by the tables below:

Table 5-2: Load during Normal Season (Widatalla and Zinko)

Table 5-3: Load during Tourist Season (Widatalla and Zinko , Wikipedia 2014)

Simulation

The following equipment was configured in the HOMER screen to consider for simulation and

optimized working combination. The

objective of this simulation is to compare the

efficiency or usefulness of All Vanadium

Redox Flow Batteries to the Valve Regulated

Lead Acid (VRLA) batteries for a hotel. The

Figure 5-7: HOMER configuration using Vanadium Redox Flow Battery

67

images show a graphically represented image of AC lines being connected to the Grid, along with a generator to a load of a hotel that uses 250 kWh/d of power. The load is also supported by a PV array and a battery energy storage system. Figure 7 represents the system configured for a

Vanadium Redox Flow Battery while Figure 8 shows the simulation configured for a stack of

Figure 5-8: HOMER configuration using VRLA battery stack

Valve Regulated Lead Acid (VRLA) batteries.

The lifetimes of the components considered are described below:

 Lifetime of Flow battery was considered to be 12 years

 Lifetime of the inverter was considered to be 7 years

 Lifetime of the VRLA batteries was considered to be 3 years

 Lifetime of the solar panels was considered to be 25 years

 Lifetime of the generator was considered to be 15,000 hours

Additionally, the annual expenditure of the hotel on energy is $70,400. This includes a base case of the cost of running diesel generators burning 175 liters of diesel a day at $1 per liter to make up for power cuts.

68

Table 5-4: Variations in Equipment Type considered with Flow Batteries

Gen VRB-ESS VRB-ESS Converter Configuration PV (kW) (kW) (kW) (kWh) (kW) Grid (kW) FLOW1 10 3 3 10 11 1000 FLOW2 25 3 1 4 20 1000 FLOW3 25 10 1 4 20 1000 FLOW4 40 3 1 20 30 1000 FLOW5 30 5 0.5 20 30 1000

Table 5-5: Variations in Equipment Type considered with VRLA Batteries

Configuration PV (kW) Gen (kW) 6FM200D Converter (kW) Grid (kW)

VRLA1 10 3 5 11 1000

VRLA2 10 3 4 11 1000

VRLA3 25 3 4 20 1000

VRLA4 40 3 4 30 1000

VRLA5 30 10 4 30 1000

Economic Considerations

Net Present Value

The Net Present Value is also known as Net Present Worth. The Net Present Value is the difference between the Amount of an investment made and the present value of future cash flow of an investment made. The Present Value of incoming or outgoing cash flow can be calculated by using a discounting method along with a rate of return. 69

The Net Present value is used by companies in making financially sound decisions while evaluating a project. It is capable of indicating how much value an investment towards a project would add. If the NPV>0, the project adds value and makes financial sense for the company. If the NPV<0, the investment does not add any value to the firm, but creates losses on the balance sheet.

The formula for NPV is given by the equation:

푁 푅 푁푃푉 = ∑ 푡 (1 + 푖)푡 푡=0

Where 푅푡 is the net cash inflow or outflow during time t, at a discount rate or rate of return, i (Wikipedia 2014). The interest rate considered for the calculations was 6%.

Levelized Cost of Electricity

The Levelized Cost Of Electricity (LCOE) is often used in expressing the value of a solar project.

It effectively gives an achievable price per each unit of energy ($/kWh) over the entire operable life time of a renewable energy project. The LCOE is simply defined and calculated by the formula,

푇표푡푎푙 퐿푖푓푒 퐶푦푐푙푒 퐶표푠푡 퐿푒푣푒푙푖푧푒푑 퐶표푠푡 표푓 퐸푙푒푐푡푟푖푐푖푡푦 = 푇표푡푎푙 퐿푖푓푒푡푖푚푒 퐸푛푒푟푔푦 푃푟표푑푢푐푡푖표푛

When multiple sources of energy generation are combined to produce electricity, LCOE can be used to combine these technologies and return an overall cost of electricity per unit. By combining

70

an overall cost, LCOE can help in comparison of several different technologies or a combination that optimizes and returns the best value (Tarn Yates 2012).

Results – No Power Outages

The following Tables 6 and 7 show a summary of the results obtained. The hotel pays an annual bill of $70,400 currently while operating from the grid alone. The results show a payback period that is of immense significance. In some cases, like Flow4, the operating cost is brought down significantly from $70,400 to just $23,473. Usage of Valve Regulated Lead Acid (VRLA) has shown similar results with payback periods being just around 3 years. In some cases like VRLA4, the initial capital costs are significantly high with a longer payback time. While considering all the cases, the Net Present Cost of Flow1 shows maximum Net Present Costs.

Table 5-6: Flow Battery Simulation Results

Initial Operating cost COE Payback Configuration capital ($/yr) Total NPC ($/kWh) Period FLOW1 $74,304 51,189 $728,671 0.623 3.87 FLOW2 $138,299 31,883 $545,865 0.431 3.59 FLOW3 $145,107 27,958 $502,498 0.397 3.42 FLOW4 $218,409 23,473 $518,467 0.341 4.65 FLOW5 $167,473 25,389 $492,030 0.366 3.72

71

Table 5-7: VRLA Battery Simulation Results

Configuration Initial capital Operating cost ($/yr) Total NPC COE ($/kWh) Payback Period VRLA1 $57,018 51,269 $712,409 0.61 2.98 VRLA2 $56,618 51,226 $711,452 0.609 2.95 VRLA3 $133,337 32,733 $551,779 0.435 3.54 VRLA4 $210,247 23,890 $515,645 0.337 4.52 VRLA5 $167,055 25,021 $486,902 0.36 3.68

Case Study – Power Outages Modeled into the Grid

The State government has planned four hour power cut every single day for residences and five

hour power cut for industries. Rajasthan relies on hydro power as well; however, the 2014 season

has seen less than average rain fall. A nuclear power plant of 250 MW capacity is undergoing

repairs and is thus shut down. Along with it, two other nuclear power plants of 200 MW capacity

each are not functioning due to annual repairs. There is also a shortage of imported coal, which has

led to an acute shortage of power in the State. This is where the economic value of the flow battery

gives a financial edge to companies in maintaining industrial output in a cost effective

manner(IANS 2014).

Methodology

Although HOMER is a powerful simulation tool, it does have some limitations. For instance, it is

unable to model power cuts from the grid. It assumes that all equipment is available 100% of the

time. To simulate a power cut, a generator with very large capacity is used. The initial capital costs

are set to 0. The replacement costs also are set to 0. This is how a grid is available to a customer. 72

The lifetime is also set to be more than the project lifetime (>25 years). A special fuel is defined.

The energy content of the fuel is 3.6 MJ/kg. The density is 1000 kg/m3. To make the fuel 100% efficient to simulate a grid, the efficiency curve has an intercept of 0 and a slope of 1. This is because, when lights, fans or home appliances use power from a grid, there are no losses although internal losses may be there. In the case of a grid, the energy purchased in kWh is equivalent to the energy used in kWh.

To simulate load shedding, the generator working as the grid was scheduled to be force fully turned off during the hours of 12PM to 5PM every single day of the year. This simulates the current power situation at Udaipur.

Figure 5-9: Grid Schedule being Forced Off to simulate Load Shedding of 5 hours a day,

everyday

73

Simulation and Results – Flow Battery Configuration

Table below shows the most optimized configuration of the system that must be used.

Table 5-8: Component Parameters

VRB-ESS PV (kW) Gen (kW) Grid (kW) VRB-ESS (kW) (kWh) Converter (kW) 30 45 1000 15 25 40

Table shows the costs associated with the battery energy storage system, solar PV, generator

costs, Net Present Value and Levelized Cost of Energy.

Table 5-9: Economic Analysis

Operating cost Initial capital ($/yr) Total NPC COE ($/kWh) $292,836 10,619 $428,582 0.367

Table shows the payback period and savings. Renewable fraction shows the fraction of

the time renewables are used to power the building.

Table 5-10: Financial Gains

Renewable fraction Savings Payback Period 0.54 59,781 4.90

With the current configuration, the monthly average electric production can be seen in Figure

74

Figure 5-10: Monthly Average Electric Production

The state of charge for the battery is seen to fall to nearly 20% on some days when the tourist season is high. On a daily basis, the battery is seen to discharge between 12 PM to 5 PM

Figure 5-11: Battery State of Charge

Simulation and Results – Vanadium Redox Lead Acid (VRLA) Battery Configuration

Table below shows the most optimized configuration of the system that must be used.

Table 5-11: Component Specification

PV (kW) Gen (kW) Grid (kW) 6FM200D Converter (kW) 30 45 1000 24 40

Table shows the costs associated with the battery energy storage system, solar PV, generator costs, Net Present Value and Levelized Cost of Energy.

75

Table 5-12: Economic Analysis

Operating cost Initial capital ($/yr) Total NPC COE ($/kWh) $211,006 9,767 $335,863 0.288

Table shows the payback period and savings. Renewable fraction shows the fraction of the time renewables are used to power the building.

Table 5-13: Financial Gains

Renewable fraction Savings Payback Period 0.53 60,633 3.48

With the current configuration, the monthly average electric production can be seen in Figure

Figure 5-12: Monthly Average Electric Production

The state of charge for the battery is seen to fall to nearly 20% on most days.. On a daily basis, the battery is seen to discharge between 12 PM to 2:30 PM and a charging cycle begins from the generator between 2:30PM to 6 P.M.

76

Figure 5-13: Battery State of Charge

77

Chapter 6

Discussion and Conclusion

It is observed that the payback period for Valve Regulated Lead Acid (VRLA) is almost half that of the payback period of Flow batteries. However, the net present costs of using the flow battery is a little higher. Additionally, the operating costs of both are almost similar. By increasing the PV value – although initial costs are higher, the operating costs are dramatically decreased over time. The price of development of flow batteries must come down in the future to $200 per kWh and below in order for it to be feasibly less expensive immediately. Numbers don’t tell the entire story.

Over the lifetime of 25 years of solar panels, a VRLA stack of batteries may need to be replaced every 2-3 years. However, flow batteries may need to be changed just twice during the course of the lifetime of the project. Additionally, operating temperature tolerance of the flow battery is much higher when compared with VRLA batteries. Also, VRLA batteries cannot be deep discharged very often. They are able to maintain a good state of operation and health only when discharged to a shorter state. Flow batteries are built for deep discharge and charge cycles with a life time of 10 years or more. The pumps and tanks may need to be changed after 7 years.

From the results that were published by HOMER, it was evident that the flow battery offers a better option when modeled with power cuts despite the numbers showing a more short term gain with the usage of VRLA.

78

The flow battery model has an annual operating cost of $10,619 with a payback period of

4.9 years. The Net present value was $428,582 with LCOE at $0.36. The VRLA flow battery model has an annual operating cost of $9,767 with a payback period of 3.48 years. This has a Net Present

Value of $335,863 and a LCOE of $0.28. However, the VRLA stack needs to be replaced every 2-

3 years and operates inefficiently in extremely high temperatures. In conclusion, flow batteries almost match competitive results offered by VRLA, however, unless the cost of flow batteries comes down further, initial capital investment costs will still be higher. For the next few years, flow batteries may see adoption with big businesses and around wealthy neighborhoods while VRLA is adopted for power back up solutions in lower income households and small businesses.

Additionally, it does not make financial sense to switch to these technologies when there are no power cuts and electricity is inexpensive. At that point, the results evaluated had a renewable fraction of less than 24% which means the technology is not efficient when left unused owing to high costs. A small flow battery with a small generator could be installed, for emergency backup purposes. After the electricity market matures into time of use pricing and opportunities for energy arbitrage are available, it would make economic sense if proper sized equipment is installed at that point to lower expenditure on electricity.

This is an opportunity for public and private sector banks, government departments to set up programs to incentivize businesses, residences and industries to adopt flow batteries to not just solve the power situation in the short term but also manage electricity demand over the longer term by assisting in the deployment of renewable energy systems.

79

Appendix A

Appendix of Results

Model Data and Graphs – Flow Battery with Outages

System architecture

Table A- 1: System Architecture

PV Array 30 kW

Generator 1 45 kW

Grid1 999,999 kW

Battery power 15 kW

Battery storage 25 kWh

Inverter 40 kW

Rectifier 40 kW

Dispatch strategy Cycle Charging

Cost summary

Table A- 2: Cost Summary

Total net present cost $ 428,582

$ Levelized cost of energy 0.367/kWh

$ Operating cost 10,619/yr

80

Figure A- 1: Cash Flow Summary

Net Present Costs

Table A- 3: Net Present Value

Capital Replacement O&M Fuel Salvage Total Component ($) ($) ($) ($) ($) ($)

PV 150,000 0 1,534 0 0 151,534

Generator 1 43,767 7,388 27,595 61,805 -6,952 133,603

Grid1 0 0 0 33,117 0 33,117

VRB-ESS Flow Battery 91,430 2,504 7,880 0 -979 100,835

Converter 7,639 2,279 0 0 -424 9,494

System 292,836 12,170 37,009 94,922 -8,354 428,582

81

Annualized Costs

Table A- 4: Annualized Costs

Capital Replacement O&M Fuel Salvage Total Component ($/yr) ($/yr) ($/yr) ($/yr) ($/yr) ($/yr)

PV 11,734 0 120 0 0 11,854

Generator 1 3,424 578 2,159 4,835 -544 10,451

Grid1 0 0 0 2,591 0 2,591

VRB-ESS Flow 7,152 196 616 0 -77 7,888 Battery

Converter 598 178 0 0 -33 743

System 22,908 952 2,895 7,425 -654 33,527

Figure A- 2: Cash Flows by Type

82

Electrical

Table A- 5: Electrical Factors

Production Fraction Component (kWh/yr)

PV array 72,232 63%

Generator 1 9,920 9%

Grid1 32,383 28%

Total 114,534 100%

Figure A- 3: Monthly Average Electric Production

Table A- 6: Load Fraction

Consumption Fraction Load (kWh/yr)

AC primary load 91,250 100%

Total 91,250 100%

Table A- 7: Load Factors

Quantity Value Units

83

Excess electricity 16,752 kWh/yr

Unmet load 0.0000849 kWh/yr

Capacity shortage 0.00 kWh/yr

Renewable fraction 0.536

PV

Table A- 8: Capacity Factors

Quantity VValue UUnits

Rated capacity 30.0 kW

Mean output 8.25 kW

Mean output 198 kWh/d

Capacity factor 27.5 %

Total production 72,232 kWh/yr

Table A- 9: PV Output

Quantity Value UUnits

Minimum output 0.00 kW

Maximum output 32.1 kW

PV penetration 79.2 %

Hours of operation 4,397 hhr/yr

Levelized cost 0.164 $/kWh

84

Figure A- 4: PV Output

Generator 1

Table A- 10: Generator 1 Statistics

Quantity Value Units

Hours of operation 615 hr/yr

Number of starts 291 starts/yr

Operational life 24.4 yr

Capacity factor 2.52 %

Fixed generation cost 9.26 $/hr

Marginal generation cost 0.257 $/kWhyr

Table A- 11: Output Parameters

Quantity Value Units

Electrical production 9,920 kWh/yr

Mean electrical output 16.1 kW

Min. electrical output 13.5 kW

85

Max. electrical output 42.0 kW

Table A- 12: Quantity Values

Quantity Value Units

Fuel consumption 4,694 L/yr

Specific fuel consumption 0.473 L/kWh

Fuel energy input 46,189 kWh/yr

Mean electrical efficiency 21.5 %

Figure A- 5 : Generator 1 Output

Grid1

Table A- 13: Grid 1 Operational Values

Quantity Value Units

Hours of operation 5,519 hr/yr

Number of starts 504 starts/yr

86

Operational life 39.9 yr

Capacity factor 0.000370 %

Fixed generation cost 0.00 $/hr

Marginal generation cost 0.0800 $/kWhyr

Quantity Value Units

Electrical production 32,383 kWh/yr

Mean electrical output 5.87 kW

Min. electrical output 0.00453 kW

Max. electrical output 37.4 kW

Table A- 14: Consumption Values

Quantity Value Units

Fuel consumption 32,383 kg/yr

Specific fuel consumption 1.000 kg/kWh

Fuel energy input 32,383 kWh/yr

Mean electrical efficiency 100.0 %

Battery

Figure A- 6: Grid1 Output

87

Table A- 15: Quantity Statistics

Quantity Value Units

Cell stack capacity 15.0 kW

Useable storage capacity 25.0 kWh

Autonomy 2.40 hr

Battery wear cost 0.005 $/kWh

Average energy cost 0.148 $/kWh

Figure A- 7: Monthly Statistics

Converter

Figure AFigure- 8: Battery A- 9 :Bank Frequency State ofHistogram Charge

88

Table A- 16: Capacity outputs

Quantity Inverter Rectifier Units

Capacity 40.0 40.0 kW

Mean output 5.7 0.1 kW

Minimum output 0.0 0.0 kW

Maximum output 31.0 15.0 kW

Capacity factor 14.4 0.3 %

Figure A- 10: Inverter Output Power

Figure A- 11: Rectifier Output Power

Emissions

89

Table A- 17: Emissions from Pollutants

Pollutant Emissions (kg/yr)

Carbon dioxide 12,030

Carbon monoxide 241

Unburned hydocarbons 26.7

Particulate matter 18.2

Sulfur dioxide 24.8

Nitrogen oxides 2,150

Model Data and Graphs – Valve Regulated Lead Acid (VRLA) Battery with Outages

System architecture

Table A- 18: System Architecture

PV Array 30 kW

Generator 1 45 kW

Grid 999,999 kW

Battery 24 Vision 6FM200D

Inverter 40 kW

Rectifier 40 kW

Dispatch strategy Cycle Charging

Cost summary

90

Table A- 19: Cost Summary

Total net present cost $ 335,863

Levelized cost of energy $0.288/kWh

Operating cost $ 9,767/yr

Figure A- 12: Cash Flow Summary

Net Present Costs

Table A- 20: Net Present Costs

Capital Replacement O&M Fuel Salvage Total Component ($) ($) ($) ($) ($) ($)

PV 150,000 0 1,534 0 0 151,534

Generator 1 43,767 0 12,878 37,801 -3,719 90,726

Grid 0 0 0 35,989 0 35,989

91

Vision - 9,600 33,405 6,136 0 48,120 6FM200D 1,021

- Converter 7,639 2,279 0 0 9,494 424

System 211,006 35,683 20,548 73,791 -5,164 335,863

Annualized Costs

Table A- 21: Annualized Costs

Capital Replacement O&M Fuel Salvage Total Component ($/yr) ($/yr) ($/yr) ($/yr) ($/yr) ($/yr)

PV 11,734 0 120 0 0 11,854

Generator 1 3,424 0 1,007 2,957 -291 7,097

Grid 0 0 0 2,815 0 2,815

Vision 6FM200D 751 2,613 480 0 -80 3,764

Converter 598 178 0 0 -33 743

92

System 16,506 2,791 1,607 5,772 -404 26,273

Figure A- 13: Cash Flows

Electrical

Table A- 22: Electrical Component Properties

Production Fraction Component (kWh/yr)

PV array 72,232 63%

Generator 1 7,351 6%

Grid 35,192 31%

Total 114,774 100%

93

Figure A- 14: Monthly Average Electric Production

Table A- 23: Load Parameters

Consumption Fraction Load (kWh/yr)

AC primary load 91,250 100%

Total 91,250 100%

Table A- 24: Electricity Values

Quantity Value Units

Excess electricity 15,604 kWh/yr

Unmet load 0.000134 kWh/yr

Capacity shortage 0.00 kWh/yr

Renewable fraction 0.534

PV

Table A- 25: PV Capacity

Quantity Value Units

Rated capacity 30.0 kW

94

Mean output 8.25 kW

Mean output 198 kWh/d

Capacity factor 27.5 %

Total production 72,232 kWh/yr Table A- 26: PV Outputs

Quantity Value Units

Minimum output 0.00 kW

Maximum output 32.1 kW

PV penetration 79.2 %

Hours of operation 4,397 hr/yr

Levelized cost 0.164 $/kWh

Figure A- 15: PV Output

Generator 1

Table A- 27: Generator 1 Parameters

Quantity Value Units

Hours of operation 287 hr/yr

Number of starts 223 starts/yr

95

Operational life 52.3 yr

Capacity factor 1.86 %

Fixed generation cost 9.26 $/hr

Marginal generation cost 0.257 $/kWhyr Table A- 28: Generator 1 Output

Quantity Value Units

Electrical production 7,351 kWh/yr

Mean electrical output 25.6 kW

Min. electrical output 13.5 kW

Max. electrical output 45.0 kW

Table A- 29: Fuel Consumption Values

Quantity Value Units

Fuel consumption 2,871 L/yr

Specific fuel consumption 0.391 L/kWh

Fuel energy input 28,250 kWh/yr

Mean electrical efficiency 26.0 %

Figure A- 16: Generator 1 output

96

Grid

Table A- 30: Grid Parameters

Quantity Value Units

Hours of operation 5,523 hr/yr

Number of starts 504 starts/yr

Operational life 39.8 yr

Capacity factor 0.000402 %

Fixed generation cost 0.00 $/hr

Marginal generation cost 0.0800 $/kWhyr

Table A- 31: Electrical Output

Quantity Value Units

Electrical production 35,192 kWh/yr

Mean electrical output 6.37 kW

Min. electrical output 0.00453 kW

Max. electrical output 38.8 kW

Table A- 32: Fuel Consumption Values

Quantity Value Units

97

Fuel consumption 35,192 kg/yr

Specific fuel consumption 1.000 kg/kWh

Fuel energy input 35,192 kWh/yr

Mean electrical efficiency 100.0 %

Figure A- 17: Grid Output

Battery

Table A- 33: Battery Configuration

Quantity Value

String size 2

Strings in parallel 12

Batteries 24

Bus voltage (V) 24

Table A- 34: Capacity Parameters

Quantity Value Units

Nominal capacity 57.6 kWh

Usable nominal capacity 34.6 kWh

98

Autonomy 3.32 hr

Lifetime throughput 22,008 kWh

Battery wear cost 0.488 $/kWh

Average energy cost 0.164 $/kWh

Table A- 35: Energy Parameters

Quantity Value Units

Energy in 7,425 kWh/yr

Energy out 5,940 kWh/yr

Storage depletion 2.11 kWh/yr

Losses 1,483 kWh/yr

Annual throughput 6,641 kWh/yr

Expected life 3.31 yr

Figure A- 18: Frequency Histogram

99

Figure A- 19: Monthly Statistics

Figure A- 20: Battery Bank State of Charge

Converter

Table A- 36: Converter Parameters

Quantity Inverter Rectifier Units

Capacity 40.0 40.0 kW

Mean output 6.0 0.4 kW

Minimum output 0.0 0.0 kW

Maximum output 36.2 16.8 kW

Capacity factor 15.0 1.0 %

100

Table A- 37: Energy Usage

Quantity Inverter Rectifier Units

Hours of operation 4,000 4,270 hrs/yr

Energy in 58,476 3,918 kWh/yr

Energy out 52,629 3,330 kWh/yr

Losses 5,848 588 kWh/yr

Figure A- 21 : Inverter Output Power

Figure A- 22: Rectifier Output Power

Emissions

Table A- 38: Emissions Output

Pollutant Emissions (kg/yr)

Carbon dioxide 7,201

Carbon monoxide 247

101

Unburned hydocarbons 27.4

Particulate matter 18.7

Sulfur dioxide 15.2

Nitrogen oxides ,208

102

References

1. . "Dimension Info." from http://www.dimensionsinfo.com/20ft-container-size/.

2. (2012). "Electricity storage: Location, location, location … and cost." from http://www.eia.gov/todayinenergy/detail.cfm?id=6910#tabs_ElecStorage-1.

3. (2014). "Products." from http://www.vfuel.com.au/.

4. (2014). Truck freight rates up 4-5 % in May. The Hindu.

5. A. Landgrebe, F. M. (1989). Atmospheric impacts of and battery technology. Proceedings of the Symposium on Fuel Cells. San Francisco, California.

6. Aaron, D., et al. (2012). "Dramatic performance gains in vanadium redox flow batteries through modified cell architecture." Journal of Power sources 206: 450-453.

7. B. FANG, Y. W., T. ARAI, S. IWASA and M. KUMAGAI (2003). "Development of a novel redox flow battery for electricity storage system." Journal of Applied Electrochemistry 33: 197-203.

8. Bito, A. (2005). Overview of the sodium-sulfur battery for the IEEE Stationary Battery Committee. Power Engineering Society General Meeting, 2005. IEEE.

9. Catalog (2003). Lithium ion technical catalog. G. p. I. Ltd. Taiwan.

10. Catherino, H. A., et al. (2004). "Sulfation in lead–acid batteries." Journal of Power Sources 129(1): 113-120.

11. Ch. Fabjan a, J. G. b., B. Harrer a, L. Jo¨rissen b, C. Kolbeck a, F. Philippi a,G. Tomazic F. Wagner d (2001). "The vanadium redox-battery: an efficient storage unit for 12. photovoltaic systems." Electrochimica Acta 47.

13. Chieng, S., et al. (1992). "Preparation and evaluation of composite membrane for vanadium redox battery applications." Journal of power sources 39(1): 11-19.

14. Dave, A. (2014). India's coal imports rise 20 percent to help fuel new power plants. Reuters. India.

15. Dongyang Chen, S. W., Min Xiao, Dongmei Han, Yuezhong Meng (2010). "Sulfonated poly (fluorenyl ether ketone) membrane with embedded silica rich 16. layer and enhanced proton selectivity for vanadium redox flow battery." Journal of Power Sources 195: 7701-7708.

17. Engineers, U. S. A. C. o. (2012).

18. Eric Yep, S. C. (2014). India Runs Short on Coal, Despite Global Price Slump. Wall Street Journal.

103

19. F. Mohammadia, P. T., S. Zhonga, C. Padestea, M. Skyllas-Kazacosa, "Overcharge in the vanadium redox battery and changes in electrical resistivity and surface functionality of graphite-felt electrodes." Journal of Power Sources 52(1): 61-68.

20. Feng, F., et al. (2001). "Electrochemical behaviour of intermetallic-based metal hydrides used in Ni/metal hydride (MH) batteries: a review." International Journal of Hydrogen Energy 26(7): 725-734.

21. FICCI (2012). Lack of Affordable & Quality Power: Shackling India's Growth Story. New Delhi, BRIEF: Bureau of Research on Industry and Economic Fundamentals.

22. Hagerty, K. "DEEP CYCLE BATTERIES INTRODUCTION." from http://www.altestore.com/howto/Solar-Electric-Power/Design-&-Components/Deep- Cycle-Batteries-Introduction/a85/.

23. Hamann, C. H., et al. (2007). "Electrochemistry. 2nd." Completely Revised and Updated Edition, New York.

24. Hamidi, V., et al. (2010). Smart grid technology review within the transmission and distribution sector. Innovative Smart Grid Technologies Conference Europe (ISGT Europe), 2010 IEEE PES, IEEE.

25. Hiroko Kaneko, K. N., Yutaka Wada, Takamaichi aoki, Akira Negishi and Masayuki Kamimoto (1991). "Vanadium Redox Reactions and Carbon Electrodes for Vanadium Redox Flow Battery." Electrochemica Acia 36(7): 1191-1196.

26. Huang, K.-L., et al. (2008). "Research progress of vanadium redox flow battery for energy storage in China." Renewable energy 33(2): 186-192.

27. Huang, K.-L. L., Xiao-gang Liu, Su-qin Tan, Ning Chen, Li-quan (2008). "Research progress of vanadium redox flow battery for energy storage in China." Renewable Energy 33(2): 186-192.

28. Hwang, G.-J. and H. Ohya (1996). "Preparation of cation exchange membrane as a separator for the all-vanadium redox flow battery." Journal of membrane science 120(1): 55-67.

29. IANS (2014). Power shortage hits Rajasthan. sify news. Rajasthan, Sify Technologies Ltd.

30. Ibrahim, H., et al. (2008). "Energy storage systems—Characteristics and comparisons." Renewable and Sustainable Energy Reviews 12(5): 1221-1250.

31. John, J. S. (2014). Imergy Uses Recycled Vanadium to Cut Materials Costs for Flow Batteries. Silicon Valley, Greentechmedia.

104

32. Kear, G., et al. (2012). "Development of the all-vanadium redox flow battery for energy storage: a review of technological, financial and policy aspects." International Journal of Energy Research 36(11): 1105-1120.

33. Knehr, K. and E. Kumbur (2011). "Open circuit voltage of vanadium redox flow batteries: Discrepancy between models and experiments." Electrochemistry Communications 13(4): 342-345.

34. Larsson, A. (2009). Evaluation of flow battery technology: an assessment of technical and economic feasibility, Massachusetts Institute of Technology.

35. Li, L., et al. (2011). "A Stable Vanadium Redox‐Flow Battery with High Energy Density for Large‐Scale Energy Storage." Advanced Energy Materials 1(3): 394-400.

36. Li L, Z. B., Huang K (2003). Preparation method of electrode for all vanadium flow battery. 37. . Patent Application, CN03159533.2.

38. MacLeman, D. (2009). "A transmission and Distribution Prospective." Electricity Storage Association.

39. Maria Skyllas‐Kazacos, C. P. a. M. C. (1999). "Evaluation of Precipitation Inhibitors for Supersaturated Vanadyl Electrolytes for the Vanadium Redox Battery 40. " Electrochemical and Solid State Letters 2: 121-122.

41. Misra, S. S. (2007). "Advances in VRLA battery technology for telecommunications." Journal of Power Sources 168(1): 40-48.

42. Mohammadi, T. and M. Skyllas-Kazacos (1995). "Use of polyelectrolyte for incorporation of ion-exchange groups in composite membranes for vanadium redox flow battery applications." Journal of power sources 56(1): 91-96.

43. Munshi, P. U., et al. (2009). "ENERGY STORAGE: GAME-CHANGING COMPONENT OF THE FUTURE GRID." Piper Jaffray Investment Research.

44. NASA (2014). "Surface meteorology and Solar Energy." 2014, from https://eosweb.larc.nasa.gov/cgi-bin/sse/[email protected]+s01#s01.

45. Nieto, A., et al. (2014). "Combined Life Cycle Assessment and Costing Analysis Optimization Model Using Multiple Criteria Decision Making in Earth-Resource Systems." Natural Resources 2014.

46. Nourai, A. (2002). Large-scale electricity storage technologies for energy management. Power Engineering Society Summer Meeting, 2002 IEEE.

47. NREL (2005). "Getting Started Guide " Homer Version 2.1.

48. NREL (2014). Hybrid Optimization of Multiple Energy Resources, HOMER Energy LLC 105

49. Oei, D.-G. (1985). "Permeation of vanadium cations through anionic and cationic membranes." Journal of applied electrochemistry 15(2): 231-235.

50. Ponce de León, C., et al. (2006). "Redox flow cells for energy conversion." Journal of Power Sources 160(1): 716-732.

51. Preetesh U. Munshi, J. P., Elaine S. Kwei (2009). Energy Storage: Game Changing Component of the Future Grid. Guides for the Journey, PiperJaffray Investment Research.

52. Rychcik, M. and M. Skyllas-Kazacos (1988). "Characteristics of a new all-vanadium redox flow battery." Journal of Power Sources 22(1): 59-67.

53. Shigematsu, T. (2011). "Redox flow battery for energy storage." SEI Technical Review 73.

54. Silberberg, M. (2006). Chemistry: The Molecular Nature of Matter and Change, McGraw-Hill Education.

55. Skyllas-Kazacos, M. (2003). "Novel vanadium chloride/polyhalide redox flow battery." Journal of Power Sources 124(1): 299-302.

56. Skyllas-Kazacos, M., et al. (2011). "Progress in flow battery research and development." Journal of the Electrochemical Society 158(8): R55-R79.

57. Skyllas-Kazacos, M., et al. (1991). "Characteristics and performance of 1 kW UNSW vanadium redox battery." Journal of Power Sources 35(4): 399-404.

58. Skyllas-Kazacos, M., et al. (2008). Membrane properties and behaviour in the Generation 2 Vanadium Bromide Redox Flow batteries. Materials Forum.

59. Skyllas-Kazacos, M., et al. (2003). Metal bromide redox flow cell, PCT Application, PCT/GB2003/001757, April.

60. Skyllas‐Kazacos, M., et al. (2010). "Recent advances with UNSW vanadium‐based redox flow batteries." International Journal of Energy Research 34(2): 182-189.

61. Skyllas‐Kazacos, M. K., George Poon, Grace Verseema, Hugh (2010). "Recent advances with UNSW vanadium‐based redox flow batteries." International Journal of Energy Research 34(2): 182-189.

62. Spellman, K., et al. (2013). "Economic Report on Vanadium Redox Flow Battery with Optimization of Flow Rate."

63. Still, J. (Jul 9, 2014). "Imergy Power Systems Achieves Technological Breakthrough in Energy Storage: Flow Batteries Made from Recycled Vanadium."

106

64. Sukkar, T. and M. Skyllas-Kazacos (2003). "Water transfer behaviour across cation exchange membranes in the vanadium redox battery." Journal of membrane science 222(1): 235-247.

65. Tarn Yates, B. H. (2012). "Levelized Cost of Energy." from http://solarprofessional.com/articles/finance-economics/levelized-cost-of- energy/page/0/6.

66. Thackeray, M. M., et al. (2000). "Science and Applications of Mixed Conductors for Lithium Batteries." MRS Bulletin 25(03): 39-46.

67. Theresa Sukkar, M. S.-K. (2003). "Water transfer behaviour across cation exchange membranes in the vanadium redox battery." Journal of Membrane Science 222.

68. Tian, B., et al. (2004). "Proton conducting composite membrane from Daramic/Nafion for vanadium redox flow battery." Journal of membrane science 234(1): 51-54.

69. Toledo, O. M., et al. (2010). "Distributed photovoltaic generation and energy storage systems: A review." Renewable and Sustainable Energy Reviews 14(1): 506-511.

70. Vafiadis, H. and M. Skyllas-Kazacos (2006). "Evaluation of membranes for the novel vanadium bromine redox flow cell." Journal of membrane science 279(1): 394-402.

71. Vafiadis, H. and M. Skyllas-Kazacos (2006). "Evaluation of membranes for the novel vanadium bromine redox flow cell." Journal of Membrane Science 279(1-2): 394-402.

72. WeatherSpark (2013). "Average Weather." 2014, from https://weatherspark.com/averages/33936/Jaipur-Rajasthan-India.

73. Widatalla, A. M. and H. Zinko "Designing a Photovoltaic Solar Energy System for a Commercial Building Case Study: Rosa Park Hotel in Khartoum-Sudan."

74. Wikipedia (2014). "Net Present Value." from http://en.wikipedia.org/wiki/Net_present_value.

75. Ye Qin, J.-G. L., You-Ying Di, Chuan-Wei Yan, Chao-Liu Zeng, and Jia-Zhen Yang (2010). "Thermodynamic Investigation of Electrolytes of the Vanadium Redox Flow 76. Battery (II): A Study on Low-Temperature Heat Capacities and Thermodynamic 77. Properties of VOSO4 ·2.63H2O(s)." J. Chem. Eng. Data 55: 1276-1279.

78. Yue, L., et al. (2010). "Highly hydroxylated carbon fibres as electrode materials of all- vanadium redox flow battery." Carbon 48(11): 3079-3090.

79. Zhao, P., et al. (2006). "Characteristics and performance of 10kW class all-vanadium redox-flow battery stack." Journal of Power Sources 162(2): 1416-1420.

80. Zhu, H. Q., et al. (2008). "Graphite–carbon nanotube composite electrodes for all vanadium redox flow battery." Journal of Power Sources 184(2): 637-640.

107