Full Document Text
Publ ic Interest Energy Research (P IER) Program
FINAL PROJECT REPORT
2020 STRATEGIC ANALYSIS OF
ENERGY STORAGE IN CALIFORNIA
Prepared for: California Energy Commission
Prepared by: University of California, Berkeley School of Law
University of California, Los Angeles
University of California, San Diego
NOVEMBER 2011
CEC ‐500 ‐2011 ‐047
Prepared by:
University of California, Berkeley School of Law
Jessica Intrator
Ethan Elkind, Principal Investigator (joint
appointment at UCLA School of Law)
Steven Weissman
Morgan Sawchuk
Emily Bartlett
University of California, Los Angeles
Andris R. Abele, Principal Investigator
Bruce S. Dunn, Ph.D.
Tsu-Chin Tsao, Ph.D.
Rita Blaik
Christopher Lim
David Luong
Leland Smith
University of California, San Diego
Byron Washom, Principal Investigator
Contract Number: 500-02-004
Work Authorization: MRA-02-088
Prepared for:
California Energy Commission
Cathy Turner
Contract Manager
Avtar Bining, Ph.D.
Project Manager
Mike Gravely
Office Manager
Energy Systems Research Office
Laurie ten Hope
Deputy Director
ENERGY RESEARCH AND DEVELOPMENT DIVISION
Robert P. Oglesby
Executive Director
DISCLAIMER
This report was prepared as the result of work sponsored by the California Energy Commission. It does not
necessarily represent the views of the Energy Commission, its employees or the State of California. The Energy
Commission, the State of California, its employees, contractors and subcontractors make no warrant, express or
implied, and assume no legal liability for the information in this report; nor does any party represent that the uses
of this information will not infringe upon privately owned rights. This report has not been approved or disapproved
by the California Energy Commission nor has the California Energy Commission passed upon the accuracy or
adequacy of the information in this report.
ACKNOWLEDGMENTS
The project titled 2020 Strategic Analysis of Energy Storage in California is funded by the California
Energy Commission’s Public Interest Energy Research (PIER) Program, through the California
Institute for Energy and Environment.
The authors would like to thank Dr. Carl Blumstein, Dr. Merwin Brown, and Mr. Larry Miller
with the California Institute for Energy and Environment for their help with this project, as well
as Jan Kleissl, Ph.D. of the University of California, San Diego. Project support provided by
Kristen Bargmeyer at the University of California, Berkeley School of Law is gratefully
acknowledged. In addition, the authors thank Avtar Bining and Mike Gravely from the
California Energy Commission for their dedication to the project and input on the drafts.
This report benefitted greatly from its Project Advisory Committee, which offered crucial
guidance and input during an in‐person meeting and through electronic feedback on multiple
drafts. The authors are grateful for the committee members’ dedicated support and insightful
comments. Any errors and all policy conclusions contained in this document are strictly those
of the authors. The Project Advisory Committee for the “2020 Strategic Analysis of Energy
Storage in California” consists of the following academic, nonprofit, public sector, and industry
experts:
• Victor Babbit, RES Americas
• Matt Barmack, Calpine Corporation
• Deborah Behles, Environmental Law and Justice Clinic, Golden Gate University
School of Law
• Nicole A. Blake, Consumer Federation of California
• Ed Cazalet, MegaWatt Storage Farms, Inc.
• Larry Chaset, Sustainable Energy Futures
• Stephen R. Clarke, Applied Intellectual Capital
• Carlos Coe, Xtreme Power Solutions, Inc.
• Michael Colvin, California Public Utilities Commission
• Eric Cutter, Energy and Environmental Economics, Inc.
• Paul Denholm, National Renewable Energy Laboratory
• Jim Detmers, Power Systems Resources, Inc.
• Marcelo E. Di Paolo, Los Angeles Department of Water and Power
• J. Douglas Divine, Eagle Crest Energy Company
• Steve Eckert, Brookfield Renewable Power Inc.
• Jack Ellis, Resero Consulting, Inc.
• Jim Eyer, Distributed Utility Association
• Adam Green, Solar Reserve
• Cris Grieco, Gravity Power, LLC
i
• Ross Guttromson, Sandia National Laboratories
• Dave Hawkins, KEMA, Inc.
• Joe Heinzmann, Altairnano
• Eric Hsieh, A123 Systems, Inc.
• Judith Judson, Beacon Power Corporation
• David Kates, Elsinore Valley Municipal Water District (EVWMD)
• Praveen Kathpal, AES Energy Storage, LLC
• Chris King, eMeter Strategic Consulting
• Samantha King, International Quality and Productivity Center (IQPC)
• Monica Lamb, First Solar, Inc.
• Janice Lin, California Energy Storage Alliance
• Chris Marnay, Lawrence Berkeley National Laboratory
• Jim McIntosh, California Independent System Operator
• Alex Morris, Southern California Edison
• David Nemtzow, Ice Energy, Inc.
• Robert Pratt, Pacific Northwest National Laboratory
• Dan Rastler, Electric Power Research Institute
• Adam Rauwerdink, SustainX, Inc.
• Mark Rawson, Sacramento Municipal Utilities District
• Stacey Reineccius, Power and Energy Consulting
• Johannes Rittershausen, Southern California Edison
• Gwen Rose, Vote Solar
• Heather Sanders, California Independent System Operator
• Klaus Sch??ss, KS Engineers
• Chris Shelton, AES Energy Storage, LLC
• Kevin Snyder, HDR/DTA Inc.
• John Eric Thalman, Pacific Gas and Electric Company
• Sarah Thomas, California Public Utilities Commission, Division of Ratepayer
Advocates
• William Torre, San Diego Gas & Electric Company
• Rick Winter, Primus Power Corporation
ii
PREFACE
The California Energy Commission Public Interest Energy Research (PIER) Program supports
public interest energy research and development that will help improve the quality of life in
California by bringing environmentally safe, affordable, and reliable energy services and
products to the marketplace.
The PIER Program conducts public interest research, development, and demonstration (RD&D)
projects to benefit California.
The PIER Program strives to conduct the most promising public interest energy research by
partnering with RD&D entities, including individuals, businesses, utilities, and public or
private research institutions.
PIER funding efforts are focused on the following RD&D program areas:
• Buildings End‐Use Energy Efficiency
• Energy Innovations Small Grants
• Energy‐Related Environmental Research
• Energy Systems Integration
• Environmentally Preferred Advanced Generation
• Industrial/Agricultural/Water End‐Use Energy Efficiency
• Renewable Energy Technologies
• Transportation
The report 2020 Strategic Analysis of Energy Storage in California (Contract Number 500‐02‐004,
Work Authorization Number MRA‐02‐088), conducted by the University of California, Berkeley
School of Law; University of California, Los Angeles; and University of California, San Diego,
contributes to PIER’s Energy Systems Integration program area.
When the source of a table, figure or photo is not otherwise credited, it is the work of the author
of the report.
For more information about the PIER Program, please visit the Energy Commission’s website at
www.energy.ca.gov/research/ or contact the Energy Commission at 916‐327‐1551.
iii
ABSTRACT
As California progresses towards its goal of 33 percent renewable electricity generation , the
potential for energy storage to help integrate renewable resources and maintain a reliable and
efficient electric grid takes on great significance. In 2010, the California Legislature enacted
Assembly Bill 2514 (Skinner, Chapter 469, Statutes 2010), directing the California Public Utilities
Commission (CPUC) to convene a proceeding to define energy storage procurement targets, if
any, for investor‐owned utilities. Under the statute, similar targets, on a slightly larger scale,
would also be required for publicly owned utilities. This report presents a strategic analysis of
energy storage for California by 2020. The report assesses current energy storage technologies,
discusses the diverse policies affecting deployment in California, and outlines critical
technology gaps, future research needs, and policy reforms. It also provides a reference
framework for the Energy Commission, CPUC, and other regulatory agencies to use as they
develop solutions for how commercially ready energy storage technologies can be cost‐
effectively applied in California to reduce costs to ratepayers, reduce emissions from fossil fuel
generation, and enable and accelerate the implementation of more renewable generation and its
integration in California’s electricity system. The report also provides a roadmap for the
California Energy Commission to use when allocating public interest energy research funds for
developing, improving and demonstrating energy storage technologies.
Keywords: energy storage, renewable energy, renewable portfolio standards, AB 2514, variable
renewable energy integration, distributed generation, distributed energy resources, smart grid,
peak load shaving, permanent load shifting, ancillary services, frequency regulation
Please use the following citation for this report:
Andris Abele, Ethan Elkind, Jessica Intrator, Byron Washom, et al (University of California,
Berkeley School of Law; University of California, Los Angeles; and University of
California, San Diego) 2011, 2020 Strategic Analysis of Energy Storage in California,
California Energy Commission. Publication Number: CEC-500-2011-047.
iv
TABLE OF CONTENTS
ACKNOWLEDGMENTS ........................................................................................................................... i
PREFACE…...……………………………………………………………………………………………iii
ABSTRACT……………… ........................................................................................................................ iv
TABLE OF CONTENTS ............................................................................................................................ v
LIST OF FIGURES ..................................................................................................................................... ix
LIST OF TABLES ........................................................................................................................................ x
EXECUTIVE SUMMARY .......................................................................................................................... 1
Summary of Energy Storage Technologies ........................................................................................ 1
2020 Energy Storage Vision and Policy Priorities ............................................................................. 6
CHAPTER 1: Introduction: The Importance of Energy Storage and Technology Status ............. 13
1.1 Electrochemical Energy Storage ............................................................................................. 16
1.1.1 Lead‐Acid Batteries........................................................................................................ 17
1.1.2 Nickel‐Based Batteries ................................................................................................... 18
1.1.3 Lithium Ion Batteries ..................................................................................................... 19
1.1.4 Flow Batteries ................................................................................................................. 22
1.1.5 Sodium‐Beta Alumina Batteries ................................................................................... 25
1.1.6 Metal Air Batteries ......................................................................................................... 28
1.1.7 Electric Double Layer Capacitors ................................................................................ 30
1.1.8 Pseudocapacitors ........................................................................................................... 32
1.1.9 Hybrid Capacitors .......................................................................................................... 33
1.2 Mechanical Energy Storage .................................................................................................... 34
1.2.1 Pumped Hydro ............................................................................................................... 34
1.2.2 Compressed Air Energy Storage ................................................................................. 40
1.2.3 Flywheels ........................................................................................................................ 48
1.3 Thermal Energy Storage .......................................................................................................... 56
1.3.1 Solar Thermal ................................................................................................................. 56
1.3.2 Thermal Storage for Heating, Ventilation, and Air Conditioning (HVAC) .......... 58
1.4 Hydrogen as an Energy Storage System ............................................................................... 61
v
1.4.1 Description of Technology ............................................................................................ 61
1.4.2 Technology Status .......................................................................................................... 64
1.4.3 Technology Limitations ................................................................................................ 66
1.4.4 R&D Needs and Opportunities .................................................................................... 66
1.5 General Cost Data for Energy Storage Technologies .......................................................... 67
CHAPTER 2: Regulatory and Policy Framework............................................................................... 70
2.1 Energy Storage Policy and Regulatory Landscape ............................................................. 70
2.1.1 California’s Clean Energy Policy Push ....................................................................... 70
2.1.2 Legislative Mandates: AB 2514, AB 32, and California’s RPS ................................. 74
2.1.3 Institutional and Regulatory Action: Regulatory Proceedings and Rate Cases .... 77
2.1.4 Comparative Cases: Energy Storage Policy Out of State and Abroad ................... 97
CHAPTER 3: Technology Gaps, Research and Development Needs, and Policy Challenges ... 104
3.1 Technology Status Findings: Gaps and R&D Priorities .................................................... 104
3.2 Policy and Regulatory Findings: Barriers and Challenges ............................................... 105
3.2.1 Valuation and Cost Recovery ..................................................................................... 105
3.2.2 Awareness and Collaboration .................................................................................... 111
3.2.3 Competing Policy Priorities........................................................................................ 112
CHAPTER 4: Applications and Scenarios Analyses for Energy Storage in California –
Background, Purpose, and Approach ....................................................................... 114
4.1 California’s Grid Needs and Applications Selection ........................................................ 115
4.2 Possible Drivers for Energy Storage Deployment in California ...................................... 117
4.2.1 Procurement Targets ................................................................................................... 117
4.2.2 Tariff and Market Reform ........................................................................................... 118
4.2.3 Loading Order and Resource Adequacy Requirements......................................... 119
4.2.4 Financial Incentives ..................................................................................................... 120
4.2.5 Targeted Research and Development Funding ....................................................... 123
CHAPTER 5: Applications and Scenarios Analyses for Energy Storage in California ............... 132
5.1 Exemplary Application #1: Area and Frequency Regulation .......................................... 132
5.1.1 Area and Frequency Regulation: Definition and Application ............................... 133
vi
5.1.2 Area and Frequency Regulation: Technology Status and Market Fit ................... 134
5.1.3 Area and Frequency Regulation: Technology Potential ......................................... 137
5.1.4 Area and Frequency Regulation: Regulatory Status and Policy Changes ........... 138
5.1.5 Area and Frequency Regulation: Benefits and Costs .............................................. 140
5.1.6 Area and Frequency Regulation: Recommended Next Steps and Milestones .... 145
5.2 Exemplary Application #2: Renewables Grid Integration ................................................ 146
5.2.1 Renewables Grid Integration: Definition and Application .................................... 148
5.2.2 Renewables Grid Integration: Technology Status and Fit ...................................... 150
5.2.3 Renewables Grid Integration: Technology Potential .............................................. 153
5.2.4 Renewables Grid Integration: Regulatory Status and Policy Changes ................ 156
5.2.5 Renewables Grid Integration: Benefits and Costs ................................................... 160
5.2.6 Milestones and Next Steps for Energy Storage for Renewables Grid Integration
………………………………………………………………………………………….165
5.3 Exemplary Application #3: Community Energy Storage / Distributed Energy Storage
Systems (DESS) ...................................................................................................................... 167
5.3.1 Community Energy Storage: Definition and Application ..................................... 167
5.3.2 Community Energy Storage: Technology Status and Fit ....................................... 168
5.3.3 Community Energy Storage: Technology Potential ................................................ 170
5.3.4 Community Energy Storage: Regulatory Status and Policy Changes .................. 171
5.3.5 Community Energy Storage: Benefits and Costs ..................................................... 173
5.3.6 Milestones and Next Steps for Community Energy Storage ................................. 177
CHAPTER 6: 2020 Energy Storage Vision and Recommendations ............................................... 179
6.1 Energy Storage Vision for 2020 ............................................................................................ 179
6.1.1 33 percent or More Renewable Energy Generation Under the 2020 Renewables
Portfolio Standard……………. ..................................................................................................... 179
6.1.2 Implementation of the Smart Grid ............................................................................ 180
6.1.3 Proliferation of Microgrids ......................................................................................... 180
6.1.4 Increased Residential Needs for On‐Site Solar, Electric Vehicles, and Home Area
Networks. ........................................................................................................................................ 181
vii
6.1.5 Opening of Electricity Markets to Competition From Energy Storage and
Incentive Programs ........................................................................................................................ 182
6.1.6 Natural and Human‐Caused Disasters ..................................................................... 183
6.1.7 Likely Price Decreases for Various Energy Storage Technologies ........................ 183
6.1.8 Future Research Needs to Implement the Vision .................................................... 183
6.2 Policy Drivers for Energy Storage Deployment ................................................................ 185
6.2.1 CPUC and AB 2514 Target Potential ......................................................................... 185
6.2.2 Additional Policies for Energy Storage Deployment .............................................. 187
6.2.3 Agency Coordination .................................................................................................. 187
6.3 Conclusion .............................................................................................................................. 187
Glossary………….. ................................................................................................................................. 188
References……… ................................................................................................................................... 191
viii
LIST OF FIGURES
Figure 1: Comparison of Power and Discharge Time for Various Energy Storage Devices ......... 15
Figure 2: Technical Maturity of Various Energy Storage Systems .................................................... 16
Figure 3: Principles of Sodium‐Sulfur Battery Operation .................................................................. 26
Figure 4: Structure of Li‐air Battery ....................................................................................................... 29
Figure 5: Cutaway Diagram of a Typical Pumped Hydro Plant ....................................................... 35
Figure 6: Proposed Pumped Storage Projects in the United States ................................................... 37
Figure 7: Gravity Power ModuleTM........................................................................................................ 39
Figure 8: CAES Plant With Underground Storage .............................................................................. 41
Figure 9: Aerial View of CAES Plant in Huntorf, Germany .............................................................. 42
Figure 10: Aerial View of CAES Plant in McIntosh, Alabama ........................................................... 42
Figure 11: Conceptual AA‐CAES System ............................................................................................. 44
Figure 12: Conceptual 2nd Generation CAES With Air Injection ..................................................... 46
Figure 13: Schematic of CAES With Bottoming Cycle and Humidification .................................... 47
Figure 14: Schematic of AA‐CAES With Isobaric Storage .................................................................. 48
Figure 15: TES HVAC Unit With Cooling Coil Integrated Into Existing A/C ................................. 59
Figure 16: Hydrogen Electrolyzer/Fuel Cell Energy Storage System. .............................................. 62
Figure 17: Concept of a Neighborhood Using DESS ......................................................................... 168
ix
x
LIST OF TABLES
Table 1: Comparison of Various Flow Batteries .................................................................................. 24
Table 2: Flywheel Shape Factors ............................................................................................................ 51
Table 3: ARRA‐Funded Energy Storage Technology Demonstration Projects ............................. 124
Table 4: Area and Frequency Regulation Performance Targets for Energy Storage .................... 134
Table 5: Energy Storage Options for Frequency Regulation ............................................................ 135
Table 6: Adjustable Speed Pumped Storage Characteristics ............................................................ 136
Table 7: Renewables Integration Short‐Term Metrics ....................................................................... 149
Table 8: Renewable Energy Time Shift Applications – Long‐Duration Metrics ........................... 150
Table 9: Bulk Energy Storage for Long‐Duration Renewables Integration ................................... 152
Table 10: Application Value Estimates ............................................................................................... 163
Table 11: Distributed Energy Storage System Metrics ...................................................................... 169
Table 12: DESS Uses and Benefits ........................................................................................................ 174
EXECUTIVE SUMMARY
On April 12, 2011, California Governor Jerry Brown approved Senate Bill 2 Simitian Energy:
renewable energy resources. Through this bill, California has set a 2020 goal of 33 percent
electricity generation from renewable energy resources. The energy storage technologies’ role
to assist in the integration of variable renewable resources and the maintenance of a reliable and
efficient electric grid takes on even greater importance as California progresses towards
achieving this goal. Numerous policy and planning documents at the state and federal level
have cited the importance of evaluating the potential benefits of energy storage in the
generation, transmission, and distribution equation. On September 29, 2010, Assembly Bill 2514
(Skinner, Chapter 469, Statutes 2010) was signed into law directing the California Public
Utilities Commission (CPUC) to convene a proceeding to determine energy storage
procurement targets, if any, for investor‐owned utilities. Under the statute, similar targets
would be required for publicly owned utilities on a slightly later time frame. Meanwhile,
electricity industry participants and technology developers have renewed efforts to develop,
demonstrate, and implement a wide range of energy storage technologies.
A strategic analysis of the various energy storage technologies that may present options for
California by 2020 has been conducted. The current status of energy storage technologies and
many diverse policies affecting the deployment of energy storage in California has been
analyzed, and critical technology gaps and research needs and policy reforms have been
identified. This analysis provides a reference framework for the Energy Commission, CPUC,
and other regulatory agencies to use as they work to develop commercially ready energy
storage technologies that can be cost‐effectively deployed in California to reduce costs to
ratepayers, means to reduce emissions from fossil fuel generation, and opportunities to enable
and accelerate the implementation of 33 percent renewable generation and its integration in
California’s electricity system.
Summary of Energy Storage Technologies
Overall Findings
Grid operators are already deploying large energy storage technologies, such as pumped
hydroelectric and compressed air energy storage (CAES). There are significant challenges that
must be solved in order to achieve desired storage goals. These goals include: finding
appropriate sites for these facilities, obtaining necessary permits from various agencies and
levels of government, overcoming regulatory hurdles associated with environmental review,
meeting high capital costs for construction, and addressing a lack of access to transmission
lines.
Manufacturers are demonstrating modular technologies, such as flywheels and various forms of
batteries, in grid applications. Key challenges relate primarily to cost (although the modularity
of the technologies may offer promise for cost reduction through volume production), to the
ability to manufacture and deploy on a large‐scale basis, to durability, and because of limited
experience in grid applications.
1
Electrochemical Energy Storage
Batteries take in electricity from another producing source, convert the electricity to chemical
energy, and store it as a liquid or solution. When operators need energy from the battery, an
electric charge chemically converts the energy back into electrons, which then move back into a
power line on the electric grid.
Batteries used to store power from renewable energy sources must be reliable, durable, and
safe. Ultimately, affordability will be a key to widespread deployment. There are several
promising battery technologies for grid energy storage applications including advanced lead‐
acid, lithium‐ion, flow, and sodium‐sulfur batteries.
Advanced Lead‐Acid
During discharge in a traditional lead‐acid battery, sulfuric acid reacts with the lead anode
(positive electrode) and cathode (negative electrode) to create lead sulfate. The process reverses
during charge. This conversion produces a short, powerful burst of energy, such as needed to
jump start a vehicle. Over time, a lead‐acid battery can lose its charge due to the gradual
crystallization and buildup of lead sulfate within the batteryʹs core. The corrosive acid also can
eat away at a batteryʹs core.
Lead‐acid batteries are a mature and proven technology in use in a number of applications
including frequency regulation, bulk energy storage for variable renewable energy integration,
and distributed energy storage systems. Technology development of lead‐acid battery
technology is ongoing. Researchers have found that adding carbon to the battery seems to
minimize or prevent the detrimental crystallization from occurring, thus improving the life
cycle and overall lifespan of the battery. This technology has potential for storing renewable
energy, but engineers must work to understand the technology’s limitations and to find ways to
bring down the cost.
Lithium‐Ion
In a Lithium‐ion (Li‐ion) battery cell, positively charged lithium ions migrate through a liquid
electrolyte (fluids that conduct electricity) while electrons flow through an external circuit. Both
move back and forth from one side to the other. This movement creates and stores energy. Li‐
ion batteries store energy in various compounds, composed of layers of different elements, such
as lithium, manganese, and cobalt.
Li‐ion batteries are most commonly found in consumer products and electric vehicles. The
relatively high energy and power capacity offered by Li‐ion batteries, when compared to other
technologies, has made Li‐ion batteries the most promising option for transportation
applications such as electric vehicles. Developers are considering and demonstrating Li‐ion
batteries in the same applications as lead‐acid batteries.
Although Li‐ion batteries have been a success for small electronics such as cell phones and
laptop computers, larger versions are expensive, prone to overheating, and susceptible to
electrical shorting. While engineers have made substantial progress over recent years toward
2
improving this technology, they will need to make further advances to extend life, improve
safety, and reduce materials cost for this to be an attractive alternative for stationary
applications.
Flow Batteries
A flow battery is a type of rechargeable battery that stores electrical energy in two tanks of
electrolytes. When operators need energy, they pump liquid from one tank to another. During
this slow and steady process, the technology converts the chemical energy from the electrolyte
to electrical energy. When operators need to store energy, they reverse the process. The size of
the tank and the amount of electrolyte the battery can hold determine the amount of energy the
battery can store.
Flow batteries may be good candidates for backup energy storage up to 12 hours. They may
also support integration of variable renewable energy. This technology has potential for grid
applications if developers can manufacture it in a variety of sizes and make it portable and
more affordable.
Sodium‐Sulfur
The sodium‐sulfur battery uses sulfur combined with sodium to reversibly charge and
discharge, using sodium ions layered in aluminum oxide within the batteryʹs core. The battery
shows potential to store lots of energy in a small space. In addition, its high energy density and
rapid rate of charge and discharge make it an attractive candidate for applications that require
short, potent bursts of energy.
Sodium‐sulfur batteries are a commercial energy storage technology with applications in
electric utility distribution grid support, wind power integration, and high‐value electricity
services. However, materials are expensive, and safety concerns remain with respect to the
high operating temperature of the battery. Researchers believe that modifying the shape of the
battery can improve efficiency, lower the operating temperature, and reduce cost.
Mechanical Energy Storage
Operators can store energy in water pumped to a higher elevation using pumped storage
methods, in compressed air, or in spinning flywheels.
Pumped Hydroelectric
Pumped hydroelectric uses two water reservoirs, separated vertically. During off‐peak hours,
operators pump water from the lower reservoir to the upper reservoir. The operators reverse
the water flow to generate electricity.
Pumped hydroelectric energy storage is a large, mature, and commercial utility‐scale
technology that utilities use at many locations in the United States and around the world. This
application has the highest capacity of the various energy storage technologies that experts have
assessed. However, pumped storage plants generally entail long construction times and high
capital expenditure for both construction of the plants and needed transmission lines.
3
Compressed Air Energy Storage
Compressed air energy storage technology stores low cost off‐peak energy, in the form of
compressed air, typically in an underground reservoir. Operators then heat the air with the
exhaust heat of a standard combustion turbine and release it during peak load hours.
Operators convert the heated air to energy through expansion turbines to produce electricity.
Compressed air energy storage systems suffer from reduced roundtrip efficiency associated
with the cooling/reheating process. Air cooling between compression stages, although
necessary, results in a loss of heat energy. Compressed air energy storage systems also produce
carbon dioxide (CO2) emissions from the reheating process, usually performed by direct
combustion with natural gas. Some compressed air energy storage systems under
development, such as advanced adiabatic compressed air energy storage, use a thermal energy
storage unit that absorbs heat from the hot compressed air and saves the heat energy for later
use to reheat the air before expansion, thus avoiding CO2 emissions.
Flywheels
Flywheel energy storage works by accelerating a rotor (flywheel) to a very high speed,
maintaining the energy in the system as rotational energy. When operators extract energy from
the system, they reduce the flywheelʹs rotational speed as a consequence of the principle of
conservation of energy. Adding energy to the system correspondingly results in an increase in
flywheel speed.
Developers have matured flywheel technology through the advent of strong, lightweight
materials, microelectronics, and magnetic bearing systems. Manufacturers are currently
developing and demonstrating megawatt‐scale flywheel plants with cumulative capacities of 20
megawatts to maintain a uniform quality electricity supply often also termed as frequency
regulation applications. Overall, manufacturers have proven flywheels to be an ideal form of
energy storage due to their high efficiency, long life cycle, wide range of operating temperature,
and higher power and energy density on both a mass and volume basis. FESs still present high
costs and technology limitations, including modest energy storage capacity and efficiency losses
associated with the bearings.
Thermal Energy Storage
Thermal energy storage comprises a number of technologies that store thermal energy in energy
storage reservoirs for later use. Operators can employ them to balance energy demand between
daytime and nighttime. Operators maintain the thermal reservoir at a temperature above
(hotter) or below (colder) than that of the ambient environment. The applications include
concentrating sunlight to produce electricity from solar thermal energy during non‐solar
periods and the production of ice, chilled water, or salt solution at night, or hot water, which the
devices use to cool / heat environments during the day.
4
Solar Thermal Storage Integration
The integration of thermal energy storage with solar energy offers a direct grid application for
thermal energy storage. Unlike solar photovoltaic (PV) generation, concentrating solar power
uses the thermal energy of sunlight to generate electricity. An advantage of concentrating solar
power is the potential for storing solar thermal energy to be subsequently used during non‐
solar periods and to dispatch it as needed. Thermal energy storage allows concentrating solar
power to achieve higher annual capacity factors from 25 percent without thermal storage up to
70 percent or more with thermal storage. Large concentrating solar power facilities using
molten salt energy storage are in construction and/or operation in Spain and the United States.
Plans are underway for facilities offering thousands of megawatts of additional generating
capacity that will also use this storage technology.
Recent and ongoing improvements in solar thermal generation technologies, coupled with the
need for more renewable sources of energy, have caused an increased interest in concentrating
solar power. The key challenges lie in further cost reductions and perfecting designs to store
solar heat later into the peak electrical period.
Thermal Storage for Heating, Ventilation, and Air Conditioning (HVAC)
In thermal energy storage systems, a device chills a storage medium during periods of low
cooling demand and then uses the stored cooling to meet air‐conditioning load or process
cooling loads. The system consists of a storage medium, such as a water/ice system in a tank, a
packaged chiller or built up refrigeration system, and interconnecting piping, pumps, and
controls. Heating, ventilation, and air conditioning (HVAC) thermal energy storage systems
shift cooling energy use to nonpeak times.
Thermal energy storage for commercial HVAC systems is a mature technology. The key to
maximizing the effectiveness of such systems to shift cooling load and thus support the electric
grid is appropriate engineering design and implementation.
Hydrogen as an Energy Storage System
Hydrogen as an energy storage system involves four processes. First, a device must produce
hydrogen. In a grid energy storage application, the most appropriate production technology is
the electrolysis of water using electricity. Second, after electrolysis produces the hydrogen, a
device must store it, either in gaseous or liquid form. Third, in many instances, the hydrogen
must be transported by truck or pipeline to a distant location. Fourth, to return electric power
to the grid, the devices must convert hydrogen to electricity by either a fuel cell or a combustion
engine or gas turbine generator.
The primary limitations of hydrogen energy storage systems include the maturity of the fuel
cell technology; the durability of fuel cells and electrolyzers; and the capital cost of fuel cells,
electrolyzers, and, to a lesser extent, storage vessels. The scale of fuel cells and electrolyzers
with respect to grid storage applications and the efficiency of fuel cells and electrolyzers also
limits the use of the technology, with roundtrip (electricity into the system to produce hydrogen
5
relative to the electricity produced by the hydrogen fuel cell) energy efficiencies of 31‐35
percent.
Recommendations
Suggested Research and Development Priorities
• Bulk energy storage demonstrations for variable renewable energy integration (for
example, pumped hydro, concentrating solar power, and solar thermal).
• Field demonstrations of modular energy storage technologies (for example, batteries,
flywheels) in various grid applications.
• Evaluation/demonstration of aggregated storage, for example Thermal energy storage
HVAC or electric vehicle (EV) batteries, especially in a smart grid scenario.
• Develop simulations, analytical tools, and intelligent control systems for planning,
designing, and dispatching energy storage devices for multiple applications and
benefits.
• Quantification of costs and benefits of energy storage in grid applications.
• Modeling of the impact of 33 percent renewable energy on California’s electricity grid to
determine needs for energy storage to support the grid, including sensitivity analysis to
address cost variables of storage and other needed energy resources, environmental
impacts, and emerging smart grid performance enhancements.
2020 Energy Storage Vision and Policy Priorities
As policy makers anticipate future grid needs and how energy storage can address them, they
should consider a number of key developments that are likely to transpire between now and
2020 and that will influence the role of energy storage.
33 Percent or More Renewable Energy Generation Under the 2020 Renewable Portfolio
Standard
The California Renewable Energy Resources Act (Senate Bill 2, Simitian Energy: renewable
energy resources), signed by Governor Jerry Brown into law in April 2011, requires California
to procure 33 percent of its electricity from renewable sources by 2020. However, integrating
these variable and intermittent sources, such as solar and wind, will require an array of policy
mechanisms. Policy makers should explore all cost‐effective options, given California’s
environmental and greenhouse gas goals and need to avoid curtailment of renewable resources.
Studies indicate that California may require between 3,000 to 4,000 megawatts of fast‐acting
energy storage by 2020 to integrate the projected increase in renewable energy.
Implementation of the Smart Grid
The deployment of Smart Grid, a network of interconnected electronically enhanced and
digitally communicated and controlled electricity generation and transmission and distribution
systems, in California may both improve and lessen the need for energy storage. Because
utilities may be able to reduce end‐user load with the new, smarter technology, they may not
6
require energy storage to shave peak load or maximize the efficiency of renewable energy
electricity production. As customers, particularly from the business and industrial sectors,
access real‐time pricing information and identification of periods of peak electricity rates, they
may consider purchasing on‐site distributed energy storage systems, particularly batteries, to
avoid peak charges. In addition, residential and business consumers can use the smart grid to
increase their purchases of thermal energy storage technologies for HVAC systems. Utilities
can also coordinate and aggregate these systems on the utility side of the meter, resulting in
expanded markets for thermal energy storage technologies.
Proliferation of Microgrids
The rapidly growing market of microgrids for industrial/commercial customers requires
extremely high reliability and availability of electricity supply. A microgrid is a local area
network of interconnected electricity generation, transmission and distribution systems.
Electric energy storage is a mandatory element in the design and architecture of a microgrid
that does not rely on fossil fueled electricity generation. Energy storage is central to achieving
the higher reliability and availability of energy supply, as well as maintaining the power quality
on distribution circuits.
Increased Residential Needs for On-Site Solar, Electric Vehicles, and Home Area Networks
The proliferation of residential PV systems and the potential large‐scale deployment of electric
vehicles (EVs), as well as an increase in zero net energy buildings (meaning buildings requiring
no external energy supply to meet their energy needs), have the capacity to fundamentally
change the characteristics of the 2020 market for energy storage. Customers can potentially use
behind‐the‐meter energy storage systems, particularly batteries, to capture and balance
electricity from on‐site renewable sources. They may also be able to use their battery systems to
participate in power markets by providing grid services when not in use. Meanwhile, large‐
scale adoption of electric vehicles will likely drive down battery costs for lithium ion batteries as
manufacturers increase production to meet demand and will provide opportunities for energy
storage to lessen the stress on the distribution system from increased delivery of electric power.
Vehicle owners in 2020 or sooner may be able to sell or otherwise recycle used batteries for
second‐use applications, such as providing distributed storage opportunities for the grid. They
may also be able use their batteries to supply power to the grid often termed as vehicle‐to‐grid
applications to offer ancillary services to utilities.
Opening of Electricity Markets to Competition From Energy Storage and Incentive Programs
The opening of various electricity markets to energy storage competition through the Federal
Energy Regulatory Commission (FERC) and the California Independent System Operator
(California ISO) rulings will likely spur greater deployment of technologies that can compete
better in a less restrictive market. In addition, the California ISO could implement imbalance
charges for intermittent resources, like wind and solar, by modifying rules governing the use of
these resources potentially leading renewable energy suppliers to deploy energy storage
technologies to reduce variability. Meanwhile, regulatory entities like the FERC and the
California Public Utilities Commission can ensure greater certainty of cost recovery for
investments in energy storage technologies by developing a valuation method to help monetize
7
the benefits provided by energy storage. Complementing these regulatory initiatives, the state
and federal governments could promote incentive programs, either by developing new
programs or by expanding existing ones, to help finance energy storage projects, bring down
costs and help manufacturers provide better data to spur further investment by 2020. California
could also promote energy storage deployment in 2020 through various policy mechanisms,
most prominently the state’s passage of AB 2514.
Natural and Human-Caused Disasters
Disasters over the next decade may spur greater investment in energy storage. Possible
scenarios range from costly blackouts that jeopardize business functions and perishable goods
storage to nuclear meltdowns, earthquakes, terrorist attacks, and other catastrophes. One
response to these disasters could be regulations that require on‐site backup energy storage to
provide uninterrupted and continuous electricity and to help the grid to restart as quickly as
possible.
Likely Price Decreases for Various Energy Storage Technologies
The result of the potential increase in investment and deployment of energy storage
technologies, driven by the aforementioned policy changes and technological advances that
may occur over the next decade, is that California and the nation will likely experience
significant price drops in the cost of various energy storage technologies by 2020. Meanwhile,
the likely competitor for energy storage over the next decade, natural gas, may not be able to
maintain current prices and may experience price increases by 2020, leading to greater market
opportunities for energy storage. In addition, the ability of energy storage technologies to
harness and dispatch carbon‐free renewable energy will become more important as the state
implements carbon pricing and other legal and regulatory mechanisms aimed at reaching its
ambitious greenhouse gas reduction goals. By 2020, the need for energy storage will be
significantly greater than today.
Future Research Needs to Implement the Vision
This report cites a number of technical challenges that need further research and policy
adjustments that the state, primarily through the California Energy Commission, should
undertake to achieve the vision of energy storage in 2020. These needs include:
• A valuation method for energy storage that accounts for its costs and benefits to
ratepayers and the public, given the state’s energy and environmental goals.
• Research into communication technologies that enable participation and lower cost of
participation by nonutility and end‐user owners with grid connected energy storage
technologies.
• Studies on how energy storage could operate in conjunction with demand response
technologies.
• Studies to review the effects of recent tariff changes at the New York Independent
System Operator, ISO New England, PJM Interconnection, and other grid operators on
energy storage technologies attempting to compete in the area and frequency regulation
market.
8
• Studies to review the technical requirements and cost implications of intrahour and
shorter scheduling at California ISO in connection with regional balancing.
• Research into the potential effects of using energy storage for needed regulation over
existing fossil fuel‐based power plants.
Valuation Method
To determine the appropriate amount and role of energy storage by 2020, California should first
determine a method to valuate or monetize the benefits provided by energy storage
technologies, located on either side of the utility meter. The state lacks such a valuation
framework. The high costs of current energy storage technologies and investment risk will
persist without adoption of a valuation framework that monetizes the independent benefits and
creates opportunities for cost recovery.
The CPUC is proceeding with a two‐phase rulemaking process under AB 2514, with a second
phase addressing energy storage application values, costs, and means to determine cost‐
effectiveness. A critical feature of this challenge is how to allocate the costs and benefits of
storage across the range of services that are affected, including generation, transmission,
distribution, and regulation. Some benefits may not be aggregated due to institutional barriers
or technical/operational conflicts. Policy makers will need to devise a framework that addresses
the multiple values and potential overlapping nature of energy storage’s benefits.
Considerations that the CPUC will need to address include:
• Grouping operational uses and associated benefits, such as by application(s) and
location. This method should use the most promising applications, such as the three
identified in this report: frequency regulation, integrating variable renewable energy,
and developing community or distributed energy storage systems.
• Prioritizing certain applications to meet the objectives set forth by AB 2514.
• Compensating energy storage owners/ operators for services not covered by California
ISO markets.
• Determining which services/applications and related value streams may be aggregated
to maximize financial return to a storage system without double‐counting benefits or
committing the same resource to incompatible uses at one time.
Policy makers should consider broad categories of benefits that could be monetized in the
valuation method. For example, societal benefits include reduced reliance on fossil fuel and
increased energy security, reduced criteria air pollutant and greenhouse gas emissions, and
achieving superior operation of the existing generation fleet. In addition, energy storage can
provide transmission and distribution deferral and other avoided costs, such as for variable
distributed generation integration. Additional value for customers could come from avoided
transmission and distribution fees, such as when large distributed energy storage modules
provide on‐peak energy closer to load, relieving some of the on‐peak energy flow in congested
transmission lines.
9
Key Energy Storage Policies
The state can boost appropriate deployment of energy storage by setting targets for
procurement under AB 2514, ideally in a two‐phase process with short‐term and long‐term
targets. Setting procurement targets would ensure that conventional energy storage
technologies do not have an unfair advantage over newer or less proven options that may
nonetheless become more cost effective over time. Should the state determine that targets are
not appropriate, numerous policy actions at the state and federal level still can unleash
incentives for energy storage procurement, such as opening ancillary services and capacity
markets to energy storage, expanding existing incentive programs to cover energy storage, and
focusing additional research and development dollars on energy storage demonstration projects
that prove grid applications and seek ways to reduce costs.
The CPUC can also affect the market for energy storage in the following key ways:
• Implement energy storage procurement targets under AB 2514.
• Ensure the Self‐Generation Incentive Program covers energy storage resources,
providing potentially crucial financial incentives for energy storage investment.
• Refine its Resource Adequacy (RA) program, which ensures that California ISO has
access to sufficient resources to operate the grid safely and reliably and promotes the
siting and construction of new resources needed for future reliability. All load‐serving
entities (LSE) within the CPUC’s jurisdiction are subject to the determined RA
requirements. The CPUC could promulgate rules to allow certain categories of energy
storage procurement to count toward an LSE’s RA obligations.
• Use existing proceedings related to the development of the smart grid, permanent load
shifting, demand response, long‐term procurement process, and alternative‐fueled
vehicles to explore the cost‐effective promotion of energy storage technologies as they
relate to these proceedings.
• Implement real‐time pricing and strong price differentials between peak and off‐peak
rates to boost the market for peak‐load shaving technologies.
• Encourage the California ISO and FERC to open markets to competition from energy
storage.
• Develop a valuation methodology for energy storage that accounts for its costs and
benefits to ratepayers and the public, given the state’s energy and environmental goals.
This method could value technologies based on the most promising applications and
ensure that benefits to additional applications are appropriately counted. This report
identifies three potentially promising applications for energy storage: frequency
regulation, integrating variable renewable energy, and developing community or
distributed energy storage systems.
10
California’s policy makers will need to balance the promotion of energy storage with concerns
about containing costs and not favoring less optimal technologies. They will also need to
ensure proper agency coordination, given overlapping jurisdictions, as they implement a
strategic vision for deploying energy storage in California by 2020 to realize the benefits of
energy storage to ratepayers such as clean low‐cost renewable electricity and reliable and
sustainable energy system in California.
The California Energy Commission can help encourage the use of energy storage technologies
by using public interest energy research funds to develop, improve and demonstrate energy
storage technologies. The benefits of energy storage technologies should be viewed as a key
element in ensuring electrical grid stability in everyday operations, helping to integrate the use
of intermittent energy resources into the electrical grid, providing backup energy storage to
ensure uninterrupted electricity in the event of a catastrophic event, and helping business and
residential customers achieve zero net‐energy use goals.
11
CHAPTER 1: Introduction: The Importance of Energy
Storage and Technology Status
The Importance of Energy Storage
California’s ambitious clean energy agenda places it at the forefront of efforts nationally and
globally to increase efficiency, reduce greenhouse gas emissions, and shift to a cleaner and more
sustainable energy future. The changes taking place to implement this vision are wide ranging
and encompass legislative and regulatory action, private sector and industry initiatives, public‐
private partnerships, and inter‐agency coordination and collaboration. As the state progresses
towards its 2020 targets for substantially increased renewable energy resources and reduced
greenhouse gas emissions, the potential for energy storage to assist in the integration of
renewable resources and the maintenance of a reliable and efficient electric grid takes on great
significance.
Certain types of energy storage, such as pumped hydro, have long been part of the electricity
grid, and numerous policy and planning documents, at the state and federal level, have cited
the importance of evaluating the potential benefits of energy storage in the generation,
transmission, and distribution equation. These include the California Energy Commission’s
(Energy Commission) 2009 Integrated Energy Policy Report (IEPR) and 2011 IEPR Scoping
Document, the joint energy agencies California’s Clean Energy Future – Collaborative Vision for
2020, and President Obama’s Policy Framework for the 21st Century Grid released by the
Executive Office. The United States Department of Energy has awarded hundreds of millions of
stimulus dollars to support private and public sector research in energy storage and smart grid
technologies. And the Federal Energy Regulatory Commission (FERC) has undertaken action to
enable greater participation by energy storage and other non‐generation assets.
In September, 2010, the California legislature enacted Assembly Bill 2514 (AB 2514 [Skinner,
Chapter 469, Statutes 2010]), directing the California Public Utility Commission (CPUC) to
convene a proceeding by March 1, 2012 to determine energy storage procurement targets, if
any, for investor‐owned utilities. Under the statute, similar targets would be required for
publicly‐owned utilities on a slightly later time frame. In December 2010, the CPUC issued an
Order for Rulemaking Pursuant to AB 2514, initiating the process more than a year ahead of the
statutory deadline. The adoption of targets under AB 2514 could potentially impact the electric
utilities’ procurement planning processes, which already are subject to the April 2011‐enacted
California Renewable Energy Resources Act, Senate Bill 2 (SB 2) requiring 33 percent renewable
electricity generation by 2020. The procurement planning process also must take into account
impending regulations to be promulgated by the California Air Resources Board (ARB) under
AB 32, the state’s greenhouse gas emissions reduction law.
12
This report, a project of the California Energy Commission, presents a strategic analysis of
energy storage technology for California by 2020, in part to inform the AB 2514 process. The
report assesses current energy storage technologies and the diverse policies affecting
deployment in California, outlining critical technology gaps and research needs and policy
reforms. It also provides a framework for the Energy Commission, CPUC, and other regulatory
agencies, as they create a roadmap for how commercially ready energy storage technologies can
be cost‐effectively applied in California to reduce costs to ratepayers, reduce emissions from
fossil fuel generation, and enable and accelerate the implementation of more renewable
generation and its integration in California’s electricity system.
In addition, a primary objective of this report is to determine the current status of energy
storage technologies and identify related issues, research gaps, barriers, and opportunities as
well as specific targets or milestones and specific actions necessary for development and
deployment of energy storage technologies in California. Under this direction, the authors of
this report worked with the PIER Energy Storage Program Team, industry experts, utility
representatives, and other leading practitioners to identify targets and milestones, as well as
specific actions necessary for development and deployment of energy storage technologies in
California and articulate it in the PIER Energy Storage Program Strategic Vision. The authors
also presented findings from the draft report at two CPUC hearings on energy storage and one
Energy Commission meeting, as well as in a workshop for key stakeholders. The purpose of the
targets and milestones is to clearly identify the 2020 Energy Storage goals over time that the RD
and D activities must be working toward achieving. The project advisory committee members,
listed in the acknowledgments, provided critical input to the findings of this report. All
mistakes and conclusions, however, are ultimately those of the authors.
The Energy Commission’s ultimate goal is to establish a 2020 Energy Storage Vision for the
state. As such, the agency initiated this report specifically to develop scenarios for how energy
storage maybe applied throughout California’s electric power system and to estimate the costs
and benefits of such a vision as compared to a scenario without deployment of energy storage.
The Energy Commission envisioned this report as a useful planning tool to help the CPUC and
other regulatory agencies develop a vision for how commercially ready energy storage
technologies can be cost effectively applied in California to reduce costs to ratepayers, reduce
emissions from fossil fuel generation and enable and accelerate the implementation of more
renewable generation and its integration in California’s electricity system. Given the
complexity of the regulatory and market structures that affect energy storage deployment in
California, as well as the wide range of commercially available and emerging energy storage
technologies and grid applications they may serve, this report will hopefully serve as a
13
significant step toward developing a 2020 Energy Storage Vision that the public and private
sectors alike can continue to bolster with follow‐on research.
Technology Status
Any discussion of a strategic vision for energy storage must begin with an assessment of the
diverse technologies that comprise the sector. To assess the potential for future deployment of
energy storage, policy makers and the public need an understanding of the current state of
these technologies and the likely technological advancements that may occur in the near term.
As such, this technology status chapter surveys the various energy storage technologies,
describing how they function, assessing their current state of development, and citing any
critical research and development needs to advance the technology and lower costs. In
addition, this section presents some cost data on existing technologies based on their
performance in the field.
The chapter is divided into electrochemical, mechanical, thermal and hydrogen energy storage,
under which exists a multitude of separate technologies. Each technology has its strengths,
limitations, and appropriateness for the large and diverse set of applications for energy storage.
This study does not attempt to pick the energy storage technologies that may commercially
penetrate these application markets. Although this chapter references previous analyses that
align particular technologies’ rated capacity with either storage and discharge time and that
suggest suitability to a specific application, the inclusion of these references does not indicate a
preference in this report for the particular companies or technologies involved.
Figure 1 compares the power and discharge time of various energy storage devices (note that
this graph does not reflect the latest advances in energy storage technology capabilities but is
representative nonetheless).
14
Figure 1: Comparison of Power and Discharge Time for Various Energy Storage Devices
(Yang, 2010)
Scientists classify energy storage applications by the same criteria, power and time of discharge.
They consider those applications that charge and discharge on a time scale less than an hour to
be ‘power applications,’ while they call those that charge and discharge over times longer than
one hour ‘storage applications’ (Nexight Group, 2010a, p. 7).
There are a wide range of energy storage devices currently available, and many have been in
operation for decades. Figure 2 below is from a report by Chen et al., 2009 showing the
maturity status of many major energy storage technologies1. Analysts expect advancements in
energy storage to occur with the maturation of new technologies, such as metal air batteries,
and the application of new materials and designs to proven technologies, like lead‐acid.
1 The definition of “maturity” as a metric of technology status is held in the eye of the customer.
The historically risk‐averse utility sector in a regulated environment has a different definition of
“maturity” than the unregulated customer or energy services provider. This report will express
maturity in MW‐years of commercial operations of the existing state of the art for each
particular EES technology.
15
Figure 2: Technical Maturity of Various Energy Storage Systems
(Chen et al., 2009)
1.1 Electrochemical Energy Storage
Scientists classify electrochemical energy storage (EES) devices in terms of power, often
normalized by weight (such as kilowatts per kilogram) or volume (such as kilowatts per liter) to
enable performance comparisons, and the discharge time for which the devices can sustain that
power. The energy that the devices store is the product of the power and discharge time.
All EES devices are composed of two basic parts, the electrodes and the electrolyte.
Conventional battery technologies store energy in the electrodes, while ion transport, the
essential feature of electrochemical processes which creates the flow of electricity, occurs
through the electrolyte. Flow batteries are rechargeable batteries in which electrolyte
containing one or more dissolved electro‐active species flows through an electrochemical cell
that converts chemical energy directly to electricity. Flow batteries differ from more
conventional batteries in that they store energy in the electrolytes. Advances in materials, and
especially in controlling the interfaces between the electrodes and electrolyte, will drive
improvements in EES performance. Potential opportunities also exist for hybridization of
devices. For example, engineers have made achievements in combining the high energy density
of batteries with the high power density of capacitors, which are devices that store electric
surface charges for fast reacting power discharge.
Device improvements can increase the specific power output of the device or the specific energy
capacity. However, making wide‐scale adoption of EES economically feasible will require
improvements in both energy and power density. Scientific research in EES is often concerned
16
with energy and power density normalized by units of weight or volume. Scientists choose
these units in part because in the traditional EES applications, such as portable electronics and
transportation, the weight and volume of a device can affect its usefulness. However, for
stationary EES applications, such as grid stability, load shifting and renewables integration,
weight and volume are not such important parameters. Instead, cost, cycleability (the capability
to complete numerous charge and discharge cycles), and reliability are the key parameters for
EES on the grid. For instance, if electric power companies use EES in transmission as a
substitute for upgraded power lines, the reliability of the EES system must match the reliability
of power lines, which is 99.9percent (Nexight Group, 2010a, p. 12). Thus, over the next decade,
engineers must make advances not only in the physics and chemistry of EES, but also in
improving the manufacturing processes to produce reliable and long‐lasting devices more
cheaply. They should also incorporate materials into the device designs that are
environmentally benign and abundant.
1.1.1 Lead-Acid Batteries
1.1.1.1 Description of Technology
Lead‐acid batteries are the most mature of all EES systems that exist today and offer an
immediate solution while engineers continue to develop other EES systems. Lead‐acid batteries
are particularly attractive for early deployment due to their relatively low cost, ease of
manufacture, rapid electrochemical reaction kinetics, and good cycle life under controlled
conditions (Huggins, 2010, p.237).
1.1.1.2 Technology Status
Although lead‐acid is the most commercially mature rechargeable battery technology in the
world, the technology is undergoing a period of rapid development amid renewed interest.
The maximum theoretical specific energy (the energy content per unit mass) of fully
compounded active material in a lead acid battery is 171 Watthours per kilogram (Wh/kg)
(Huggins, 2010, p. 239). However, conventional lead acid batteries lose much of this potential
due to the need for internal connectors and grids to structurally support the heavy lead
material. Conventional lead acid batteries typically achieve 20‐30 Wh/kg and possibly much
more by replacing or eliminating the non‐active lead components. Maintenance free valve‐
regulated lead‐acid (VRLA) batteries, also known as sealed lead acid batteries, have largely
replaced conventional high maintenance “flooded cell” batteries in a variety of applications,
including automotive, marine, telecommunications, and uninterruptible power supply (UPS)
systems. However, for large scale grid storage, many lead acid experts and manufacturers still
view flooded lead‐acid batteries as the best alternative.
More recently, engineers have made significant advances in the chemistries and formulations
used and even the basic architecture. Assuming 2 Volts (V) per cell, the maximum theoretical
17
specific energy of lead‐carbon batteries is 171 Wh/kg (Huggins, 2010, p. 239). However, due to
the need for water and other components, working values for specific energy vary from 30‐50
Wh/kg. The power density of lead‐acid batteries has reached 4 kiloWatts per kilogram (kW/kg)
over 10 seconds in prototype batteries (Wagner, 2002).
The lifetime of lead‐acid batteries varies significantly depending on the application, discharge
rate, and number of deep discharge cycles, which can reduce life. A 1 megawatt (MW)/1.5
Megawatt hours (MWh) lead‐acid system by GNB Industrial Power and Exide has been
operating for 12 years at a remote island location in Alaska. In that project, the battery system
exhibited very little visible degradation upon post‐test analysis, and operators replaced it in
2008. Operators have deployed other lead‐acid energy systems in sizes of 10 to 20 MW. The
cost of lead can influence the battery price. For bulk energy storage, costs for lead‐acid batteries
range from $425‐$980 /kWh, depending on efficiency, capacity, and power (EPRI, 2010a, p. 4‐
25).
1.1.1.3 Technology Limitations and Opportunities
Because grid operators could not control the cycle conditions for a renewable energy‐ intensive
grid application (particularly for variable sources such as solar and wind), traditional lead‐acid
batteries would experience short cycle life and significant maintenance requirements.
Manufacturers have made improvements to lead‐acid systems by using carbon‐enhanced
electrodes to create advanced lead‐carbon batteries. Adding carbon components to the
electrodes does not change the basic electrochemistry but acts to increase the power output and
reduce detrimental chemical reactions within the battery cells (Nexight Group, 2010a, p. 18).
Manufacturers can also obtain additional performance enhancement in lead‐acid batteries by
changing the electrolyte material.
Key issues for the deployment of lead acid batteries are:
• Need for careful thermal management, including a high likelihood for active cooling.
• Need for custom‐designed grid storage system using lead acid batteries.
1.1.2 Nickel-Based Batteries
1.1.2.1 Description of Technology
Nickel‐based batteries come in the form of Nickel/Cadmium (NiCad) and Nickel/metal hydride
(NiMH) systems. The NiCad battery uses nickel oxide hydroxide as the positive electrode and
metallic cadmium as the negative electrode. The NiMH battery has a number of technical
features similar to a NiCad battery. However, the NiMH battery uses a hydrogen‐containing
alloy for the negative electrode instead of cadmium, thus changing the electrochemistry and
resulting in two to three times the energy storage capacity. NiMH batteries are best suited for
applications such as hybrid electric vehicles, where they are already in widespread use.
18
1.1.2.2 Technology Status
Currently, the largest application for NiCad batteries is in portable electronics, similar to
primary batteries for the same devices. Compared with lead‐acid batteries, NiCad batteries
have longer cycle life, higher energy densities, and lower maintenance (Nair and Garimella,
2010). Nickel metal hydride batteries are a feasible alternative to NiCad batteries due to their
better performance and environmental benefits. In comparison to lead‐acid and NiCad
batteries, NiMH does not contain toxic substances (cadmium or lead). The energy density of
NiMH cells is 25–30percent better than NiCad cells (Crompton) but well below rechargeable Li‐
ion cells (Nexight Group, 2010a, p. 21).
1.1.2.3 NiCad Battery Technology Limitations and Opportunities
Many of the advancements in NiCad batteries have been in device architecture, using the jelly‐
roll design that is similarly akin to primary batteries of the same type. In this design,
manufacturers lay down an anode (negative electrode) material, apply a separator, and then lay
down a cathode (positive electrode) material. They then roll this assembly into a cylinder,
leading to the term “jelly roll.” Although NiCad batteries are relatively inexpensive and quite
reliable for a number of applications, especially consumer products, they face some of the same
limitations as lead‐acid batteries: the need for large volumes, toxicity, and the severity of the
self‐discharge, a phenomenon in batteries in which internal chemical reactions reduce the stored
charge in the battery without any connection to the electrodes. Therefore, while NiCad
batteries are a potentially competitive replacement for lead‐acid batteries in the near term, they
are not a suitable long‐term solution.
1.1.2.4 NiMH Battery Technology Limitations And Opportunities
NiMH batteries, like NiCad batteries, also suffer from severe self‐discharge, making them
inefficient for long term energy storage. As a result, NiMH batteries represent another
possibility for interim applications but not for long term solutions.
1.1.3 Lithium-Ion Batteries
1.1.3.1 Description of Technology
Lithium (Li) is an attractive material for batteries because it has a high reduction potential,
which is a tendency to acquire electrons, (‐3.04 Volt (V) versus a standard hydrogen electrode)
and is lightweight. However, Li is also highly reactive with water and oxygen, so Li batteries
must use non‐aqueous electrolytes such as lithium hexafluorophosphate (LiPF6), lithium
tetrafluoroborate (LiBF4), and lithium perchlorate (LiClO4), which are dissolved in an organic
solvent instead of water. Because these electrolytes, like Li, react explosively with water, Li
batteries must remain tightly sealed from the environment during manufacture, use, and
disposal.
19
In a Li‐ion battery, lithium ions move from the negative electrode to the positive electrode
during discharge, and then back to the negative electrode when charging. In a conventional LI‐
ion cell, manufacturers produce the negative electrode from carbon (most commonly graphite),
while the positive electrode is a transition metal oxide which undergoes oxidation‐reduction or
redox reactions , which are the fundamental reactions that transfer electrons between species in
electrochemical cells (the standard being lithium cobalt oxide). The electrolyte that separates
the two electrodes is a lithium salt dissolved in an organic solvent.
1.1.3.2 Technology Status
Rechargeable Li‐ion batteries are commonly found in consumer electronic products, which
make up most of the worldwide production volume of 10 to 12 Gigawatt hours (GWh) per year.
Already commercial and mature for consumer electronic applications, plug‐in hybrid electric
vehicle (PHEV) and ‐electric vehicle (EV) manufacturers are positioning Li‐ion to be the leading
technology platform for their products, which will use larger‐format cells and packs with
capacities of 15 to 20 kWh for PHEVs and up to 50 kWh for all‐electric vehicles (EPRI, 2010a, p.
4‐17).
Compared to the long history of lead‐acid batteries, Li‐ion technology is relatively new. Many
different Li‐ion chemistries exist, each with specific power versus energy characteristics.
Analysts expect the EV and energy storage markets to benefit from the high‐volume production
and safety experience that large‐scale lithium‐ion battery manufacturers have gained through
producing batteries for consumer electronics. Large‐format batteries are currently the subject of
intense R and D, scale‐up, and durability evaluation for near‐term use in hybrid electric
vehicles. However, these cells are only available in limited quantities as auto equipment
manufacturers increase production of PHEVs and EVs. However, Li‐ion battery original
equipment manufacturers (OEMs) are increasing their global manufacturing capabilities to
meet the future needs of the automotive market. Analysts anticipate the manufacturing scale of
Li‐ion batteries (estimated to total approximately 35 GWh by 2015) to result in an over‐capacity
of supply and lower‐cost battery packs, which could be integrated into systems for grid‐support
applications of energy storage duration (EPRI, 2010a, p. 4‐17).
The high energy density and relatively low weight of Li‐ion systems make them an attractive
choice for vehicles and other applications where space and weight are important. Given their
attractive cycle life and compactness, in addition to high roundtrip energy efficiency that
exceeds 85 percent–90 percent, Li‐ion battery manufacturers are demonstrating their products
for several utility grid‐support applications, including distributed energy storage systems
(DESS) at the community scale, transportable systems for grid‐support, commercial end‐user
energy management, home back‐up energy management systems, frequency regulation, and
wind and photovoltaic smoothing (EPRI, 2010a, p. 4‐18).
20
Although their operating voltage, specific energy, lifetimes, and safety varies, lithium ion
batteries are relatively light, compact, and work with a voltage on the order of 4 V with a
specific energy ranging between 100 Wh/kg and 150 Wh/kg (Scrosati and Garche, 2010), with a
theoretical specific energy of 450 Wh/kg (Beck and Rüetschi, 2000). A recent demonstration
involving the 8 MW/32 MWh Southern California Edison Tehachapi Project has a total project
cost of $57.2 million (with the Energy Commission’s PIER program providing over $6 million in
funding). The project costs include all expenses, such as the building, the interconnection, and
operation and maintenance (O and M). In this project, the battery energy cost is then
~$1,788/kWh. Cost projections for utility‐support applications or commercial applications,
where these batteries are currently better suited, range from $900‐$1,900/kWh (EPRI, 2010a, p. 4‐
24).
1.1.3.3 Technology Limitations and Opportunities
Recent improvements in Lithium‐ion batteries technology have been primarily in the design of
novel, three‐dimensional architectures and the use of nanoscale materials. These approaches
have been employed to improve power. To improve the specific energy, engineers have been
actively researching and developing the replacement of graphite and of lithium cobalt oxide
with higher capacity and lower cost electrode materials. Engineers are currently exploring
Lithium‐metal alloys as a replacement for the carbon electrode. For the positive electrode,
manganese‐based compounds and lithium metal phosphates have shown promise. All of these
initiatives are at the development stage.
Engineers are also examining the replacement of the electrolyte solutions with safer and more
reliable alternatives. Concerns with the current standard electrolyte, a LiFP6‐organic carbonate
solution, include:
• the narrow voltage stability, preventing the use of higher voltage cathodes
• the high vapor pressure and flammability
• the toxicity of LiFP6
Efforts to solve these issues include (i) additives to build‐up a stable solid‐electrolyte interface
and/or enhance its thermal stability; (ii) redox additives to protect from overcharge; (iii) shut‐
down separators to prevent thermal runaway; and (iv) lithium salts as an alternative to LiPF6,
to reduce toxicity (Scrosati and Garche, 2010). Engineers may also be able to develop a polymer
electrolyte, allowing for a solvent‐free, Li‐ion conducting membrane. Overall, Li‐ion batteries
are a promising mid‐term solution and a potential longer‐term solution.
Another approach to reducing costs of Li‐ion batteries would be through a “second use.” Once
Li‐ion batteries for plug‐in hybrid and electric vehicles (PHEV/EV) degrade to 70percent‐
21
80percent of their original power/capacity, they are no longer sufficient for automotive use.
However, electric power companies could reuse these aged batteries in other applications. Such
second‐use applications could increase the total lifetime value of the battery and thus reduce its
cost to the automotive user. Given that the high cost of the batteries significantly reduces the
affordability of PHEV and EVs, sale of the batteries for reuse could boost the PHEV and EV
market by driving down overall costs. Possible additional secondary uses for Li‐ion batteries
include residential and commercial electric power management, power grid stabilization to
help provide reliable electricity to users, and renewable energy system firming. The latter
involves using batteries to make the power provided by variable resources such as wind and
solar energy more useable for the grid. The DOE National Renewable Energy Laboratory
(NREL) will partner with an industry‐academia team led by the California Center for
Sustainable Energy (CCSE) to investigate secondary applications for aged battery packs from
electric and hybrid electric vehicles (Neubauer and Pesaran, 2010).
1.1.4 Flow Batteries
1.1.4.1 Description of Technology
A flow battery is a form of rechargeable battery in which electrolyte containing one or more
dissolved electro‐active species flows through an electrochemical cell that converts chemical
energy directly to electricity. When the devices introduce these materials into the cell during
operation, they produce an oxidation‐reduction (redox) reaction. A membrane or other
separator isolates the electrolytes to allow for ion transport. The devices store additional
electrolyte externally, generally in tanks, and usually pump it through the cell (or cells) of the
reactor. Flow batteries can rapidly ʺrechargeʺ by replacing the electrolyte liquid (in a similar
way to refilling fuel tanks for internal combustion engines) while simultaneously recovering the
spent material for re‐energizing. This differs from more conventional storage batteries, which
store the electrolytes within the battery cell itself.
1.1.4.2 Technology Status
A number of potential redox flow battery chemistries exist and are highlighted in Table 1.
Vanadium redox battery technology is one of the most mature types of flow battery systems
available. Flow batteries store energy as charged ions in two separate tanks of electrolytes, one
of which stores electrolyte for positive electrode reaction while the other stores electrolyte for
negative electrode reaction. Vanadium redox systems use one common electrolyte, which
provides potential opportunities for increased cycle life. When power companies need
electricity, the electrolyte flows to a redox cell with electrodes, generating current. Operators
can reverse the electrochemical reaction by applying an over‐potential, as with conventional
batteries. This application allows the system to repeatedly discharge and recharge. Like other
flow batteries, many variations of power capacity and energy storage are possible depending on
the size of the electrolyte tanks (EPRI, 2010a).
22
Manufacturers can design vanadium redox systems to provide energy for two hours to more
than eight hours, depending on the application. Manufacturers claim that cycling does not
strongly affect the lifespan of flow‐type batteries because the devices store the electrolytes
externally. Suppliers of vanadium redox systems estimate the lifespan of the cell stacks to be 15
or more years, while the balance of plant and electrolyte can have lifetimes of over 25 years.
System suppliers also say they have achieved cycling capability of 10,000 or more cycles at
100percent depth of discharge. The physical scale of vanadium redox systems tends to be large
due to the high volumes of electrolyte required when sized for utility‐scale (megawatt‐hour)
projects. Vanadium redox battery costs are in the range of $620‐$740 /kWh for bulk energy
storage applications (EPRI, 2010a, p. ES‐18).
For bulk energy storage applications, Zn/Br batteries are currently in the demo stage of
development. For example, the Energy Commission’s PIER program helped fund a Primus
Power Corporation 25 MW/75 MWh grid‐connected Zinc‐based flow battery energy storage
system to provide renewable firming, strategic local peak shaving, automated load shifting, and
ancillary services. Recent studies project the cost of these projects at $290‐$350/kWh (EPRI,
2010a, p. ES‐18). Analysts project Fe/Cr batteries, currently emerging into commercial
demonstrations, to cost $360‐$380/kWh (EPRI, 2010a, p. ES‐18). As in vanadium redox systems,
the Zn/Br battery charges and discharges in a reversible process as the device pumps
electrolytes through a reactor vessel. While the technology lacks significant field experience,
vendors claim estimated lifetimes of 20 years, long cycle lives, and operational ac‐to‐ac
efficiencies of approximately 65percent to 70percent (EPRI, 2010a, p. 4‐25).
23
Table 1: Comparison of Various Flow Batteries
(Nguyen and Savinelli, 2010)
1.1.4.3 Technology Limitations and Opportunities
Flow batteries are an attractive energy storage option for the grid because of their ability to
store large amounts of energy and their potential long cycle life and high efficiency. However,
the technology is still young with significant cost barriers. For flow batteries that employ
different active species in the anode and cathode electrolytes in particular, cross‐transport
through the cell can lead to efficiency loss and contamination. Flow batteries can also suffer
from thermal and chemical stability problems in the membranes and electrolytes, which can
impact operating range and battery life. Additionally, the stack design of flow batteries may
present a scaling problem, particularly the generation of electrical shorts. Leaks, pump
reliability, and seal and tubing life are also important issues for flow battery technologies.
Finally, the delivery system for the electroactive materials needs improvements in robustness
and cost. To address these issues, engineers must make improvements to the membrane, stack
design, monitoring systems, novel materials and chemistry, and cell design.
24
The relatively short field experience of flow batteries suggest that the claims of 20 ‐ 30 year
lifetimes are still to be proven.
1.1.5 Sodium-eta Alumina Batteries
1.1.5.1 Description of Technology
Sodium (Na) is a good battery material because it has a high reduction potential, meaning a
tendency of a chemical species to acquire electrons (‐2.71V versus a standard hydrogen
electrode), is lightweight, non‐toxic, inexpensive, and relatively abundant. However, sodium is
also highly reactive with water and oxygen. As a result, engineers must tightly seal sodium
batteries from the environment. In contrast to other battery systems, sodium beta‐ alumina
batteries consist of liquid electrodes and a solid electrolyte separator. These batteries operate at
elevated temperatures (300 to 350 °C) to ensure that the electrolyte (sodium beta‐alumina) and
the electrodes have high sodium ion conductivity to enable the flow of electrons and that the
electrodes have low viscosity to facilitate the flow of reactants enabling sodium batteries to
work. Sodium beta‐alumina batteries discharge by transporting sodium ions from a liquid
sodium anode, through a beta‐alumina solid electrolyte (BASE), and into a cathode
compartment. The cathode is either sodium polysulfide (Na‐S battery) or a porous metal/metal
halide structure impregnated with molten NaAlCl4 as a second electrolyte. The latter, Na‐metal
halide batteries, have a higher cell voltage compared to Na‐S (2.58 V at 300 oC versus. 2.21 V at
350 oC). Commercially available sodium‐beta batteries have energy densities around 100 Wh/kg.
Figure 3 shows the charging and discharging processes for a Na‐S battery; the round trip
efficiency between charging and discharging the battery is up to 90percent. BASE is an
excellent sodium ion conductor at elevated temperatures, and its low electronic conductivity
prevents electrical shorting. Because sodium‐beta batteries must operate at high temperatures,
300‐350 oC, thermal management is an integral part of any sodium‐beta battery system.
25
Figure 3: Principles of Sodium-Sulfur Battery Operation
(NGK, 2011)
1.1.5.2 Technology Status
NGK Insulators has commercialized a sodium‐beta battery that is at the forefront of grid‐scale
energy storage. Since 2002, the Japanese company has been manufacturing Na‐S battery
modules capable of operating at 50 kW for 6 hours. Operators collect these modules into larger
MW scale units that vendors market mainly to utilities. As of April 2009, NGK had a total of
300 MW distributed at 200 locations worldwide (NGK, 2011). The main applications for these
units are load leveling, power quality maintenance, and renewable energy management. The
largest application of NGK’s sodium‐beta batteries is a 34 MW installation coupled to a 51 MW
wind farm built by the Japanese Wind Development Company. NGK also has a partnership
with Younicos, a German company that has established a grid‐scale energy experiment on the
Azorean Island of Graciosa. The goal of this experiment is to power the island’s isolated
electricity grid with 70‐90percent wind and solar energy using 10‐30 MW backup generators
and a 1 MW NGK Na‐S battery system (Younicos, 2009).
NGK is facing growing competition from General Electric (GE), which in 2007 acquired Beta R
and D and in 2010 began marketing sodium‐beta batteries based on Na‐metal halide technology
under the trade name Durathon. Durathon designed the modules to be smaller than NGK’s
and to have the capability for 8 kW discharge for 2 hours (GE). GE is targeting slightly different
markets, including backup power supply for hospitals, data centers and wireless
communication towers. In these areas, they compete more directly with lead‐acid technology.
Nonetheless, GE’s modules are capable of grid‐scale storage, and GE will likely make a stronger
push into that market as demand grows. The competition between NGK and GE technologies,
with their varying chemistries and module designs, should help increase performance and
drive down costs in the near term.
26
1.1.5.3 Technology Limitations and Opportunities
One of the principal limitations of sodium‐beta batteries is their high temperature of operation.
As indicated above, both electrodes must be in the liquid state to be sufficiently fluid, to
conduct sodium ions effectively, and to ensure that the electrolyte is very conductive, thus
minimizing ohmic losses. Freeze/thaw cycles dramatically reduce the working life of sodium
batteries, so operators require thermal management systems which, however, can increase the
size of sodium‐beta batteries and potentially decrease their overall efficiency. An improved
sodium‐beta battery would have lower melting point electrodes and a solid electrolyte that is
sufficiently conductive at low temperatures, which would reduce thermal losses, improve
efficiency, reduce material costs, and improve durability. Sodium‐beta batteries also suffer
from decreasing liquid volume in the sodium anode during discharge, resulting in reduced
contact between the BASE and the liquid sodium, which can reduce efficiency and output.
Manufacturers can improve this interaction by using a variety of methods, including gravity,
wicking, or gas pressure. Wicking is the most practical method, and manufacturers
predominately use this method in commercial sodium‐beta batteries (Yang, 2010). Surface
treatment and surface science studies of the BASE could enhance understanding of the
electrode‐electrolyte interface (Nexight Group, 2010a, p. 25).
The molten polysulfide cathode in the Na‐S battery is extremely corrosive and tends to degrade
the BASE. If the BASE fails, the liquid cathode and anode will penetrate the cracks and, upon
contact, react violently. The semi‐solid cathode in the Na‐metal halide battery has a much safer
mode of cell failure that results in low resistance so that other Na‐metal halide cells in series can
continue to function. Developers of Na‐metal halide batteries also have the potential to
improve the battery performance by using alternative transition metal chlorides, which can
lower operating temperatures, improve safety and increase reliability. Future research should
address corrosive interactions between the electrodes and electrolytes. Less corrosive electrodes
would justify thinner electrolytes, which would help lower the manufacturing costs of sodium‐
beta batteries.
Manufacturers could lower costs in the near term by improving BASE manufacturing
techniques, which include materials preparation by sol‐gel chemistry, co‐precipitation, or spray‐
freeze/freeze drying. The purity of the sodium beta‐alumina is important because the presence
of different alumina phases or impurities like calcium or silicon can reduce the membrane’s
conductivity significantly. Currently, sodium beta‐alumina membranes are composed of tubes
with a wall thickness of 1‐2 mm, which is necessary for structural stability (Nexight Group,
2010a, p. 25). Researchers have shown that smaller and more mono‐disperse grain sizes lead to
better structural performance, which could allow engineers to build sodium‐beta batteries with
thinner and less expensive membranes (Yang, 2010).
27
1.1.6 Metal Air Batteries
1.1.6.1 Description of Technology
Metal‐air batteries get their name from reactants that provide energy by undergoing
electrochemical reactions. In zinc‐air cells, for example, energy is released by oxidizing zinc
(which is held within the battery case) with oxygen from the air. Insofar as the oxidizer is not
stored in the battery but supplied continuously from an external source—air—these kinds of
batteries are similar to fuel cells, in which neither the oxidizer nor the fuel is packaged inside
the cell. Metal‐air batteries reduce atmospheric oxygen to form oxides and/or peroxides with
various metals. Metal oxidation reactions have high energy density, and because the batteries
use atmospheric oxygen as the active cathode material, the weight of the cathode reactant is not
included in the weight of a metal‐air battery. These two factors combine to give metal‐air
batteries high theoretical energy densities such as 1,300, 8,100 and 11,100 Wh/kg for zinc,
aluminum and lithium, respectively (Armand, 2008; Douche, 1997). Because current Li‐ion
batteries generate 120 Wh/kg from a theoretical energy density of just 450 Wh/kg, investors
have shown significant interest in metal‐air batteries.
Engineers use primary Zn‐air batteries to power everyday items like watches and hearing aids,
and some scientists have proposed using refillable Zn‐air fuel cells for automotive applications.
However, grid storage applications require flexibility in charging and discharging, which
necessitates the use of secondary (rechargeable) batteries. Building secondary metal‐air
batteries entails a number of technical challenges, and each of the metals proposed for metal‐air
applications has its own unique properties. For example, Li reacts violently with water while Al
develops a thin oxide layer when exposed to air that could interfere with battery operation. The
general structure of a lithium‐air battery is shown in Figure 4.
28
Figure 4: Structure of Li-air Battery
(Armand, 2008)
In this cell design, engineers separate the solid Li metal from a porous carbon composite air
electrode with a polymer gel electrolyte. Li ions migrate to the air electrode and react with O2 to
form Li2O2. This reaction has a specific capacity of 1,200 mAh/g and has the unique advantage
of reacting without O2 cleavage to atomic oxygen (O atoms), which has extremely high
activation energy and requires expensive catalysts like platinum to reduce activation energy of
chemical reactions (Armand, 2008).
1.1.6.2 Technology Status
Scientists developed commercial Zn‐air batteries as early as 1932 for applications like remote
communication and ocean buoys. In the 1970s, fuel cell research into air electrodes enabled
development of the now ubiquitous Zn‐air button cell. Zn‐air fuel cells with mechanical
recharging have been used to power buses and for stationary power generation. The general
structures of these fuel cells are a thin air cathode composed of a nonprecious metal catalyst
over a carbon current collector, polymer membrane separator, alkali electrolyte and powdered
zinc anode. The oxygen reduction catalysts used in the cathodes of these fuel cells are typically
composed of manganese oxides made by using sol‐gel chemistry (Neburchilov et al, 2010).
However, catalysts for secondary metal‐air batteries must be able to both reduce and evolve
oxygen. Currently, Arizona‐based Fluidic Energy is developing zinc‐air batteries with an ionic
liquid electrolyte for utility‐scale applications. The company’s choice of catalyst remains
proprietary (Yang).
Manufacturers have been unable to develop other metal‐air batteries as efficiently as Zn‐air. Al‐
air batteries have a high energy density, but their prohibitive cost, limited shelf life, and
29
unreliability have thus far prevented them from making much headway in the commercial
market (Woodbank, 2011). Currently, the military has been the primary consumer of Al‐air
batteries for various applications. In 1996, Abraham developed the first lab‐scale Li‐air battery,
and manufacturers built Li‐air battery with demonstrable rechargeability ten years later
(Abraham, 2006; Armand, 2008). To date, manufacturers have not developed Li‐air batteries for
the commercial market, and rechargeable Li‐air batteries are unlikely to be available in the near
term.
1.1.6.3 Technology Limitations and Opportunities
There are numerous challenges preventing the effective implementation of rechargeable metal‐
air batteries. While using atmospheric oxygen is advantageous from an energy density
perspective, it presents many practical difficulties in building devices. To support higher
discharge rates, manufacturers need to increase the transport of oxygen to the cathode and
develop new catalysts with low oxygen overpotentials (Nexight Group, 2010a, p. 38). Many
possible materials deserve further investigation, such as peroversuskite, pyrochlore, and spinel‐
type metal oxides (Fujiwara). Another common research problem for all air‐metal battery
systems is how to reversibly plate the metal during charging. When metals do not plate evenly,
they form dendrites that, if they grow long enough, could lead to short circuits and result in
exothermic reactions. A classic strategy for smoothing deposited metal surfaces in
electroplating is to use chelating ligands, or connecting molecules. However, researchers have
not yet applied this method to Li‐air batteries (Armand, 2008). In the case of Li‐air batteries,
catalysts are often composed of a mixture of metal oxides, conductive binders, and porous
carbon materials such as aerogels or nanotubes. Li2O2 has a tendency to clog the air‐catalyst
pores. Researchers will need a better understanding of reaction kinetics, as well as finer control
over the morphology of the air‐electrode, to prevent this clogging. In addition, because Li is
highly reactive, manufacturers must take care to prevent the diffusion of water and CO2
through the air‐electrode and electrolyte and into the bulk Li. The use of ionic liquids and
anion exchange membranes represents a promising approach for accomplishing this objective.
1.1.7 Electric Double Layer Capacitors
1.1.7.1 Description of Technology
An electric double‐layer capacitor (EDLC), also known as supercapacitor, supercondenser,
pseudocapacitor, electrochemical double layer capacitor, or ultracapacitor, is an electrochemical
capacitor with relatively high energy density. Compared to conventional electrolytic capacitors,
the energy density is typically on the order of hundreds of times greater. Conventional
capacitors store energy by the removal of charge carriers, typically electrons, from one metal
plate and depositing them on another. This charge separation creates a potential between the
two plates, which an external circuit can harness. The total energy stored in this fashion
increases with both the amount of charge stored and the potential between the plates. The
30
amount of charge stored per unit voltage is essentially a function of the size, the distance, and
the material properties of the plates and the material in between the plates (the dielectric), while
the potential between the plates is limited by breakdown of the dielectric. The dielectric
controls the capacitorʹs voltage. Optimizing the material leads to higher energy density for a
given size of capacitor.
EDLCs do not have a conventional dielectric. Rather than two separate plates separated by an
intervening substance, these capacitors use ʺplatesʺ that are in fact two layers of the same
substrate, and their electrical properties, the so‐called ʺelectrical double layerʺ, result in the
effective separation of charge despite the vanishingly thin (on the order of nanometers) physical
separation of the layers. The lack of need for a bulky layer of dielectric permits the packing of
plates with much larger surface area into a given size, resulting in high capacitances in
practical‐sized packages. EDLCs are applicable for power applications that require very fast
response times, high cycleability and less energy density than batteries. The amount of energy
stored by a capacitor is proportional to the active surface area of the electrode material and to
the square of the voltage under which the charge is stored. Manufacturers make traditional
EDLCs by placing an aqueous electrolyte like potassium hydroxide (KOH) or sulfuric acid
(H2SO4) between two carbon‐based electrodes, usually activated carbon. The aqueous electrode
has a limited voltage window because water will break down via hydrolysis at just over 1 V.
Also, activated carbon is not optimized for charge storage and has regions of nanodimensional
pores which are inaccessible to the electrolyte and thus electrochemically inactive. EDLCs
possess an enormous advantage over other technologies due to their extended lifetime of up to
a million cycles – a benchmark unmatched by other forms of EES.
1.1.7.2 Technology Status
NEC first commercialized the aqueous electrolyte supercapacitor in 1971 and used the
technology as backup power for computer memory. In 2011, commercial EDLCs use organic
electrolytes like acetonitrile that have lower conductivity but have a maximum useful voltage,
or breakdown voltages, as high as 2.7 V. These features dramatically increase the energy
density of the devices because of the V2 dependence of energy density on voltage. As of 2008,
EDLCs were available with energy densities around 5 Wh/kg and peak power densities up to 10
kW/kg (Simon and Gogotsi, 2008). As the performance of EDLCs improves, consumers are
using them for an increasingly diverse array of applications. Customers can harness EDLCs as
standalone energy storage for short‐term back up power supply, regenerative braking in electric
and hybrid electric vehicles, and load leveling. They can also couple the devices to other energy
sources, including batteries and fuel cells, to increase the power density of those devices. As an
energy storage device, however, EDLC’s are relatively expensive, with costs in the $10,000/kWh
range.
31
1.1.7.3 Technology Limitations and Opportunities
Control of electrode pore size is a major area of research for EDLCs. Research suggests that the
optimum pore size is just slightly smaller than the mobile ion size. In this case, the pore walls
strain the solvation shell that surrounds the mobile ion, increasing the energy stored by the ion
(Largeot et al, 2008). Researchers must also consider the distribution of pore sizes in the carbon
electrodes, because the balance between large and small pores determines the balance between
specific capacitance and resistivity. While small pores store charge efficiently, larger pores are
necessary for ions to migrate quickly toward and away from the electrode surface. For EDLCs
to maintain sub‐second response times with improved energy density, manufacturers must
make electrodes with a carefully controlled structure that provides tiny pores for energy density
connected by a network of large channels for fast ion conduction.
Improved EDLC designs will take advantage of new nanostructured materials such as
nanotubes, aerogels, xerogels, and carbide derived carbons. In addition, new EDLCs feature
exotic organic or ionic liquid electrolytes with higher breakdown voltages. Engineers can
immobilize liquid electrolytes in solid nanostructures and use them as thin film electrolytes that
dramatically increase the power and energy density of the device. Because of the relationship
between ion size and optimum pore size, manufacturers must develop electrode materials in
tandem with new electrolytes so that new EDLCs can maximize energy density through both
increased active surface area and higher voltage windows.
1.1.8 Pseudocapacitors
1.1.8.1 Description of Technology
Pseudocapacitors represent a class of capacitive storage materials that use fast and reversible
surface, or near surface, redox reactions to achieve charge storage. The redox reactions result
from the adsorption of small ions, typically Li+ or H+, onto the surface of the electrodes.
Manufacturers use two classes of materials, conducting polymers and transition metal oxides
(TMOs), in pseudocapacitor electrodes.
1.1.8.2 Technology Status
Pseudocapacitors have yet to match the low cost and reliability of EDLCs. Therefore, EDLC
charge storage remains the lone commercial technology for capacitor applications.
1.1.8.3 Technology Limitations and Opportunities
The electric charge transfer, or Faradaic, reactions in conducting polymers occur by reversibly
doping and undoping the polymer. Doping intentionally introduces impurities into an
extremely pure semiconductor for the purpose of modulating its electrical properties. The first
conducting polymers only had the potential for p‐doping, but developers eventually introduced
n‐doping polymers. Researchers have only shown polythiophenes to be n‐dopable at voltages
that are compatible with electrolyte and polymer stability. Recently, manufacturers showed that
32
n‐doped and p‐doped polymers could be paired in a device to expand the operating voltage
window. Manufacturers have opportunities to further improve this design by designing new
conducting polymers that can undergo additional charge transfer reactions (Huggins, 2010, p.
70).
TMOs exhibit very large specific capacitance with only modest surface area. Because of charge
transfer reactions, the area‐normalized specific capacitance for TMOs is over 100 microfarads
per square centimeter (μF/cm2). In contrast, EDLCs are on the order of 10 μF/cm2. By far, the
most significant TMO is ruthenium oxide (RuO2), which has a specific capacitance of 380 Farads
per gram (F/g) when anhydrous, or not hydrated with water. Interestingly, the capacitance
doubles to 768 F/g for the hydrated material. RuO2 has three distinct oxidation states within a
1.2 V window, which is an important contributor to its large capacitance (Simon and Gogotsi,
2008). While RuO2 is a very potent pseudocapacitive material, the technology is also expensive.
For this reason, manufacturers are attempting to reduce the RuO2 content in electrode materials
while maintaining high specific capacitance. One strategy is to make binary Ru‐metal oxides
with a second, cheaper metal oxide. Promising prototypes with this design have been made
with 5 Wh/kg and 750 W/kg (Huggins, p. 77). Another strategy for reducing Ru‐content is to
embed RuO2 particles into a traditional EDLC carbon electrode. Coupling RuO2 with carbon
can improve the capacitance of the electrode at higher scan rates when the RuO2 faradic reaction
does not have time to occur. Engineers must develop pseudocapacitive oxide materials that do
not contain ruthenium. However, to identify and evaluate new materials, researchers must
better understand the kinetics of pseudocapacitance. Potential substitutes for RuO2 which have
attracted investor attention include manganese dioxide (MnO2), iron oxide (Fe3O4), niobium
pentoxide (Nb2O5), and lithium titanate spinel oxide (Li4Ti5O12). If manufacturers are to
successfully develop pseudo capacitive devices, they must not only achieve higher energy
density but also maintain the power density and cycleability that have enabled EDLC‐based
technology.
1.1.9 Hybrid Capacitors
The strategy of coupling EDLC electrode materials with pseudocapacitive materials into either
the same electrode or the same device is a promising way to make new devices that bridge the
gap between capacitors, and their fast response, high power capabilities, and batteries and their
energy storage capabilities. Manufacturers can make hybrid electrodes by coating carbon
nanotubes with conducting polymers. This design resulted in electrodes with 170 F/g
capacitance in aqueous electrolytes, compared to just 80 F/g for a typical carbon nanotubes
electrode. Manufacturers can make hybrid devices by coupling carbon negative electrodes with
pseudocapacitive transition metal oxides (TMO) or conducting polymer positive electrodes. In
this design, the ratio of the weights of positive and negative active material on the two
33
electrodes becomes an important device parameter. Researchers also have opportunities for
hybridization of capacitors with even mature battery technologies. The lead‐carbon battery is a
hybridization of traditional lead‐acid technology with carbon capacitors, and large‐scale
demonstration projects are in progress for the stationary power regulation market (Huggins,
2010, p. 247). Typically, hybridization results in increased energy density with decreased
cycleability, so manufacturers must understand the application requirements when deciding
whether a hybrid device is desirable.
1.2 Mechanical Energy Storage
1.2.1 Pumped Hydro
1.2.1.1 Description of Technology
Conventional pumped hydro uses two water reservoirs, separated vertically. During off peak
hours, operators pump water from the lower reservoir to the upper reservoir. When required,
the operators reverse the water flow to generate electricity. Some hydro plants have large
reservoirs resulting in storage capability, and operators can dispatch them in a similar manner
as a pumped hydro – meaning they can stop and start the units depending on grid needs, while
retaining the water in the large reservoir to support rapid ramping and regulation.
Underground pumped storage, using flooded mine shafts or other cavities, is also technically
possible. Industry representatives are currently investigating these options. Pumped storage
developers have also used the open ocean as the lower reservoir. Engineers in Japan first built
the 30 MW seawater pumped hydro plant in 1999 (Yanbaru Project).
34
Figure 5: Cutaway Diagram of a Typical Pumped Hydro Plant
(MacKay, 2007)
Taking into account evaporation losses from the exposed water surface and electrical
conversion losses, operators can regain approximately 70percent to 85percent of the electrical
energy used to pump the water into the elevated reservoir (ESA, EPRI, 2010a, p. 4‐2). The
technique is currently the most cost‐effective means of storing large amounts of electrical
energy on an operating basis, but it entails significant capital costs due to the size of the
projects. The presence of appropriate geography is also a critical factor. Pumped storage,
however, is cost competitive with other technologies when looked at on a $/kW and $/kWh
perspective.
The relatively low energy density of pumped storage systems requires either a very large body
of water or a large variation in height. For example, 1,000 kilograms of water (1 cubic meter) at
the top of a 100 meter tower has a potential energy of about 0.272 kWh. Pumped storage has
the ability to store a significant amount of energy by having a body of water located on a hill
relatively near – but as high as possible above – a second body of water. In some places this
layout occurs naturally, but in other locales, engineers have had to create one or both bodies of
water. Analysts refer to projects in which both reservoirs are artificial and in which no natural
waterways are involved as ʺclosed loopʺ and have much less environmental impact.
Developers are currently investigating over sixty sites as closed loop pumped storage around
the country, with the potential for over 50 GWs of energy storage.
This system may be economical from a regional perspective because it flattens out load
variations on the power grid. As a result, pumped hydro allows thermal power stations (such
35
as coal‐fired plants and nuclear power plants) and renewable energy power plants that provide
base‐load electricity (base load power plants) to continue operating at peak efficiency, while
reducing the need for ʺpeakingʺ power plants that use costly fuels. However, capital costs for
purpose‐built hydro‐storage are relatively high and may require the construction of new
transmission lines.
Over 129 GW of pumped storage are in operation worldwide, totaling approximately 3percent
of global generation capacity. Pumped storage is the most widespread energy storage system in
use on power networks. Its main applications are for energy management, energy imbalance,
frequency control, and provision of system reserves. Italy and Switzerland first developed
pumped hydro in the 1890’s. By 1933, engineers had made reversible pump‐turbines with
motor‐generators available. Operators are now using adjustable speed machines to improve the
operating range of the units and, more importantly, provide fast response for frequency
regulation. Pumped hydro is available at almost any scale with discharge times ranging from
several hours to a few days.
1.2.1.2 Technology Status
Pumped hydroelectric energy storage is a large, mature, and commercial utility‐scale
technology currently that utilities use at many locations in the United States and around the
world. This application has the highest capacity of the energy storage technologies that experts
have assessed.
Developers may size these projects up to 4,000 MW. Pumped hydro plants have very long lives
(typically 50 years or so) and fast response times enable them to participate equally well in
applications, including:
• Voltage and frequency regulation ‐ Balancing supply and demand by providing
ancillary services to the grid.
• Spinning reserve and non‐spinning reserves markets ‐ Generation capacity that is on‐
line but unloaded and that can respond within 10 minutes to compensate for generation
or transmission outages. “Frequency‐responsive” spinning reserve responds within 10
seconds to maintain system frequency. Spinning reserves are the first type used when
shortfalls occur.
• Energy arbitrage – Storing energy when electricity is abundant/cheap, and discharging
into the grid when electricity is more valuable.
• System capacity support – Providing additional energy resources to supplement
generation capacity.
While the siting, permitting, and associated environmental review processes can take many
years, investors and advocates have exhibited a growing interest in re‐examining the
36
opportunities for pumped hydro in the United States, particularly with the large amounts of
wind and new nuclear power generation that various states may deploy over the next few
decades. As can be seen in Figure 6 below, the number of FERC preliminary permits under
development are significant and are primarily in the western interconnect. Analysts consider
most of these sites as closed loop systems with no new dams on main stem rivers, which may
ease the regulatory process and mitigate environmental concerns. Also, developers of many of
these projects are exploring adjustable speed technologies which provide greater frequency
regulation and reserves capability.
Figure 6: Proposed Pumped Storage Projects in the United States
(HDR/DTA, 2011)
In 2011, EPRI began undertaking research to better estimate the future costs of building new
pumped hydro facilities.
1.2.1.3 Technology Limitations and Opportunities
Pumped storage plants generally entail long construction times and high capital expenditure.
To expand the availability of sites, developers have investigated the use of underground
37
reservoirs and proposed several new underground pumped storage projects. Recent examples
include the proposed Summit project in Norton, Ohio and the Mount Hope project in New
Jersey, which would have used a former iron mine as the lower reservoir. Although cost
estimates for these projects are higher than for surface projects, their use might greatly expand
the number of pumped storage sites. Eagle Crest Energy Company proposed to develop the
Eagle Mountain Pumped Storage Project (Eagle Mountain) located near the towns of Eagle
Mountain and Desert Center in Riverside County, which could nominally provide up to 1,300
MW of generating capacity. The upper and lower reservoirs would be formed from existing
mining pits. In addition, Riverbank Power, which merged with Symbiotics LLC in December
2010, developed a number of Closed Loop Pumped Storage Projects (CLPS) throughout the
United States. The Parker Knoll CLPS (FERC project 13239) would provide flywheel‐type
regulation services for up to 500 MW using state‐of‐the‐art, variable speed pump generating
units.
In addition to CLPS projects, whereby Riverbank builds both the upper and lower reservoir, the
company is also actively pursuing underground storage. The company has created an
underground pumped storage hydro‐electric generation concept called Aquabank. Riverbank’s
Aquabank technology combines traditional pumped storage hydro‐electric power generation
with deep mining techniques to generate power. These alternative power generation facilities
will use an existing water source to create electricity from turbines located underground. The
typical facility will have an installed capacity of 1,000 megawatts, representing enough
electricity to power more than 300,000 homes (Riverbank). The Aquabank system temporarily
diverts water from a surface source down 600 meter shafts to an underground powerhouse. At
the powerhouse, the water turns turbines creating emission‐free electric power. A transformer
then harnesses this newly generated power and sends it to the power grid to help address
periods of peak consumption. Once through the turbines, underground reservoirs temporarily
store the water at approximately the same depth as the powerhouse. To complete the cycle,
operators empty the reservoirs by pumping the water back to its original source using lower
cost power from traditional and renewable power sources. Developers are currently
considering two underground projects, one in Wiscasset, Maine, and another in Sparta
Township, New Jersey.
Another example of innovating underground storage is a Santa Barbara, California‐based
startup called Gravity Power, LLC. This company is developing a system called the Gravity
Power Module™ (GPM). The heart of the system is a reversible pump‐turbine and motor‐
generator sitting atop an underground “water circuit” consisting of two sealed water‐filled
vertical shafts – a main bore called the “storage pipe” and a secondary “return pipe” bore.
Within the main bore rests a large piston called the “weight stack.” The weight stack is the
38
means of storing energy. Water flows through the two sealed bores at the top and bottom,
circulating within the system around the weight stack.
Drawing on grid power, the reversible drive turbine pumps the water down through the return
pipe and up the main bore below the weight, lifting the weight to store energy. To release
stored energy when power demand exceeds supply, the weight descends, forcing the water up
through the return pipe, driving the pump‐turbine, and generating electricity in the process.
The speed the weight drops determines the power of the GPM, while the length of the storage
pipe and the mass of the weight determines the amount of energy stored in a GPM.
Figure 7: Gravity Power ModuleTM
(Gravity Power, LLC, 2011)
A “peaking” capacity GPM facility will have a drive shaft 10 meters (33 feet) in diameter and
awatts
s of
1,000‐2,000 meters deep (about 3,200 to over 6,500 feet deep). The company claims that
operators can cluster the system together. Eight such shafts could provide up to 150 meg
each for four hours. Up to sixteen GPMs clustered together could supply a total of 2,400MW on
2.5 acres of land – essentially 1 gigawatt per acre which operators can locate anywhere they can
construct a deep hole, independent of the need for an upper and lower reservoir. Gravity
Power also claims a ramp speed from zero to full power in about 20 seconds and efficiencie
39
75 to 80 percent over a broad power range, from smaller ancillary service applications to full‐
scale peaking plants.
While the Gravity Power system will provide pumped storage performance with location
flexibility at the low end of traditional pumped storage cost, engineers can take advantage of
existing abandoned mine shafts that provide an already constructed hole. The technology may
also be well‐suited to desert environments where emergency water supply has value.
Another new concept involves using wind turbines or solar power to drive water pumps
directly, in effect creating an “Energy‐Storing Wind or Solar Dam.” This option could provide a
more efficient process and smooth out the variability of energy captured from the wind or sun
(Buenoa and Cartab, 2006; Popp).
In addition, engineers can use pumped sea water to store energy. For example, the 30 MW
Yanbaru project in Okinawa was the first demonstration of seawater pumped storage.
Developers have recently proposed a 300 MW seawater‐based project on Lanai, Hawaii and
several seawater‐based projects in Ireland. Tidal lagoons could present another seawater
option. Operators allow seawater to flow into the lagoon at high tide when the water level is
roughly equal to either side of the barrier. The potential energy difference is close to zero at this
point. Then the operators release the water at low tide when a head of water has been built up
behind the barrier and there is a far greater potential energy difference between the two bodies
of water. As operators recover the energy used to pump the water, they will have multiplied
the energy to a degree depending on the head of water built up behind the barriers. Developers
can further enhance the process by pumping more water at high tide with intermittent
renewables, thus increasing the head with renewable energy (MacKay, 2007). In addition to
potential environmental impacts, downsides of this method include having to locate the
generator below sea level and the frequent need to remove the marine organisms that tend to
grow on the equipment and disrupt operation.
1.2.2 Compressed Air Energy Storage
1.2.2.1 Description of Technology
Compressed air energy storage (CAES) uses off‐peak electricity to compress air and store it in a
reservoir, either in an underground cavern or aboveground pipes or tanks. When utilities need
electricity, operators heat the compressed air by combustion, thus expanding the air and
passing it through a conventional turbine to develop electricity (EPRI, 2010a, p. 4‐4).
40
Figure 8: CAES Plant with Underground Storage
(EPRI,
2010a, p. 4-6)
1.2.2.2 Technology Status
Developers built the worldʹs first CAES plant (290 MW) in Huntorf, Germany and have
operated it since 1978. Developers then built a modern 110 MW plant with a storage capacity of
2,700 MWh in McIntosh, Alabama, which has been operational since 1991 (Cavallo, 2007).
Currently these two plants are the only CAES plants that are operational in the world. These
plants are so‐called “first generation” CAES plant because they rely only on the basic CAES
structure (see description of technology). They are differentiated from “second generation”
CAES, which in concept improves the energy recovery efficiency due to modifications of the
structure.
The Huntorf facility has an efficiency of 42percent and the McIntosh facility has an efficiency of
54percent (RWE, 2010). At times, operators have used the Huntorf facility to level the variable
power from wind farms (Greenblatt, 2007).
41
Figure 9: Aerial View of CAES Plant in Huntorf, Germany
(Photo Credit: Alstom)
Figure 10: Aerial View of CAES Plant in McIntosh, Alabama
(Photo Credit: Energy Storage and Power, LLC)
Other plants are in development, including a 260 MW plant in Iowa and a 2,700 MW plant in
Ohio (Greenblatt, 2007). Developers intend the Iowa facility to provide storage for wind
energy. A number of researchers have studied the significance of CAES for wind power storage
(Swider, 2007; Cavallo, 2007; Bullough, 2004). Other current research on CAES focuses on its
limitations and how to address them. The following sections discuss this analysis.
1.2.2.3 Technology Limitations
Current compressed air energy storage (CAES) systems suffer from two primary limitations:
42
1. Reduced roundtrip efficiency associated with the cooling/reheating process, and
2. CO2‐emissions produced by the reheating process. Air cooling between compression
stages, although necessary, represents a loss of heat energy. Direct combustion with
natural gas usually performs the air reheating before the expansion turbine, producing
CO2 emissions.
1.2.2.4 R and D Needs and Opportunities
Researchers have focused efforts to improve the current state of CAES by addressing these
limitations. They have proposed new or modified CAES systems, with the most notable being
the advanced adiabatic (AA) CAES system. Unlike current CAES systems, which use
intercooling and aftercooling for heat removal, the AA‐CAES system uses a thermal energy
storage unit that absorbs heat from the hot compressed air and saves the heat energy for later
use to reheat the air before expansion (Bullough, 2004). Theoretically, this process can achieve
higher roundtrip efficiency because the thermal energy storage unit stores heat energy that
would otherwise be lost by cooling. In addition, because the thermal energy storage unit
accomplishes the reheating, the process produces no CO2‐emissions.
Presently, the AA‐CAES concept is in a research and development phase. RWE Power, working
jointly with General Electric, Zublin, and DLR, are developing an AA‐CAES demonstration by
2013 (RWE, 2011). If the demonstration is to be successful, the group must overcome a number
of key technical challenges. For example, without the use of intercooling, the air temperature
inside the compressor can exceed 650 C. This effect, coupled with the high pressure of
compression (up to 100 bar), creates an environment that is too aggressive for modern
compressors. Therefore, a primary R and D need for AA‐CAES is a high pressure/high
temperature compressor design with considerations for material selection, thermal expansion
and thermostresses, sealing concepts, and thermal limitations for bearings and lubrication
(Finkenrath, 2009).
43
Figure 11: Conceptual AA-CAES System
(Alstom Power Ltd., 2011)
Another primary R and D need for AA‐CAES is the development of an efficient thermal energy
storage (TES) unit. At present, researchers have identified a number of candidate options, but
they must undergo further R and D efforts to find the optimum TES system. The main
consideration for researchers is whether they should use a liquid or solid media TES. In a liquid
TES, the compressed air indirectly transfers heat to a liquid media via a heat exchanger. This
setup has the advantage of not requiring a pressurized vessel for the liquid media, but the
process adds cost and complexity due to the addition of the heat exchanger equipment.
Because the TES must operate over a broad range of temperatures (from 50 to 650 C),
researchers must consider suitable forms of liquid media. Engineers have proposed dual media
approaches, such as salt and oil. In a solid TES, the compressed air directly contacts and
transfers heat to a solid media. A solid TES system has the advantages of large surface area for
heat transfer and generally uses inexpensive storage materials. However, it requires a large
pressurized container for the solid media and compressed air. For the solid TES, manufacturers
must pay special attention to the material selection. Possible materials include natural stone,
concrete, fireproof material, and metal. Manufacturers will need to undertake further research
and testing to identify additional candidate materials that absorb and release heat quickly. In
addition, engineers must design the TES pressurized vessel to be sufficiently durable against
gas permeation and leakage and against cyclic stresses caused by high fluctuations in
operational temperatures and pressures. In either TES system, engineers must consider similar
durability for insulation and pipelines.
Manufacturers will need further R and D to address the challenges presented by underground
cavern and turbine technologies used in current CAES plants. For example, the humidity of
44
underground caverns can corrode the underground bore‐hole equipment, the cavern heads,
pipes, and fittings (RWE, 2010). In addition, engineers have not designed modern turbines to
handle efficiently the high pressure and flow rate fluctuations required of the CAES process.
Researchers have proposed a solution using a turbine with adaptive stages, also known as a
sliding pressure air turbine.
In addition to the AA‐CAES modification, engineers have also proposed so‐called 2nd
generation CAES cycles (Nathamkin, 2008). These methods retain the cooling/reheating process
of conventional CAES but make augmentations elsewhere in the cycle to improve the overall
efficiency. Two such 2nd generation CAES augmentations use bottoming cycles with air
injection and air chilling. In either case, an auxiliary compressor‐turbine cycle, or bottoming
cycle, produces hot combustion gas that reheats the stored air via a recuperator heat exchanger.
After the recuperator, the stored air expands across a turbine. In the air injection setup,
operators inject some air from the turbine exit into the bottoming cycle at the point of
combustion. The rest of the air undergoes a second stage expansion across another turbine. In
the air chilling setup, the air at the turbine exit is colder, and the bottoming cycle compresses
the air before combustion. These 2nd generation CAES concepts have the benefit of only
requiring standard turbomachinery technology and can be readily implemented. R and D
efforts in this area should focus on a cost/benefit analysis of such systems and development of
demonstration systems.
45
Figure 12: Conceptual 2nd Generation CAES with Air Injection
(Energy Storage and Power LLC, 2011)
(Energy Storage and Power LLC, 2011)
Yet another CAES modification uses an air saturator to increase the humidity of the stored air.
The humid air is injected into a bottoming cycle at the point of combustion. The resulting
exhaust gas expands across a turbine.
46
Figure 13: Schematic of CAES with Bottoming Cycle and Humidification
(Pearl Street Inc, 2011)
Up until this point, the discussion has focused on large‐scale CAES using underground caverns
to store compressed air. However, when underground caverns are not available, engineers
should consider aboveground storage using man‐made tanks. This type of CAES system,
however, is only suitable for small‐scale storage. Engineers need to undertake further R and D
efforts to compare small‐scale CAES with other small‐scale storage technologies to determine if
and when CAES is a better option, in terms of cost and performance.
In all the aforementioned CAES concepts, operators allow the storage pressure to vary
according to the supply and demand of compressed air. Purtz proposes a new CAES concept
that calls for the use of an isobaric (constant pressure) storage system (Purtz 2010). Operators
maintain constant pressure in the storage tank by submerging the tank underwater in a large
lake or ocean and exposing the air to the hydrostatic pressure of the surrounding water. The
immediate benefit of a constant pressure air storage device is that the turbine input will also be
at a constant pressure. As a result, developers can use currently available turbines instead of
specialized turbines to handle variable pressure. They can also combine this concept with AA‐
CAES. However, these benefits come at the cost of added complexity. Developers must
consider material wear and insulation in a high pressure, possibly saline environment, costs and
technical challenges associated with the construction of a floating CAES platform with flexible
pipeline to the underwater tank, and concerns about the overall environmental impact.
47
Figure 14: Schematic of AA-CAES with Isobaric Storage
(Jurgen Purtz, 2011)
Developers are pursuing other CAES concepts at a research and development scale, including
the concept of Regenerative Air Energy Storage (RAES). In this approach, operators compress
and expand air nearly isothermally (that is, at close to a constant temperature). As a result, the
technology wastes less energy and creates more energy in the form of compressed air. A
promising approach to maintaining near‐isothermal conditions in the compression or expansion
chamber is to spray a liquid with high heat capacity into the chamber to cool it during
operation. If a sufficient quantity of the liquid is present, the great majority of the heat will
transfer to the liquid, so that the overall rise in temperature is small. Operators can manage the
low‐temperature heat in the liquid much more easily than very high‐temperature air. They can,
for example, use the heat for space heating or store it as warm liquid that can be re‐used during
expansion.
1.2.3 Flywheels
1.2.3.1 Description of Technology
Flywheel energy storage (FES) works by accelerating a rotor (flywheel) to a very high speed and
maintaining the energy in the system as rotational energy. When engineers extract energy from
the system, they reduce the flywheelʹs rotational speed as a consequence of the principle of
conservation of energy; adding energy to the system correspondingly results in an increase in
48
the speed of the flywheel. Most FES systems use electricity to accelerate and decelerate the
flywheel, but engineers are developing devices that directly use mechanical energy. Advanced
FES systems have rotors made of high strength carbon filaments, suspended by magnetic
bearings, and spinning at speeds from 20,000 to over 50,000 rpm in a vacuum enclosure. Such
flywheels can come up to speed in a matter of minutes.
Flywheels are mechanical devices that have large amounts of inertia and store energy in the
form of rotational motion. The main flywheel components include a rotor, a stator, and
bearings. Operators accelerate a cylindrical assembly, often referred to as the rotor, to high
speeds from an energy source, such as grid electricity, and store the energy in rotational kinetic
form. Operators maintain the energy as long as the flywheel. When the operators need energy,
they convert the kinetic energy (back to electricity for example) by slowing the flywheel.
The stator, which is stationary relative to the rotor and is connected to the utility grid through
electronics, typically supports the massive rotating cylinders. Excess grid electricity drives a
motor that spins the flywheel, transferring and storing the electricity in the form of rotational
kinetic energy. Operators then discharge the device by running a generator that feeds back into
the grid.
Modern flywheel energy storage (FES) systems rely on magnetically levitated bearings that
mitigate wear and tear on bearings, thereby extending the system’s life. Operators maintain
efficiency by locating the FES in a vacuum‐like environment to reduce drag. Developers first
used mechanical bearings, but recent advances to improve performance and cost‐effectiveness
have made high temperature superconductor bearings (HTS) a popular choice for bearing
systems to reduce frictional losses and extend energy storage time.
The choice of material for the cylinders depends on the application, and manufacturers must
consider the system requirements for cost, weight, size, and performance. One of the primary
limits to flywheel design is the tensile strength of the material used for the rotor. Generally
speaking, the stronger the disc, the faster operators can spin it and the more energy the system
can store. The flywheel will shatter when it exceeds its tensile strength, releasing all of its
stored energy at once. Cylinders made of solid and dense steel are limited to a lower tip
velocity at the rim (200 to 375 m/s) than those made of lighter and stronger composite materials,
which can attain much higher speeds (600 to 1,000 m/s) (ESA).
1.2.3.2 Technology Status
Developers have matured flywheel technology through the advent of strong, lightweight
materials, microelectronics, and magnetic bearing systems. Operators have tested the present
designs of US Flywheel Systems and showed that power densities at designed speed of 110,000
RPM will exceed 11.9 kW/kg with an in‐out efficiency of 93percent (Vit et al, 2004). Beacon
49
Power is currently developing and demonstrating megawatt‐scale flywheel plants with
cumulative capacities of 20 MW for frequency regulation applications.
1.2.3.2.1 Maturity and Availability
Researchers have primarily focused on improving the rotor component, which is responsible for
sustaining energy through spinning
Aerospace and uninterrupted power supply (UPS) applications have driven the development
and purchase of high‐powered flywheels. Recent interest from investors has spurred
developments for commercial FES. Researchers and suppliers are striving for low‐cost and
efficient systems that can store energy for longer operation periods on the scale of hours.
FES activities and efforts in the 1990s have shown promise in development and have created the
push for larger‐scale systems. In 1998, Chubu Electric Power/Mitsubishi Heavy Industry
succeeded in developing a 1 kWh FES using an axial‐type superconducting magnetic bearing
(Nagaya et al, 2003). The US Argonne National Laboratory developed a 2.25 kWh FES using a
high temperature superconductor (HTS) bearing (Mulcahy et al, 2011). Boeing is currently
developing a 5kWh/50kW FES in collaboration with Argonne (Boeing Phantom Works). And in
Germany, the Piller group is attempting to realize a 10 kWh/2MW class FES for uninterrupted
power supply (Siems et al, 2004).
Recently, manufacturers have deployed systems for telecommunication applications. Such
systems are rated at 2 kWh and 6 kWh. A flywheel farm approach can also attain megawatts
for minutes and hours. Beacon Power has successfully developed a flywheel network
consisting of forty 25 kWh wheels capable of storing 1 MW for an hour, both efficiently and
with a small footprint (Beacon, 2011).
1.2.3.2.2 Performance
Two classes of flywheel rotors exist. Advanced composite materials, such as carbon fiber and
graphite, have high strength‐to‐weight ratios, resulting in high specific energy storage by virtue
of the higher rotational speeds thus enabled. The second class relies on steel as the main
structural material, resulting in large diameters, slower rotational speeds, and lower power and
energy densities [Liu and Jang, 2007].
Material choice is important in determining the energy storage in flywheels. The energy stored
is proportional to its mass and the square of its rotational speed, or more accurately its angular
velocity. At the same rotational speeds, higher mass flywheels are capable of more energy
storage. The tensile strength of the flywheel material limits the maximum specific energy
density (normalized by mass). Steel with fiber‐reinforced composites typically have the highest
50
tensile strength. Moreover, experts deem these materials to be safer at higher rotational speeds
during failures because the flywheels tend to delaminate and disintegrate gradually from the
outer circumference rather than explode catastrophically (Demachi et al, 2002).
1.2.3.2.3 Geometry
Engineers choose the geometry to maximize the energy density, specific energy, or both
[Wagner et al, 2002]. In practice, manufacturers achieve these goals either by placing the mass
farthest away from the axis of rotation or increasing the material density. The shape of the rotor
also influences the energy storage capability. The specific energy depends on the rotor’s shape,
characterized by a shape factor as shown in Table 2. The shape factor is a measure of the shape
efficiency of the rotor in the stress‐limited case (Liu and Jang, 2007).
Table 2: Flywheel Shape Factors
(Liu and Jang, 2007)
1.2.3.2.4 Length
The optimal length is directly related to dynamic considerations (Liu and Jang, 2007). Rotating
bodies undergo rigid body and flexural resonance modes, referred to as criticals. When these
critical resonances are excited, these modes lead to high losses. If engineers choose bearing
systems that are too stiff, these criticals are excited above the operating frequency of the rotor.
Engineers must therefore choose bearing systems at the proper stiffness so that the rotating
bodies can pass these criticals at low speeds and operate with low damping losses. Engineers
choose the length to avoid excitation of the criticals, typically the maximum safe length below
the speed of the rotor when it is running at its speed of maximum stress (Liu and Jang, 2007).
1.2.3.2.5 Bearings
Bearings support the spinning rotor. Bearings can be magnetic or can come in the form of balls.
Bearings are characterized according to weight, loss, cost, and longevity. Magnetic bearings
have the advantage of high spin speeds and good vibration‐suppression characteristics. Ball
bearings have benefited from advancement in material (ceramics and hardened steel) [Liu and
51
Jang, 2007]. Lubricant life rather than material fatigue help determine the longevity of the
bearings. Keeping operating temperatures low ensures the lubricant does not deteriorate. At
higher spin speeds, magnetic bearings are a better choice. However, below a certain speed ball
bearings have a weight advantage due to relatively low drag losses. This speed cutoff is
application dependent, but it generally falls between 20,000 to 40,000 revolutions per minute
(RPM) (Wagner et al, 2002).
In an era when batteries typically have relatively shorter equipment life and higher operational
and maintenance expenses, flywheels have become more attractive to electric power companies.
In addition, flywheels are better suited for applications experiencing frequent cycling (Federal
Technology Alert, DOE/EE‐0286).
1.2.3.2.6 Cost
The following cost data are from a DOE report outlining a federal energy management program
comparing flywheels and battery storage technologies (Federal Technology Alert DOE/EE‐
0286). Flywheel purchase costs can range from $100/kW for larger and/or lower RPM models to
$300/kW for smaller and/or higher RPM models. Installation costs, including electrical
connections, are typically simple and inexpensive, ranging from $20/kW to $40/kW.
Operating and maintenance costs depend on the flywheel type. Internal staff can handle the
routine maintenance, which typically has a minimal and low cost. Vendors offer service
contracts for about $5/kw/year that can be more expensive than internal maintenance.
However, these contracts typically include around‐the‐clock response and extended warranties.
Bearing replacements typically range from $5/kW to $15/kW, depending on the flywheel design.
Manufacturers can replace the vacuum pump at about $5/kW. Because engineers design
flywheels for a long life of about 20 years, these components do not require frequent attention,
usually between 5‐10 years. Standby power consumption is about $5/kW/year for lower RPM
flywheels using mechanical bearings or a combination of mechanical and magnetic. Higher
RPM systems using only magnetic bearings typically cost about $0.50/kW/year.
1.2.3.3 Technology Limitations
The most significant limitation of flywheels lies in their relatively modest capacity for energy
storage (Dell). They are best suited as surge‐power devices that experience frequent charge and
discharge of modest energy quantities at high‐power rates (Dell). The comparatively small
amount of electrical energy that they can store limits flywheel applications to locally generated
or distributed electricity sources.
52
The attainable speed affects the deliverable energy back to the grid. At low speeds, flywheels
cannot deliver the rated power. For example, over a 3:1 speed range, a flywheel will deliver
about 90percent of its stored energy to the electric load (ESA, 2011).
Flywheels are capable of high efficiencies, which offer a significant advantage over other energy
storage technologies. However, friction loss in the flywheel rotation is one of the significant
limitations of this technology and hinders the potential for FES as a long‐term storage option.
Analysts have estimated that a 200‐ton flywheel can experience up to 200 kW of power loss.
Assuming an instantaneous flywheel efficiency of 85percent, the overall efficiency would drop
to 78percent after five hours of operation and down to 45percent after one day (Ibrahim).
Currently, FESs using mechanical and/or magnetic bearings can deliver power for only short
durations (5 ‐ 30 seconds). The bearings are subjected to parasitic losses such as hysteresis,
friction, and eddy current losses, the latter two leading to undesirable temperature rise
(Koshizuka, 2002). This temperature increase creates a necessity for cooling systems to prevent
component degradation.
In applications where utilities use flywheels as backup power, the fact that flywheels can only
perform in durations of seconds limits their use. The backup period is commonly about 15
seconds, which is enough time for backup generators to powerfully and become activated as the
primary backup option. Thus, operators should not use flywheels alone for backup power,
without batteries and/or fuel‐fired generators. Also, operators should not use flywheels
without batteries if a fuel‐fired generator cannot reliably come up to full power in 10 seconds or
less (Federal Technology Alert DOE/EE‐0286).
1.2.3.4 R and D Needs and Opportunities
Overall, manufacturers have proven flywheels to be an ideal form of energy storage due to their
high efficiency, long cycle life, wide range of operating temperature, freedom from depth‐of‐
discharge effects, and higher power and energy density—on both a mass and volume basis
(Kohari et al. 2005; Long et al., 2002; Pieroneck et al., 1997; Ries and Neumueller, 2001).
Flywheels are essentially surge‐power devices. They naturally complement battery storage
devices and would serve well in these applications. As other applications involving wind
turbines and photovoltaic arrays grow in the renewable energy development effort, flywheel‐
based buffers could help store the large, frequent, and rapid energy fluctuations. This strategy
could potentially remove the need for downstream power electronics to track such fluctuations
and improve the overall electrical efficiency (Dell). As utilities consider batteries as a storage
medium for wind power and photovoltaic array applications, they may also want to evaluate
battery‐flywheel combinations (Dell).
53
Unfortunately, FESs still entail high costs and the above‐mentioned technology limitations that
hinder widespread adoption. The areas for advancement that researchers and industry are
targeting over the next decade (Liu and Jang, 2007) include:
• Improved Safety ‐ Research into methods for predicting design and intelligent fault
protection can enhance system safety. Development into new software and checkout
procedures to address these problems can make a substantial impact on FES
management and adoption (Liu and Jang, 2007).
• Increased Specific Power/Energy Density ‐ The kinetic energy stored in a flywheel is
proportional to its moment of inertia, a function of its mass and shape. Advances in
increasing these two flywheel properties will certainly allow scalability of FES. In
addition, the kinetic energy is proportional to the square of flywheel speed. However,
the speed limit is set by the tensile strength of the flywheel, which governs the maximal
inertial load during operation. Light materials develop lower inertial loads at a given
speed. Therefore composite materials, with low density and high tensile strength, are
preferred choices for storing energy (Gabrys, 2001). While previous research efforts
have designed and tested such materials, current research is ongoing and necessary to
develop higher performance flywheel materials and fabricate them at lower cost.
• Improved In‐Out Energy Efficiency (Koshizuka, 2002) ‐ Efficiencies of power electronics
govern the capability to charge and discharge FES. By increasing the efficiencies of
electric power converters, motors, generators, and refrigeration systems, operators can
improve the overall FES efficiency.
• Predicted performance and long‐term durability in composite flywheels (Tzeng et al,
2006) ‐ Accurate models of rotor behaviors can be useful in determining their long‐term
operation and assessing the health of the system. Developers can use such models for
fault detection purposes and to predict proximity to failure. Manufacturers have
employed techniques to model elastic, viscoelastic, and fatigue to predict rotor behaviors
(Tzeng et al, 2006). They need test and design procedures to consider the long‐term
behavior of flywheels, specifically creep, stress relaxation, fatigue, and composite
fracture tendencies.
• Superconducting Magnetic Bearings (SMBs) ‐ The discovery of high temperature
superconductor (HTS) has spurred research interest and activity into SMBs for their
extremely low rotation loss and expected strong levitation forces (Hassenzahl, 2001).
Improvements to the homogeneity of the magnetic field distribution in the permanent
magnet circuit and the superconducting bulk stator have reduced rotation loss
(Koshizuka, et al, 2006). These advances have the benefit of mitigating eddy currents
54
that lead to inefficiency and material fatigue. SMBs also have the responsibility to
produce a sufficient levitation force to support the weight of massive rotors. During
operation, rotors may experience gradual fall from decreased levitation forces due to
flux creep of the superconducting material (Koshizuka, et al, 2006). Current methods
have shown to be effective in handling flux creep. Pre‐loading the SMBs and/or excess
cooling the bearings below a certain temperature can suppress the fall. SMBs have low
stiffness and damping, leading to undesired vibrations and lower efficiencies. Active
magnetic bearings (AMBs) have offered a solution at the present, though researchers
need additional efforts to reduce their rotation losses and energy consumption
(Koshizuka, et al, 2006). In using AMBs, operators will need development in control
system design.
• Fabrication Technology (Koshizuka, et al, 2006) ‐ The mechanical strength of the
flywheel rotor is critical in high speed rotations. At these speeds, the centrifugal force
applied on the outer rotor is strong. Material choice is critical to handle these conditions.
The properties of carbon fiber‐reinforced plastics are suitable to handle such forces,
though manufacturers need further research to identify the optimum conditions for the
fabrication process. Such research includes determining the ideal temperatures of the
resin and mandrel, winding speed of the fibers, and winding patterns (Koshizuka, et al,
2006). To increase in scale, FES will need to have stiff, stable, and large flywheels.
Manufacturers must keep the cost of this fabrication low, especially for HTS bulk
materials.
• Improved Overall Efficiency by Reducing Loss (Liu and Jang, 2007) ‐ Increasing voltage
decreases current losses and eliminates redundant transformer steps. But isolation
problems and limitation in power electronics arise with high voltage. So far,
manufacturers have not constructed flywheels over 10 kV for this reason. Recent
progress in semiconductor technology offers potential improvements (faster switching,
lower costs). Vacuum enclosures represent the best way to eliminate windage loss.
Manufacturers will need better heat exchangers in tandem to remove the high heat
generated by rotating parts. Engineers can achieve improvements to the radial vibration
of rotor axis by using active magnetic bearings. However, the rotation loss in using
AMBs is significant and can negate the benefits of using SMBs in the first place.
Manufacturers will need further developments to improve the basic structure of FES,
including SMB and vibration control technologies (Koshizuka, et al, 2006). According to
researchers, if the total loss fraction of FES is reduced to about 0.1percent per hour of
operation, the flywheel systems can be effectively used for a variety of applications,
including load leveling of power utilities (Koshizuka, et al, 2006).
55
1.3 Thermal Energy Storage
Gas or electric heating and electrically powered air conditioning currently meet most residential
and commercial heating and cooling needs, but use a significant amount of energy that typically
comes from less efficient “peaker” power plants. Various technologies can reduce this energy
requirement appreciably. For example, better insulation can reduce the heat transfer to and
from the environment. Thermal energy storage systems, which large‐scale power plants also
employ, represent another option. The two general types of thermal storage mechanisms
include one based upon the use of the sensible heat in various solid and/or liquid materials and
a second involving the latent heat of phase change reactions (Huggins, 2010).
Engineers can add energy to a material by simply heating it to a higher temperature. The
energy involved in changing its temperature is called “sensible heat,” and its amount is the
product of the specific heat and the temperature change. Technologies can transfer this sensible
heat to another, cooler material, or to the environment, by radiation, convection, or conduction.
Ultimately, this method stores energy in the form of heat and transfers it later.
Phase change reactions refer to the use of materials with a high heat of fusion that, after melting
and solidifying at a certain temperature, are capable of storing and releasing large amounts of
energy. The materials absorb or release heat when they change from solid to liquid and vice
versa. As a simple example of phase change material with high latent heat, water can go from
solid (ice) to liquid (water) to gas (steam).
1.3.1 Solar Thermal
1.3.1.1 Description of Technology
The integration of thermal energy storage with solar energy represents a direct grid application
for thermal energy storage. Recent and ongoing improvements in solar thermal generation
technologies, coupled with the need for more renewable sources of energy, have increased
interest in concentrating solar thermal power (CSP). Unlike solar photovoltaic (PV) generation,
CSP uses the thermal energy of sunlight to generate electricity. Two common designs of CSP
plants ‐ parabolic troughs and power towers ‐ concentrate sunlight onto a heat‐transfer fluid
(HTF), which the technologies use to drive a steam turbine. One advantage of CSP is the
potential for storing solar thermal energy to use during non‐solar periods and to dispatch when
needed. As a result, thermal energy storage (TES) allows CSP to achieve higher annual capacity
factors—from 25percent without thermal storage up to 70percent or more with it (NREL). TES
can also provide backup energy during periods with reduced sunlight caused by cloud cover
(Sioshansi and Denholm, 2010).
56
Adding TES provides several additional sources of value to a CSP plant. First, unlike a plant
that must sell electricity when solar energy is available, a CSP plant with TES can shift
electricity production to periods of highest prices. Second, TES may provide firm capacity to
the power system, replacing conventional power plants as opposed to simply supplementing
their output. Such capability would, however, require storage capacity sized for extended
durations, depending on the system design, and the plant may have to sacrifice revenues when
the sun is shining against revenues when it is not. Finally, the dispatchability of a CSP plant
with TES can provide high‐value ancillary services such as spinning reserves. The storage
medium is typically a molten salt, which has extremely high storage efficiencies in
demonstration systems.
1.3.1.2 Technology Status
There are hundreds of megawatts of concentrating solar power (CSP) facilities with molten salt
energy storage in construction and operation in Spain and in the US, with potentially thousands
of megawatts more in development pipelines. Notable domestic examples are Solar Reserve’s
110 MW Crescent Dunes and 150 MW Rice Solar Energy projects, and Abengoa’s 250 MW
Solana project – all three of which were awarded power purchase agreements (PPA) from major
United States utilities. Utility‐scale CSP facilities have been in operation in California since the
late 1970s. In the 1980s and 1990s, the US Department of Energy sponsored the 10 MW Solar
One and Solar Two demonstration facilities, establishing the characteristics of a CSP power
tower with molten salt storage. This power tower configuration uses sodium and potassium
nitrate salts to collect and store thermal energy and then dispatches that energy to a steam
turbine whenever grid operators require electricity.
1.3.1.3 Technology Limitations and Opportunities
Researchers have tested a variety of fluids to transport the sunʹs heat, including water, air, oil,
and sodium. To date, molten salt provides the best results. Manufacturers use molten salt in
solar power tower systems because 1) it can reach high temperatures at atmospheric pressure
without boiling, 2) it provides an efficient, low‐cost medium in which to store thermal energy,
3) its operating temperatures are compatible with modern high‐pressure and high‐temperature
steam turbines, and 4) it is non‐flammable and nontoxic. In addition, manufacturers use molten
salt in the chemical and metals industries as a heat‐transport fluid, so experience with molten‐
salt systems exists in non‐solar settings.
The molten salt is typically a mixture of 60 percent sodium nitrate and 40 percent potassium
nitrate, commonly called saltpeter. Engineers can include calcium nitrate in the salts mixture to
reduce costs and provide technical benefits. The salt melts at 220 °C (430 °F). Operators must
keep the salt liquid throughout the life of the plant, moving between a “cold” insulated storage
tank at 290 °C (550 °F) and a “hot” tank between 750 and 1050 degrees F (depending upon the
57
technology). The uniqueness of this solar system lies in the ability to disassociate the collection
of solar energy from the production of power. As a result, the technology can generate
electricity irrespective of momentary cloud cover or even at night by using the stored thermal
energy in the hot salt tank. Normally, manufacturers insulate the tanks thoroughly and,
depending upon storage capacity and technology capacity, store the energy for extended
periods of several days or more. Tank size and salt inventory depend highly upon operating
temperature. Lower temperature designs (in example, troughs) store a fraction of the thermal
energy per unit salt.
1.3.2 Thermal Storage for Heating, Ventilation, and Air Conditioning (HVAC)
1.3.2.1 Description of Technology
In a thermal energy storage (TES) system, the device chills a storage medium during periods of
low cooling demand and then uses the stored cooling later to meet air‐conditioning load or
process cooling loads. The system consists of a storage medium in a tank, a packaged chiller or
built‐up refrigeration system, and inter‐connecting piping, pumps, and controls. The storage
medium is generally water, ice, or a phase‐change material (sometimes called a eutectic salt).
TES systems shift cooling energy use to non‐peak times. Analysts generally classify operating
strategies as either full storage or partial storage, referring to the amount of cooling load
transferred from on‐peak to off‐peak. TES systems are applicable in most commercial and
industrial facilities, but they must meet certain criteria for economic feasibility. A system can be
appropriate when maximum cooling load is significantly higher than average load. High
demand charges and a significant differential between on‐peak and off‐peak rates also help
make TES systems economical. They may be appropriate when operators need more chiller
capacity for an existing system or back‐up or redundant cooling capacity.
Besides shifting load, TES systems may reduce energy consumption, depending on site‐specific
design, such as when operators can use chillers at full load during the night. Also, these
systems can reduce pumping and fan energy by lowering the temperature of the water and
therefore the air temperature. As a result, they can affect the quantity of air circulation
required.
Capital costs tend to be higher than a conventional direct‐cooling system, but other economic
factors can reduce such costs. In new construction, ductwork can be smaller, allowing more
usable space. Or a TES system may enable reduction in electrical capacity, reducing the cost of
electrical service for a new or expanding facility.
Load shifting is typically the main reason to install a TES system. Cool storage systems can
significantly cut operating costs by cooling with cheaper off‐peak energy and reducing or
eliminating on‐peak demand charges (P G and E).
58
1.3.2.3 Technology Status
TES is a mature technology that has been available for decades (Schiess, 2009). A number of
companies have commercial products available, including but not limited to BAC, Calmac,
Evapco, Ice Energy, and Sunwell. An example of TES systems for HVAC is the Ice Energy Ice
Bear system, illustrated in Figure 16 TES HVAC Unit with Cooling Coil Integrated into Existing
A/C. This technology is a thermal energy storage system targeted for installation on low‐rise
(fewer than three stories) buildings, which represent over 95percent of the buildings in the U.S.
The system consists of four primary parts, an insulated ice storage tank with main heat
exchanger, an ice make compressor, a refrigeration management system, and a sophisticated
controller. Vendors install the system with the existing (typically) rooftop air conditioning
system and leverage the existing blower subsystem. Operators use the condenser subsystem
to create the nighttime ice stored in the unit as well as for cooling during daytime operation.
Figure 15: TES HVAC Unit with Cooling Coil Integrated into Existing A/C
(Ice Bear Energy, 2011)
The key technology elements are the ice coil design, controller, and control interface. Engineers
designed the ice coil and refrigerant loop subsystem for maximum energy exchange at all
temperatures, using an oil‐less liquid overfeed system. Because engineers physically isolate this
subsystem from the compressor, the device can independently control the proper oil ratio
needed to maintain compressor operation. The refrigerant loop uses an adjustable pump to
regulate ice consumption during operation, and the technology limits the power draw to less
than 300 watts for the pump. The controller is capable of operating in intercept mode, whereby
it receives the signals from the existing thermostat or building control interface. In this
configuration, the end user simply sets the temperature, and the controller optimizes the use of
the stored ice in conjunction with the existing air conditioning unit. The controller is also
capable of more exotic schemes, including one known in the control industry as “PID”
59
(proportional‐integral‐derivative) control. Manufacturers have designed the controller interface
to work with a well‐known time‐series data application and to deliver information using typical
internet technology and browser interface.
Field tests have shown that these devices can shift up to 95percent of the energy needed for
cooling during operation in a single location to the off‐peak period, with one installation
realizing a 40kW peak reduction. The power needed to run the unit during its cooling cycle is
around 300 watts. With time‐differentiated rates (time of use, demand charges, critical peak
prices), single customer installations can receive a significant benefit from TES installations
(extremely short payback periods) (Gunther, 2009).
1.3.2.3 Technology Limitations and Opportunities
Thermal energy storage for commercial HVAC systems is a mature technology approach. The
key to maximizing the effective of such systems to shift cooling load and thus support the
electric grid is appropriate engineering design and implementation. A number of design
options can make TES systems more energy‐efficient than non‐storage systems. Storage
systems let chillers operate at full load all night, versus operating at full or part load during the
day (P G and E, 1997).
Depending on the system configuration, the chiller may be smaller than would be required for
direct cooling, allowing smaller auxiliaries such as cooling‐tower fans, condenser water pumps,
or condenser fans. Manufacturers can limit pumping energy by increasing the chilled water
temperature range and they can cut fan energy with colder air distribution. Storage systems
can also make increased use of heat recovery and waterside economizer strategies (P G and E,
1997).
On the core technology side, few methods of thermal storage could be simpler or more directly
usable. These systems create ice during off‐peak periods (both electricity demand and ambient
temperature) and consume it for cooling during the on‐peak period, maximizing round trip
thermal energy conversion. The devices lower existing equipment stress through the reduced
need for air conditioning compressors to operate during the high‐temperature period of the day
when cooling demand is greatest. This feature also extends the compressor service lifetime.
Likewise, the devices operate the condenser and compressor used to create the ice during the
low temperature period of the day, with the same resultant benefit. Furthermore, the system’s
output does not degrade with temperature. As a result, the maximum capacity of the base
system can be “right‐sized,” with the resulting reduction in compressor capacity (of 10‐
25percent) producing energy savings even when the machines are not using ice.
An advantage of this particular thermal storage system is the ability to deploy the units either
tactically or strategically across the power grid. Utilities can target known problem locations,
60
feeders, or zones in their system, improving localized system stability and grid conditions.
They can also leverage their operational experience and load growth forecasts to defer planned
grid enhancements for feeders, substations, and areas. In addition, a large‐scale deployment of
the units (up to 100’s of MW capacity) defers investment in new peak‐matching generation or
lessens the need to run existing peaking generation. With proper controls and response
feedback, grid operators could use this type of energy storage as a virtual spinning reserve, and
the devices could bid into demand markets or participate in any number of internal and
external load management programs. Even if operators do not use advanced controls, simply
running all available units within a territory flattens the daily load curve of the utility and
defers both investments to support the demand and the running of any peaking units to serve
that demand (Gunther, 2009).
TES systems manufacturers will still need research and investment into a few areas to
maximize the value of the technology. The operating states and control algorithms are well
known and in fact are already available to be explored through the National Renewable Energy
Laboratory (NREL) HOMER tool. As more system designers become familiar with the
operation and potential, they can provide user feedback and help improve or even modify the
operating states and control algorithms. This feedback might provide improvement over the
ongoing monitoring and diagnostics.
1.4 Hydrogen as an Energy Storage System
1.4.1 Description of Technology
Analysts consider hydrogen to be an important energy carrier for the future and, when used as
a fuel, as an alternative to fossil fuels, such as coal, crude oil, and natural gas, and their
derivatives. Hydrogen has the potential to be a clean, reliable, and affordable energy source
and has the major advantage that the primary product of its combustion with oxygen is water
rather than CO2, which is a greenhouse gas. The primary disadvantage of hydrogen is that,
although it is the most abundant element, it is predominantly found chemically‐bound, such as
in water or in hydrocarbons. As a result, the devices must separate the hydrogen from these
compounds.
Hydrogen as an energy storage system is a complicated proposition involving three separate
processes. First, a device must produce the hydrogen. In a grid energy storage application, the
most appropriate production technology is the electrolysis of water using electricity. After
electrolysis produces the hydrogen, operators must store it in either gaseous or liquid form.
Finally, to return electric power to the grid, operators must convert the stored hydrogen to
electricity by either a fuel cell or a combustion gas turbine.
61
For the application of electricity grid storage, proposed hydrogen systems would comprise an
electrolysis unit (otherwise known as an electrolyzer), a compressor and storage system, and,
most likely, a fuel cell. Hydrogen electrolyzer/fuel cell energy storage is a storage concept that
utilizes off‐peak electricity to produce hydrogen from water. The hydrogen serves as an energy
carrier that technologies can store as a liquid or a gas that fuel cells can later use to produce
electricity. In combination with power electronics, hydrogen electrolyzer/fuel cell systems can
provide renewable energy sources with peak shaving, ancillary services, and greater
dispatchability. However, hydrogen electrolyzer/fuel cell technology also can provide
important services for renewable energy in the damping of power fluctuations. Much like
batteries, fuels cells can respond rapidly to power fluctuations from wind turbines. Analysts
expect hydrogen to play a significant role in future energy systems. Developers can use it in
fuel cells to produce electricity directly, with water as the only product. Researchers have
shown that they can use hydrogen directly in internal combustion engines, requiring relatively
minor modifications, as well as in gas turbines. The energy efficiency of fuel cells ranges from
40‐60percent, depending on the fuel cell technology. Fossil fuel combustion‐based systems, on
the other hand, are typically less than 40percent efficient, although combined cycle gas turbine
systems can achieve up to 60percent efficiency. When manufacturers use high temperature fuel
cells, they can generate both electricity and useful heat for heating purposes. This cogeneration
makes it possible to achieve total energy efficiencies of up to 80percent in this manner.
Figure 16: Hydrogen Electrolyzer/Fuel Cell Energy Storage System.
1.4.1.1 Hydrogen Production by the Electrolysis of Water
Industrial production occurs mainly from the steam reforming of natural gas and less often
from more energy‐intensive hydrogen production methods like the electrolysis of water
(Florida Solar Energy Center, 2007). Most manufacturers employ hydrogen near the production
site, with the two largest uses being fossil fuel processing (for example, hydrocracking) and
ammonia production, mostly for the fertilizer market. For the application of electricity grid
storage, analysts envision production via electrolysis of water.
62
The electrolysis of water is a simple method of producing hydrogen. Devices run a low voltage
current through the water, and gaseous oxygen forms at the anode while gaseous hydrogen
forms at the cathode. Typically manufacturers make the cathode from platinum or another
inert metal when producing hydrogen for storage. The theoretical maximum efficiency
(electricity used versus energetic value of hydrogen produced) is between 80–94percent (Kruse).
However, in practice, efficiencies are closer to 60‐70percent when analysts include parasitic
loads. 2 H2O(aq) → 2 H2(g) + O2(g)
1.4.1.2 Hydrogen Storage
For the foreseeable future, manufacturers store hydrogen as a gas at high pressure or as a liquid
under cryogenic conditions. Unlike mobile applications, hydrogen density is not generally a
problem for stationary applications. As for mobile applications, stationary applications can use
established technology:
• Compressed hydrogen in a hydrogen tank (up to 700 bar)
• Liquid hydrogen in a cryogenic hydrogen tank
Underground storage represents another implementation of compressed hydrogen storage,
albeit at lower pressures (up to 100 bars) in underground caverns, salt domes, and depleted oil
and gas fields. ICI has stored large quantities of gaseous hydrogen in underground caverns for
many years without any difficulties (ECN, 1994). The storage of large quantities of hydrogen
underground can function as grid energy storage.
1.4.1.3 Conversion to Electricity
A fuel cell is an electrochemical cell that converts energy from a fuel source into electrical
energy. The reaction between a fuel supply and an oxidizing agent generates electricity. The
reactants flow into the cell, and the reaction products flow out of it, while the electrolyte
remains within it. Fuel cells can operate continuously as long as they maintain the necessary
reactant and oxidant flows.
Analysts classify fuel cells primarily by the kind of electrolyte they employ. This classification
determines the kind of chemical reactions that take place in the cell, the kind of catalysts
required, the temperature range in which the cell operates, the fuel required, and other factors
(Department of Energy, 2011).
• Polymer electrolyte membrane (PEM) fuel cells have low operating temperatures (~ 80 ‐
100 C) that allow for fast start up, as would be required in the envisioned energy storage
system. PEM fuel cells, with efficiencies of 40 ‐ 50 percent, are the primary type being
developed for transportation applications.
63
• Alkaline fuel cells (AFC) have operating temperatures of 90 ‐ 100 C with efficiencies up
to 60percent. Operators primarily use AFCs in space applications where the devices can
ensure pure hydrogen and oxygen, since AFCs are extremely sensitive to CO2 poisoning.
• Phosphoric acid fuel cells (PAFC) were the first fuel cells developers deployed
commercially for distributed generation applications. Typical operating temperatures
are 150 ‐ 200 C with efficiencies of about 40percent. Engineers can increase overall
efficiency in combined heat and power applications. PAFCs have long start‐up times,
which may restrict their use in energy storage system applications.
• Molten carbonate fuel cells (MCFC) are being developed and deployed for distributed
generation. Operating temperatures are 600 ‐ 700 C with efficiencies of 45 ‐ 50percent.
The high operating temperatures allow for internal reforming of a variety of fuels as
well as combined heat and power applications. However, the high temperatures require
long start‐up times.
• Solid oxide fuel cells (SOFC) are also high temperature fuel cells utilizing a ceramic
electrolyte. Typical operating temperatures are 700 ‐ 1000 C with efficiencies up to
60percent. Manufacturers are developing SOFCs for distributed generation and
auxiliary power applications. The high operating temperatures allow for internal
reforming of a variety of fuels as well as combined heat and power applications.
However, the high temperatures also require long start‐up times.
1.4.2 Technology status
Electrolyzers and fuel cells have undergone extensive development in recent years, and several
companies are now beginning to commercialize these devices. In comparison with other energy
storage methods, electrolyzer/fuel cell systems are at present relatively expensive, with costs of
approximately $3,000 to $5,000 per kW. However, analysts expect these costs to decline as
manufacturers make further technological advances and improvements and produce the
systems in greater volumes (Lipman et al, 2005, p. 22).
The three subsystems of a hydrogen energy storage system vary in terms of technology
maturity. Hydrogen storage and compression, for example, are well‐established, mature
technologies for various industrial applications, albeit not necessarily for the subject energy
storage application. Steel and now composite pressure vessels with service pressures of up to
700+ bar are commercially available. The main industrial players for large hydrogen storage
systems are major international companies that are selling industrial gases, such as Praxair, Air
Products, BOC, Linde, or Air Liquide.
64
Similarly, compressors compatible with hydrogen are also commercially available for industrial
applications. For the specific application in an electrolyzer/fuel cell system, the most
appropriate technology would be a diaphragm compressor, using a membrane to compress the
gas rather than conventional reciprocating pistons, which has proven to be reliable and produce
oil and contaminant‐free high pressure hydrogen. Compression energy requirements are
modest and typically in the range of 0.2 ‐ 0.3 kWh/Newton‐meter3 (Nm3) (Pdc Machines, 2011).
Electrolyzers have also found commercial success in industrial applications that require
hydrogen on demand. As with most industrial equipment, the term “high volume production”
is probably not relevant to electrolyzers ‐ serial production may be better terminology. Several
of the key suppliers include Hydrogenics, Proton Energy Systems, and Statoil Hydrogen
Systems. The primary electrolyzer products for Hydrogenics and Statoil are based on bipolar
alkaline technology. They also both have PEM based systems under development. Proton
Energy offers only PEM electrolyzers. Based on company specification sheets, the bipolar
alkaline systems of Statoil have the highest efficiency, requiring 4.3 kWh/Nm3. However, this
total only includes the electrolyzer (Statoil). Complete integrated systems, as specified by both
Hydrogenics and Proton Energy, include not only the power required for the electrolyzer but
the ancillary power for the balance‐of‐plant, including water treatment; cooling, heating and
ventilation; and power conditioning systems. Hydrogenics’ bipolar electrolyzer requires 4.8
kWh/Nm3, while its integrated systems range from 5.0 ‐ 5.4 kWh/Nm3 (Hydrogenics). Analysts
rate Proton Energy’s systems at approximately 5.2 kWh/Nm3 (Proton Energy Systems, 2011).
Manufacturers are demonstrating and deploying commercially fuel cells for distributed
generation and potentially for utility‐scale applications, with the support of subsidies to offset
capital costs. Manufacturers have demonstrated PAFC’s and deployed 250 kW and 400 kW
units commercially since about 1991, primarily by UTC Power. FuelCell Energy has been the
primary company deploying MCFCs. The company’s Direct FuelCell® (DFC®) power plants
range in output from 300 kW up to 2.8 MW, with the first system going online in 1996. Bloom
Energy is selectively deploying SOFC technology with the modular “Bloom Box” 100 kW
systems. PEMFC development has focused on transportation, material handling equipment,
and back‐up power applications, with successful demonstrations. Only Ballard Power Systems
is actively pursuing larger, distributed energy applications. Early stationary generation
demonstrations by Ballard focused on 250 kW units, with the first prototype demonstration at
BC Hydro in 1997 (Ballard, 2009). In 2010, Ballard delivered a 1 MW unit to FirstEnergy for
demonstration in Ohio (Ballard, 2011). Specifications for the 1 MW unit state an overall system
efficiency of 48percent (Ballard CLEARgen, 2011).
65
1.4.3 Technology limitations
When considering hydrogen energy storage systems, the primary limitations include the
maturity of the fuel cell technology, the durability of fuel cells and electrolyzers, capital cost of
fuel cells, electrolyzers, and, to a lesser extent, storage vessels. Efficiencies of fuel cells and
electrolyzers and the scale of fuel cells and electrolyzers with respect to grid storage
applications also limit the technology. Only Ballard has a few demonstration units of PEM fuel
cells (the most appropriate technology for energy storage systems) for stationary applications to
indicate maturity levels. As a result of this immaturity, manufacturers have not yet
demonstrated the long‐term durability of fuel cells in stationary power generation applications.
The efficiency and capital cost characteristics of electrolyzer/fuel cell systems depend on the
type of technology utilized, the manner in which operators use the systems (lower load levels
generally imply higher efficiency but at the expense of capital costs), and the time frame
considered. Costs are relatively high at present, but analysts project them to fall (at an
uncertain pace) and efficiency to increase as economies of scale and manufacturing experience
take hold and continued technology development occurs (Lipman et al, 2005, p. 22).
One major handicap for this type of system at present is the low round‐trip efficiency, reported
at approximately 40percent by some sources (Gordes et al, 200). Based on current vendor
specifications outlined above, an electrolyzer‐compressed storage‐fuel cell system today would
have a round‐trip efficiency of 31‐35percent.
In addition, manufacturers have not demonstrated hydrogen electrolyzer/fuel cell technology in
the hundreds of MW scale of other energy storage systems, such as CAES and pumped
hydroelectric storage (Lipman et al, 2005 p. 23).
1.4.4 R&D Needs and Opportunities
Developers have focused the research and development needs for hydrogen energy storage
systems on the identified technology limitations. Recent studies suggest the potential cost
effectiveness of hydrogen energy storage for grid applications, providing a rationale to support
this research, development, and the demonstration of electrolyzer‐compressed storage‐fuel cell
systems (Schoenung, 2011; Steward et al, 2010 p. 17). Specifically, the industry will need
projects that develop and demonstrate MW‐scale systems to evaluate and assess actual costs,
durability, scale‐up, and performance as a distributed energy storage system for renewable
energy grid support.
Beyond 2020, focused research and development of regenerative fuel cells has the potential to
contribute to the advancement of grid‐scale energy storage (Nexight Group, 2010a, p. 37). A
variety of technology‐specific and crosscutting activities and initiatives could overcome the
current gaps and limitations of this technology, including:
66
• Improving thermal management in endothermic electrolysis reactions and in exothermic
fuel cell reactions to create a mid‐term market impact.
• Developing alkaline membranes, which do not require the use of precious metals, and
extending nano‐structured, thin‐film catalysts to electrolyzers to increase significantly
the potential for regenerative fuel cells to improve grid storage.
Projections from industry experts suggest significant improvements in cost effectiveness
(Nexight Group, 2010a):
• Current: $4,000/kW (alkaline and electrolysis and polymer fuel cell with separate
module for electrolysis and for fuel cells)
• 2015: $2,000/kW (alkaline and electrolysis and polymer fuel cell with one module for
electrolysis and regeneration); $800–$1,000/kW (solid oxide fuel cell with one module for
electrolysis and regeneration)
• 2020: $1,500/kW (for both types of fuel cells)
• 2030: $250/kW (for both types of fuel cells)
1.5 General Cost Data for Energy Storage Technologies
This section provides an overview of cost data obtained from existing energy storage projects.
Accurate forecasting total installed costs in $/kW and $/kWh is challenging given the large
ranges of projects and the fact that many technologies are still in the formative stages of
commercialization. Some of the data comes from projects funded by the United States
Department of Energy’s (DOE) Energy Storage Program, which experienced a significant
funding boost from the American Recovery and Reinvestment Act of 2009 (ARRA). ARRA
grants, however, may not be representative of future installed costs due to the additional
administrative expenses needed for these projects (despite the requirement that they be “shovel
ready” to be eligible for ARRA funds). Non‐ARRA procurements of energy storage systems
during the past two years indicate achievement of some reductions in installed costs. Totaling
the costs for the technology alone has been difficult because the data do not differentiate
between equipment procurement, engineering services, and construction costs. Meanwhile,
utilities in Maui, Hawaii; British Columbia, Canada; and elsewhere in the United States have
increased their battery capacity in response to assessments that more energy storage capacity
67
might have helped to avert the Fukushima nuclear crisis. As a result, these purchases for
remote applications have helped drive the early adopter market.2
Overall, the DOE noted in February, 2011 that many energy storage technologies are not
competitive in capital cost and/or life cycle cost for broad market penetration, although this
assessment does not include the potential benefits from the technologies that they currently
cannot monetize. The capital cost is at least $500/kWh in terms of energy or $2,500/kW in terms
of power (Department of Energy. 2011b, p.14). Pumped hydro storage with a low life cycle cost
(¢/kWh/cycle) may be an exception, but site selection challenges, large initial investment, long
construction time, and environmental concerns may limit the deployment, as discussed
previously. The same factors may also hinder deployment of underground compressed air
storage. Meanwhile, the CPUC‐sponsored cost effectiveness study for the Self Generation
Incentive Program (SGIP) reported distributed energy storage unfavorably in terms of societal
total resource cost (STRC) (ITRON. 2011).
In recent procurement efforts, utilities and research institutions have noted a disparity between
forecasted installed costs for various energy systems from vendors’ public presentation and
turnkey pricing that is generally not publicly disclosed. KEMA, an international consulting
firm to utilities that is active in researching the technical aspects of energy storage performance
testing,3 released to industry stakeholders in June 2011 a proprietary energy storage
performance modeling tool called ES‐Select.4 KEMA model’s input data for energy storage
technologies represent a snapshot figure of costs for various technologies that are likely to be
realized over the next one to four years. The figures do not account for the duration of supplied
energy and that embed multiple costs that may not be applicable for a particular application,
location, or vendor. However, the data provide policy makers and stakeholders with a
reference point for determining current costs for specific technologies. As an example of some
of the cost data, KEMA found that high energy Lithium Ion batteries had a low scenario total
ownership cost (with warranty) of $3107/kW and a high case of $3728/kW. Sodium Sulfur
batteries had a low case of $3815/kW and a high case of $4783/kW. Below ground compressed
air ranged from $899 to $1169 per kW, while flywheels were $1454 in a low scenario and $2631
2 See “US studies Fukushima disaster for safety lessons,” TerraDaily, April 2, 2011. Available at:
http://www.terradaily.com/reports/US_studies_Fukushima_disaster_for_safety_lessons_999.ht
ml (created April 2, 2011, last accessed July 8, 2011).
3 For more information about the model tool, see KEMA, Press Release, February 23, 2011.
Available at: http://www.kema.com/news/pressroom/press-releases/2011/KEMA-opens-energy-storage-
performance-testing-lab.aspx (created February 23, 2011, last accessed July 8, 2011).
4 To view the source of these data, please access the KEMA website at the following URL:
http://www.kema.com/ES‐Select/ (created June 2011, last accessed July 8, 2011).
68
in a high scenario per kW. Pumped hydro ranged from $1500 to $4500 per kW and advanced
lead acid batteries were $2223 to $3172 per kW for a low to high scenario.
Additional cost data will likely be forthcoming from other studies on these technologies, and
KEMA may further refine these numbers in light of industry feedback. As an important caveat
for policy makers projecting grid needs for 2020, these figures represent current prices and do
not indicate potential and likely price decreases as technological breakthroughs, continued
investment, and economies of scale reduce costs.
69
CHAPTER 2: Regulatory and Policy Framework
Introduction:
The development and deployment of electric energy storage technologies in California will
depend, in large part, upon the status of a multitude of policies and regulations, particularly at
the state and federal level. These policies can act to encourage and enable private and public
utilities, electricity service providers, and end users to include energy storage as a key asset in
meeting the state’s ambitious renewable energy targets and greenhouse gas emission reduction
goals.
A review of the relevant legislative, regulatory, and market measures, as well as of the costs,
benefits, and challenges identified in recent studies relating to energy storage, provides a basis
for understanding the uses and applications for energy storage, as well as possible regulations
and market design changes that might allow for fair valuation and cost recovery on investments
in energy storage technologies. This section provides an overview of California’s clean energy
agenda as it relates to energy storage. In particular, it focuses on the statutory and regulatory
framework, and on federal and state agency activities affecting the development and
deployment of energy storage technologies. Specifically, this section will review: state planning
and policy documents; statewide legislation that relates to or affects energy storage; regulatory
and tariff proceedings underway at the CPUC, California Air Resources Board (ARB), Energy
Commission, California ISO, and the FERC; utility treatment of storage assets. It will also
discuss actions taken outside of California for comparative and illustrative purposes.
2.1 Energy Storage Policy and Regulatory Landscape
2.1.1 California’s Clean Energy Policy Push
California’s ambitious clean energy agenda places it at the forefront of efforts nationally and
globally to increase efficiency, reduce greenhouse gas emissions, and shift to a cleaner and more
sustainable energy future. The changes taking place to implement this vision are wide ranging
and encompass legislative and regulatory action, private sector and industry initiatives, public‐
private partnerships, and inter‐agency coordination and collaboration. As the state progresses
towards its 2020 targets for substantially increased renewable energy resources and reduced
greenhouse gas emissions, the potential for energy storage to assist in the integration of
renewable resources and the maintenance of a reliable and efficient electric grid takes on great
significance.
In 2003, the Energy Commission and CPUC, together with the now defunct California Power
Authority, for the first time coordinated efforts to articulate a unified approach for California to
meet its electricity needs. This effort resulted in the first Energy Action Plan (EAP), in which the
70
agencies focused on actions to eliminate outage and electricity price spikes, reward energy
efficiency and conservation, and ensure adequate, reliable, and reasonably‐priced energy
supplies. Most significant may have been the articulation of a preferential “loading order” of
cost‐effective resources to meet the state’s electricity needs. The loading order, which remains
in place today, prioritizes first energy efficiency and conservation, second renewable energy
and distributed generation, and third clean, fossil‐fuel, central station generation (Consumer
Power and Conservation Financing Authority. 2003, p. 4). The loading order also highlights the
importance of improving California’s electricity transmission and distribution infrastructure to
support the growth of demand centers and interconnection of new resources.
In 2005, the Energy Commission and CPUC produced a second Energy Action Plan, focused on
the coordinated implementation of specific action areas, consistent with the loading order
articulated and expanded to include goals for reducing California’s contribution to climate
change. The Energy Action Plan II described key actions to maximize energy efficiency
measures, implement demand response programs, and meet an accelerated Renewables
Portfolio Standard (RPS) of 17 percent renewable by 2010 and 33 percent by 2020. (California
Energy Commission. 2005, p.2).5 The Plan also emphasized the need for capital investments to
support resource adequacy, reliability, and grid infrastructure, as well as the need to encourage
research, development, and demonstration of innovative energy technologies, including energy
storage (Ibid. p.12).
The baseline policies set forth in the 2003 and 2005 EAPs received an unprecedented push
forward with the passage of AB 32, the California Global Warming Solutions Act of 2006. AB
32’s directive for the Air Resources Board (ARB) to adopt regulations to reduce statewide
greenhouse gas emissions to 1990 levels by 2020, combined with the April 12, 2011 enactment of
California Renewable Energy Resources Act (Senate Bill 2, Simitian) that statutorily mandates
the 33 percent renewable electricity generation goal for 2020,6 together have created a need for
both technological and regulatory solutions that will help California attain its clean energy
goals. The Energy Commission’s 2007 Integrated Energy Policy Report (IEPR) and the
Commission and CPUC’s joint 2008 Energy Action Plan Update build upon the framework set out
by AB 32, identifying nine major action areas for meeting California’s energy needs in a carbon
constrained world. Energy storage technologies are relevant to several of these action areas,
5 Prior to the April 12, 2011 enactment of the California Renewable Energy Resources Act
(Senate Bill 2 Simitian) requiring 33 percent renewable electricity generation by 2020, several
actions laid the foundation for the 33 percent requirement. Senate Bill 1078 (Sher, Chapter 51,
Statutes of 2002) introduced the initial RPS, setting a goal of 20 percent by 2017. The 2003
Energy Action Plan proposed accelerating the goal to 20 percent by 2010. And Senate Bill 1250
(Perata, Chapter 512, Statutes of 2006) accelerated by statute the goal to 20 percent by 2010.
6 Senate Bill 2 (Simitian, Chapter 1, Statutes of 2011).
71
including: renewable energy; electricity adequacy, reliability, and infrastructure; climate
change; and research, development, and demonstration (California Energy Commission. 2008,
p.5). Both the 2009 IEPR and 2011 IEPR Scoping Order identified energy storage as an important
subtopic in integrating preferred resources into the electricity system. In addition, Governor
Jerry Brown’s Clean Energy Plan notes the value of energy storage for renewables integration
and its economic and job creation potential if utilities were to procure storage equivalent to
5percent of their peak load demand.7
The 2009 IEPR (Integrated Energy Policy Report) gave extended treatment to energy storage. It
listed existing and emerging storage technologies as a key component to integrating large
amounts of renewable generation into the electricity grid. Specifically, the report identified the
value of energy storage technologies for generation, transmission and distribution, and end‐use
applications. At the generation level, the Energy Commission noted energy storage
technologies’ ability to provide ancillary and renewable integration services, such as grid
frequency regulation. At the transmission and distribution levels, the report noted the values of
load shifting, transmission congestion relief, increased reliability, and capital deferral. And at
the end user level, the Energy Commission noted the value for commercial and industrial users
to obtain peak shaving, backup power, and increased reliability. The Energy Commission
identified these values in the context of the loading order preferences for meeting energy
supply needs, the requirements set by AB 32 and California’s RPS, and the Energy
Commission’s ten year electricity demand forecast (California Energy Commission. 2009, pp. 5,
87, 193‐06).
The 2009 IEPR also compared the value offered by electric energy storage to counteract effects
on local system reliability due to the operating characteristics of wind and solar power to that
provided by gas‐fired generation. It recognized that “some of the firming services provided by
gas‐fired generation will need to come from existing and emerging energy storage technologies
that allow generators and transmission operators to fill the gap between the time of generation
(off‐peak) and the time of need (on‐peak) for intermittent renewable energy” (California Energy
Commission. 2009, p. 193). It found that storage can respond to intermittency and fast drop off
and pick up rates more quickly than natural gas generation, potentially better addressing the
severe ramping rates associated with wind and photovoltaic solar and reducing the overall
amount of energy needed to balance the system and maintain reliability (although energy
storage devices lose some energy through conversion). The Commission recommended that
California pursue the following applications for energy storage:
7 To review Governor Brown’s proposal, please visit: http://www.jerrybrown.org/Clean_Energy
(created June 15, 2010, last accessed July 7, 2011)
72
• Ancillary services that otherwise would be provided by conventional fired plants to
integrate large amounts of renewables;
• Utility scale storage to avoid cutting wind production in response to transmission
limits;
• Load shifting / shifting renewable production to times of higher demand, which also
helps to address surplus generation;
• Fast response storage for electricity system stability and to reduce frequency
response problems.
In 2010, the Energy Commission focused the IEPR Update on the American Recovery and
Reinvestment Act of 2009 (ARRA) stimulus funding for clean energy projects. Energy storage
appeared in the Commission’s discussion of California’s renewable electricity goals, smart grid
development, and as a beneficiary of the private‐public partnerships that leveraged ARRA
funding for research, development, and pilot projects. The 2010 IEPR Update emphasized the
role of government during the current phase of emerging clean technologies as “crucial” to
establish “policies that provide long‐term market signals, performance standards, and
incentives to encourage private investment.” (California Energy Commission. 2010a, p.11). The
role of California’s energy regulatory agencies in moving forward policies and regulatory
changes to support energy storage deployment is addressed in detail in Chapter 2 of this
report.
In this year’s IEPR, the Energy Commission will focus on electricity infrastructure needs based
on the level of preferred resources expected to be in place in the next decade. The 2011 IEPR
Scoping Order reiterates the potential value of energy storage discussed in previous IEPRs and
identifies storage as an important subtopic of this year’s report (California Energy Commission.
2011a, p.3). The expanded treatment that energy storage will receive in the 2011 IEPR is
consistent with the recent legislative mandate under AB 2514 for adoption of a more concrete
energy storage agenda for California, and with the increased role for storage systems implicated
by California’s RPS and greenhouse gas emission reduction goals. As stated by the state’s lead
energy and environmental agencies in their paper on California’s Clean Energy Future,
Collaborative Vision for 2020, “Energy storage will be a significant feature of the cleaner power
system. As a result of agency support for storage technology pilot studies, increased utility‐scale
and aggregated smaller‐scale storage technologies … will be available to help facilitate
integration of intermittent wind and solar renewables resources,” as well as to shift peak
demand to off‐peak hours and to provide ancillary services to grid operations (California
Environmental Protection Agency. 2006, p.7).
Efforts by federal agencies and legislators to include energy storage in plans for the nation’s
clean energy future reinforce the steps taken by California policymakers. In an extensive effort
73
to outline policies for the twenty‐first century electric grid, the President’s Executive Office and
National Science and Technology Council have stated that energy storage can help facilitate a
clean energy economy:
Energy storage systems have the potential to optimize the operating
capabilities of the grid and the use of renewable energy by storing power for
later use. Such systems can work more effectively in tandem with smart grid
technology, can augment demand response resources to reduce and shift
peak energy loads, and can improve power quality and service reliability.
Significantly, utility‐sited energy storage can be used to avoid or delay
investments in new distribution capacity and substation equipment
(Executive Office of the President. 2011, p.15).
The American Recovery and Reinvestment Act invested $685 million in smart grid and energy
storage demonstration projects, laying a foundation for utilities and industry to partner with the
government and enable the real‐world testing and development necessary for energy storage to
gain ground. Other important actions at the federal level that could affect storage deployment
in California include Oregon Senator Ron Wyden’s plans to reintroduce legislation that would
create an investment tax credit specific to grid‐connected energy storage properties, the
STORAGE 2011 Act, and FERC’s regulatory processes to evaluate the treatment of energy
storage in the wholesale energy markets, discussed below in Section 2.1.3.1.
2.1.2 Legislative Mandates: AB 2514, AB 32, and California’s RPS
Several legislative mandates establish the basic framework upon which policies and regulations
relating to energy storage in California will be built. The California legislature passed AB 2514,
the foremost statute relating to utility procurement of energy storage systems, in September
2010. AB 2514 makes numerous findings regarding the value of energy storage and the
significant barriers that hinder its timely implementation. The broader mandates for the
reduction of greenhouse gas emissions under AB 32, and for increased renewable energy
generation under the Renewables Portfolio Standard, create incentives for the state to develop
and effectuate a vision for energy storage by 2020 and to evaluate targets in accordance with AB
2514. In addition are several laws that support investments in renewable energy and smart grid
systems, which may rely on or benefit from energy storage technologies. These include the Self‐
Generation Incentive Program (SB 412), Smart Grid Systems (SB 17), and the California Solar
Initiative (SB 1).
Under AB 2514, the legislature finds that expanding the use of energy storage systems can assist
load serving entities (LSEs) “in integrating increased amounts of renewable energy resources
into the electrical transmission and distribution grid in a manner that minimizes emissions of
74
greenhouse gases,” and can help “optimize the use” of the variable, intermittent, and off peak
electrical generation from wind and solar energy. It can reduce costs to ratepayers based on
deferred or avoided new peaker plants and transmission‐distribution upgrades and expansions,
and will reduce the use of fossil‐fuel generated electricity at peak, reducing carbon and criteria
pollutant emissions (AB 2514 [Skinner, Chapter 469, Statutes 2010], Sec. 1). AB 2514 identifies
certain significant barriers to obtaining these benefits, including inadequate evaluation of
energy storage as part of the state’s long‐term electricity resource planning, lack of recognition
of advancements in storage technologies, and “inadequate statutory and regulatory support”
(Ibid., Sec. 3(f)). The statute directs the CPUC to open a proceeding to determine appropriate
targets, if any, for LSEs to procure energy storage systems by December 31, 2015 and December
31, 2020 and for publicly owned utilities (POUs) to do the same on a slightly deferred schedule,
by December 31, 2016 and December 31, 2021. It also provides guidance to the CPUC regarding
the information the agency should consider in the rulemaking, the attributes and electricity
dispatch characteristics of energy storage technologies within the meaning of the statute, and
the need to weigh technological feasibility and cost effectiveness for any procurement targets
(Ibid. Sec. 2).
If the CPUC determines it appropriate to set targets for energy storage procurement, it must
adopt those targets for non‐POU electric corporations by October 1, 2013 and reevaluate them
no less than once every three years. The CPUC proceeding and stakeholder input is reviewed in
more detail below, and though it constitutes the most direct driving force behind current
discussions on energy storage, California’s broader climate change policy forms an arguably
equally important driver.
The California Global Warming Solutions Act of 2006, AB 32, represents California’s landmark
legislation and dominant force behind the state’s recent clean energy push. Its findings include
that global warming poses a serious threat to the state’s economic, environmental, and public
well‐being, and that “investing in the development of innovative and pioneering technologies
will assist California in achieving the 2020 statewide limit on emissions of greenhouse gases…
and will provide an opportunity for the state to take a global economic and technologic
leadership role in reducing emissions of greenhouse gases” (Assembly Bill 32 [Núñez, Chapter
488, Statutes of 2006], Section 38501(e)). AB 32 directs the state Air Resources Board (ARB) to
adopt regulations that will return greenhouse gas emissions for California to their 1990 levels by
2020. The ARB must design the regulations based on maximum technological feasible and cost
effectiveness to minimize costs and maximize benefits for the state’s economy, energy
infrastructure, and environment (Ibid. Sec. 28560, 38501).
In December 2008, in accordance with AB 32, the ARB adopted a Scoping Plan that sets forth the
measures to be developed and implemented by January 2012. The Scoping Plan touches on
75
nearly every sector of the state economy, from energy efficiency, building and appliance
standards and transportation to development of a regional carbon market (California Air
Resources Board, 2008). Several areas of pending regulation will affect the role of energy
storage, including the RPS, carbon cap and trade, and the Million Solar Roofs/ distributed
generation program. The development of these regulations in many cases overlaps with
proceedings underway at the ARB and the CPUC. The following section on Institutional and
Regulatory Action provides a brief review of the scope and status of these proceedings and their
potential effect on policies for energy storage system development and deployment.
The Renewables Portfolio Standard (RPS) serves as yet another major driver behind expanded
deployment of energy storage in California. Under Senate Bill 107, California’s investor‐owned
utilities (IOUs) needed to procure 20 percent of their electricity from renewable resources by
2010 (Senate Bill 107 [Simitian, Chapter 464, Statutes of 2006]). With the recent passage of the
California Renewable Energy Resources Act (Senate Bill 2 [Simitian, Chapter 1, Statutes of
2011]), California increased the percentage to 33 percent by December 31, 2020, codifying
regulations issued by ARB in response to an executive order from Governor Schwarzenegger.
The new law requires the CPUC to establish procurement targets sufficient to ensure that the 33
percent target is met and retained in subsequent years, and that an interim target of 25 percent
of retail electricity sales from qualifying renewable resources is met by December 31, 2016
(Senate Bill 2). One notable change is that the CPUC will no longer calculate procurement
targets for renewable resources in relation to a market price referent (MPR), but instead will
establish a cost limitation on how much each electrical corporation can spend on renewable
energy procurement expenditures. Exceptions to the renewable procurement standard will exist
if a retail seller cannot meet its obligations within the cost limitations set by the CPUC. The
renewable procurement requirements will apply equally to publically owned utilities, although
implementation will be the responsibility of the POU governing board, not the CPUC. A
determination by the CPUC of the load carrying capacity of wind and solar energy resources
will be used to evaluate the contribution of wind and solar toward meeting resource adequacy
requirements. The law also requires the Energy Commission to adopt regulations for the
enforcement of the RPS and for the design and implementation of an accounting system to
verify compliance. Important to the prospects for energy storage deployment, the bill requires
the California ISO to integrate and interconnect renewable energy resources to the grid in “the
most efficient means possible with the goal of minimizing the impact and cost of new
transmission facilities needed to meet both reliability standards” and the RPS targets (Senate
Bill 2, Simitian) Sec. 399.26(b) (1)).
Even prior to the passage of SB 2, California’s energy agencies have been working in
preparation of the 33 percent RPS. In the 2005 Energy Action Plan II, the CPUC and Energy
76
Commission identify the steps necessary to achieve 33 percent by 2020 (California Energy
Commission. 2005, pp. 5‐6). The AB 32 Scoping Plan similarly accounts for the 33 percent target
in its analysis of greenhouse gas emission reductions by 2020. And the ARB published its
proposed RES regulations in June 2010, which are intended to complement and build upon the
RPS program, but not to alter or replace it (California Air Resources Board, 2010a).
By 2010, the three large investor owned utilities (IOUs) had not met the 20 percent RPS target,
but reported servicing about 18 percent of their retail electricity sales from renewable energy.
The CPUC’s RPS Quarterly Report for Q4 2009 anticipated that the 20 percent RPS goal would be
met in the 2013‐14 timeframe (California Public Utilities Commission, 2009a, p.4). Because
numerous contracts for new renewable resource capacity were added in 2010, the CPUC more
recently reported that the IOUs were forecasted to achieve 20 percent RPS in the 2011‐12
timeframe (California Public Utilities Commission, 2011a, p.5). California ISO has expressed
confidence that California’s electricity system is equipped to integrate the amount of renewable
resources expected to come online in the next 1‐2 years. However, it has identified the need to
evaluate changes to the wholesale market design to accommodate the growing amount of
renewables as the state advances towards the 33 percent RPS (California Independent System
Operator, 2010a, p.3). The following section reviews the activities of California ISO and FERC,
as well as the CPUC and ARB, which are directed at facilitating the integration of renewable
generation and improving the efficiency and reliability of the grid.
2.1.3 Institutional and Regulatory Action: Regulatory Proceedings and Rate Cases
There are numerous regulatory activities underway at the state and federal level that will
impact the market for electric energy storage. These activities include tariff and market rule
reviews at California ISO and FERC, multiple proceedings underway at the CPUC, and ARB
rulemakings to implement the scoping plan under AB 32. In addition to agency activities,
investor‐owned utility rate case applications, smart grid deployment plans, and permanent load
shifting proposals as presented to the CPUC shed light on how the large IOUs intend to include
energy storage in their portfolios based on the existing regulatory framework.
The remainder of this section reviews the status of the regulatory proceedings named above,
briefly summarizing stakeholder input when appropriate, and highlights several utility
proposals for the inclusion of energy storage assets in the procurement and smart grid planning
processes.
2.1.3.1 California ISO and FERC: Wholesale Energy and Ancillary Markets
As part of the market design review for 33 percent renewables integration, California ISO has
engaged with stakeholders in a two‐phase Renewable Integration: Market and Product Review
process to identify and define new market products and improvements needed to support
77
renewable integration. Phase 1 of the market and product review process will implement near‐
term changes within 1‐2 years that are anticipated to provide benefits in any future renewable
resource scenario (California Independent System Operator, 2011b, p.1). Phase 2 will develop
longer‐term solutions, focusing on “the most effective and robust solutions to the new
challenges and requirements rather than the most expedient or easiest to implement,” and will
take “a holistic view” that accounts for the interactions between distinct market elements
(California Independent System Operator, 2011c). Some of the proposed market design
elements under consideration in Phase 2 include pay for performance regulation in accordance
with the FERC Notice of Proposed Rulemaking (NOPR) discussed below, a load‐following
reserve requirement, a methodology to allocate the costs of integrating increased amounts of
variable renewable energy resources, and modifications to intra‐day market settlements.
In conjunction with the market and product review process to support renewables integration,
California ISO has initiated changes to facilitate the provision of energy storage resources for
ancillary service8 needs. FERC simultaneously has proceedings underway to remove
unnecessary barriers to transmission service and wholesale markets for variable energy
sources9 and technologies that may facilitate their integration, the results of which are likely to
affect California ISO’s tariffs and ultimately the deployment of energy storage technologies in
California.
In 2007 and 2008, the FERC issued Orders 890 and 719, which together required regional system
and transmission operators to allow demand response technologies to compete in ancillary
services if they met technical requirements and could bid at or below market price; and for non‐
generation technologies, including energy storage, to be evaluated on a comparable basis with
generation sources to meet reliability requirements, ancillary services needs, and plans for grid
expansion (Federal Energy Regulatory Commission, 2007; Ibid., 2008a.). In response, California
ISO initiated a proceeding to consider changes to operating and technical requirements in the
ancillary services market to facilitate the participation of non‐generator resources. In March
2010, the California ISO Board approved the changes discussed below, and in May 2010 they
published draft revised tariff language for non generator resources (See California Independent
System Operator, 2010c; Ibid. 2010b, Sec. 8.2.3.5(c)).
8 According to the California ISO, for the purpose of the relevant proposed tariff language,
ancillary services are: (i) Regulation Up and Regulation Down, (ii) Spinning Reserve, (iii) Non‐
Spinning Reserve, (iv) Voltage Support, and (v) Black Start capability. See California
Independent System Operator. 2010b, p.1.
9 Variable energy resource is the term used by FERC to describe renewable energy resources
that have variable or intermittent production characteristics.
78
The modifications to the ancillary services market include: (1) a technology‐neutral definition of
ancillary services resource as ʺany resource meeting the operating characteristics and
technological requirements for each ancillary service,ʺ which definition will align with the
Western Electricity Coordinating Council’s pending change to the definition of spinning reserve
such that it is no longer limited to generation resources (2) a reduced continuous energy
requirement for ancillary services from 2 hours to 60 minutes for day ahead regulation
up/down, and to 30 minutes for real time regulation up/down, spinning reserve, and non‐
spinning reserve; (3) an altered protocol for measurement of the continuous energy requirement
that will begin at the moment the resource reaches its award capacity, not at the end of the
minimum 10‐minute ramp time; and (4) a reduced minimum rated capacity requirement from
1MW to 500KW (California Independent System Operator, 2010c, pp. 3, 10‐11) ;10. These
changes lay the groundwork for increased participation of energy storage and other non‐
generation resources in the ancillary services market and open the door to a cost‐effectiveness
methodology that incorporates ISO tariffs for unbundled ancillary services applications. For
example, the proposed tariffs allow Spinning Reserve and Regulation Up from the same
resource to be provided as separate ancillary services, provided that their sum “is not greater
than the maximum Ramp Rate of the resource (MW/minute) times ten (10)” (California
Independent System Operator, 2010b).
However, these changes still limit participation in the real time market or day ahead market to
those storage resources which can discharge or charge for more than 29 minutes or 59 minutes,
respectively. Many limited energy storage resource (LESR) technologies, such as flywheels,
some batteries, and some demand response technologies, have the ability to provide valuable
regulation services to the grid because of their fast ramping, but not in the required time scale.
To address this gap, California ISO proposed a Regulation Energy Management (REM)
methodology, which the California ISO Board of Governors approved on February 3, 2011
(California Independent System Operator Board of Governors, 2011d). The REM option
complements the changes to the ancillary services market because it allows a LESR that
otherwise would be regulated out of participation in the day ahead market to participate by
purchasing or selling energy in real‐time (California Independent System Operator, 2011e,
p.3).11 By opting for REM, a LESR allows California ISO to manage its state of charge, enabling
10 As proposed by WECC, and pending approval by FERC, the definition of spinning reserves
would change from “unloaded generation which is synchronized and ready to serve additional
demand” to a resource that “immediately and automatically responds proportionately to
frequency deviations.” See California Independent System Operator. 2010c, p.3.
11 Only limited energy resources that require a real‐time energy offset can qualify for REM;
these may include flywheels, some batteries, and some demand response resources.
79
it to meet the one hour duration requirement for regulation purchased in the day ahead market.
Based on modeling exercises of REM functionality under realistic system conditions and
renewable integration studies under 20 percent RPS, California ISO concluded that REM will
help address future system requirements by facilitating regulation by limited energy resources,
even if the specific capacity required under 33 percent RPS is yet unknown (Ibid., pp. 5, 7).
Settlement to LESRs for provision of regulation services through REM will be in the form of a
capacity payment, the same as for traditional regulation in the day ahead market (California
Independent System Operator, 2011e, p.13).12 Some stakeholders advocated for an additional
“mileage payment” to regulation resources, which would account for the sum total of all
injections to or withdrawals from the system in response to ISO regulation signals. California
ISO recognizes the potential merit of such payment and has included exploration of “pay for
performance” mileage payments in the scope of phase 2 of the Integration of Renewables, Market
and Product Review. A mandated “mileage” component for frequency regulation payments may
stem from recent activities at FERC, the results of which will determine further changes to
California ISO’s actions to facilitate increased participation of non‐generator resources in the
ancillary services markets.
On February 17, 2011, FERC issued a Notice of Proposed Rulemaking (NOPR) for Frequency
Regulation Compensation in the Wholesale Power Markets, which would affect California ISO’s
tariff for regulation services. In the NOPR, FERC found that current compensation methods in
the ISO and RTO markets may not acknowledge the benefits that fast‐ramping resources, such
as flywheels and battery systems, bring to frequency regulation over the capabilities of
traditional generators, resulting in tariffs that may not compensate regulating resources
accurately to reflect the services performed (Federal Energy Regulatory Commission, 2011,
pp.1‐3, 15‐16). To remedy any discriminatory compensation practices, FERC proposed that
frequency regulation compensation in the ISO/RTOs reflect a two‐part payment structure
derived from current best practices. The two parts are (1) a uniform capacity payment for the
amount of capacity held in reserve to provide frequency regulation service, which also would
include the marginal opportunity costs of the regulating resource, and (2) a performance or
“mileage payment” based upon the absolute MW up and down that the resource provides in
response to control signals, which also would include a measure for the resource’s response
accuracy (Ibid., pp. 21‐22). FERC sought comments to its proposed payment structure,
including whether the performance payment should be set by the market or administratively
Conventional generation resources may continue to provide regulation services in the day‐
ahead market, but may not select REM. (California Independent System Operator, 2011e, p.12).
12 A capacity payment is calculated based on the capacity set aside by the resource to provide
the service requested.
80
and how the payments should incorporate accuracy. If approved, the incorporation of mileage
and accuracy payments into the traditional capacity‐based compensation structure for
regulation services would have the potential to increase revenues for fast‐response storage
developers and to increase their market participation.
FERC also issued a Request for Comments Regarding Rates, Accounting, and Financial
Reporting for New Electric Storage Technologies to better understand the various ways in
which electric storage can be used, where those uses fall within established jurisdictional
boundaries, and the appropriate rate treatment, classification, and reporting requirements for
those uses (Federal Energy Regulatory Commission, 2010). FERC’s leaders have been
particularly interested in the development of rate policies that accommodate and value the
flexibility of storage technologies. In their Request for Comments, FERC staff outlined the
agency’s initial position on the uses of storage and highlighted a number of questions that they
seek to answer. These include: When can storage be classified and compensated as a
transmission asset? When, if ever, can storage receive compensation both as transmission and
for enhancing the value of generation or for providing ancillary services? Should policy makers
consider developing stand‐alone contracts for storage services? And do policy makers need
new accounting and reporting requirements to facilitate cost of service ratemaking for storage
technologies? (Federal Energy Regulatory Commission, 2010, Sec. 36382).
Over 60 stakeholders filed responses to FERC’s Request for Comments, providing a wide range
of feedback. Amongst the California and nationally‐based responses, a large number supported
the possibility of energy storage cost recovery in transmission rates, so long as FERC establishes
criteria and safeguards to prevent unfair advantage or double recovery. 13 For example, storage
serving transmission needs would be subject to system operator control but without the
requirement that the system operator become a market participant. As expressed by the
National Alliance for Advanced Technology Batteries, “concern about possible cross‐
subsidization, unfair competition and double recovery … are proper regulatory concerns.
However, a simplistic solution that forces storage into a single category is not sound policy.”14
Several of these parties emphasized the need for policy makers to support the emerging energy
storage market by creating new regulations that are not overly restrictive and that allow entities
13 See Federal Energy Regulatory Commission. Docket AD10‐13, Comments of: Southern
California Edison Co.; San Diego Gas & Electric; National Hydropower Association; National
Electrical Manufacturers Association; National Alliance for Advanced Technology Batteries;
Electric Power Supply Association; California Energy Storage Alliance; Electricity Storage
Association. Even those responses that oppose rate treatment for energy storage suggest
safeguards that would be necessary, should it be allowed at all. See ibid., Comments of
California Dep’t of Water Resources and State Water Project.
14 Ibid., Comments of The National Alliance for Advanced Technology Batteries, p. 5.
81
to use the diverse values and flexible functions of storage in a number of ways, both rate‐based
and market‐based.15
Numerous responses, however, disagreed, based on the complications that arise if transmission
operators participate in the energy market and because transmission applications identified for
storage devices are essentially the same as those of generation and market‐based demand
response and may be covered under those rates.16 California ISO staff cautioned FERC against
classifying an energy storage device providing a service like voltage support as a transmission
asset, because it blurs the line between transmission and generation. Instead, FERC should
limit the classification of energy storage as a transmission asset to the unique situations in
which the technology’s operation justifies the classification. But in this case, policy makers
should allow the technology to serve transmission functions only. California ISO noted that
storage devices competing directly with generation and demand response resources could
cause price distortion if policy makers allow cost recovery for services such as transmission
congestion and deferment of transmission upgrades.17
Regarding dual classification as a transmission asset and either a generation or ancillary service
enhancement, numerous responses suggested that FERC could develop rules to define the
means of participating in each market and to ensure cost recovery that does not
overcompensate or cross‐subsidize.18 Several comments among this group suggested that FERC
classify storage devices on a case‐by‐case basis depending on the intended use and capabilities,
15 Ibid., Comments of: National Electrical Manufacturers Association; National Alliance for
Advanced Technology Batteries; Electric Power Supply Association; California Energy Storage
Alliance; Project for Sustainable FERC Energy Policy, on behalf of several public interest and
environmental organizations.
16 Ibid., Comments of: Six California Cities; California ISO; Transmission Agency of North
America; Transmission Agency of Northern California; City of Santa Clara, California and the
M‐S‐R Public Power Agency; California Dept. of Water Resources and State Water Project.
17 Ibid., Comments of California ISO. See also Comments of: Transmission Agency of Northern
California; City of Santa Clara, California and the M‐S‐R Public Power Agency; California Dep’t
of Water Resources and State Water Project; American Public Power Association; Electricity
Consumers Resource Council (arguing that energy storage as both a transmission service and a
competitive energy service creates special preference discrimination for storage facilities and
opens the door to double recovery or double charging of customers for the same facilities).
18 Ibid., Comments of: Southern California Edison Co.; National Hydropower Association;
National Alliance for Advanced Technology Batteries; Electric Power Supply Association;
Electric Power Supply Association; California Energy Storage Alliance; Coalition to Advance
Renewable Energy Through Bulk Storage; Electricity Storage Association; Project for
Sustainable FERC Energy Policy, on behalf of several public interest and environmental
organizations.
82
subject to a defined criteria, and where “intended use and capabilities” included flexibility to
allow storage units to perform different functions at different times.19 But the California ISO
and others argued that allowing dual cost recovery would greatly increase the operational
complexity between transmission system and storage operators and create unacceptable risk
that the storage technology would not operate as directed because of competing economic
priorities, causing reliability and cost‐based recovery concerns,20 in addition to the
subsidization and double recovery concerns noted above. Several entities, regardless of their
position on allowing cost recovery through transmission rates, supported exploration of a
contract‐based approach to storage services, possibly using interstate natural gas storage facility
stand‐alone contract options as a model.21 Others cited the lack of any parallel between gas and
electric storage and the challenge of implementing stand alone contracts for storage.22
Stakeholders disagree over the need or benefit of creating new accounting or reporting
requirements for energy storage, as opposed to establishing additional subaccounts within
already‐existing facility functions.23 The M‐S‐R Public Power Agency, representing Modesto
Irrigation District, the City of Santa Clara, and the City of Redding, pointed out that limiting
any transmission rate recovery of storage technologies to single‐purpose facilities will reduce
the need for accounting changes. However, if FERC allowed multiple use storage projects to
recover costs through transmission rates, then accounts will need to track and ensure that costs
resulting from market functions are not charged to the transmission function.24 According to
the comments of the CPUC, FERC currently lacks sufficient information to make generalized
cost recovery and accounting and financial reporting determinations and should proceed on a
19 Ibid., Comments of: Coalition to Advance Renewable Energy Through Bulk Storage;
Southern California Edison; National Rural Electric Cooperative Association; Edison Electric
Institute.
20 Ibid., Comments of California ISO.
21 Ibid., Comments of: California ISO; Electric Power Supply Association; Edison Electric
Institute; California Energy Storage Alliance.
22 Ibid., Comments of: Southern California Edison; National Rural Electric Cooperative
Association.
23 Ibid. Compare Comments of: Six California Cities; American Public Power Association;
California Energy Storage Alliance (supporting establishment of a separate accounting
framework for energy storage) with Comments of: Southern California Edison Co.; San Diego
Gas & Electric; National Rural Electric Cooperative Association; Edison Electric Institute
(arguing storage can be accounted for under existing plant functions).
24 Ibid., Comments of: Transmission Agency of Northern California; City of Santa Clara,
California and the M‐S‐R‐ Public Power Agency.
83
case‐by‐case basis until policy makers develop a better understanding of the varied uses of
storage in the electric system.25
2.1.3.2 California ISO: Participating Intermittent Renewables Program
The Participating Intermittent Renewables Program (PIRP) represents another California ISO
program undergoing revisions as part of the Renewable Integration: Market and Product Review.
The PIRP allows intermittent resources, like wind and solar, to schedule energy in the forward
market without incurring imbalance charges when the delivered energy differs from the
scheduled amount. Instead, grid operators net deviations from ten‐minute settlement intervals
across a calendar month and settle them at a weighted‐average price (California Independent
System Operator, 2006, p.1). California ISO staff will file proposed changes to the PIRP,
stemming from the Renewable Integration stakeholder process, in the coming weeks. These
changes may include modification of intra‐hour scheduling to increase California ISO
operating flexibility, the measuring of deviations closer to real‐time, and adjusting the
settlement system for market participants (Miller, 2011, pp.21‐24). If the new PIRP implements
a policy of imbalance charges for deviating from forecasts, energy storage could play a valuable
role in firming and shaping participating wind and solar resources. Generators that couple
with onsite or co‐located storage have the potential to mitigate deviations by providing firm
power in accordance with the forecasted amount.
2.1.3.3 California ISO: Transmission Planning and Interconnection Procedures
Transmission planning, and the rules affecting transmission and interconnection cost and cost
recovery, may impact the cost and value of energy storage technologies. Energy storage can
serve to reduce transmission congestion, but transmission planning rules determine the
likelihood that utilities can recover costs for investing in these technologies. California ISO’s
Generator Interconnection Procedures (GIP), based on FERC’s conditional approval of the GIP
in December 2010, govern transmission upgrades and generator interconnection for both small
and large projects (California Independent System Operator, 2011f, p.7). One important policy
that California ISO leaders will consider in the near term is how to distinguish between network
upgrades that will receive full funding by transmission ratepayers versus network upgrades for
which the interconnection customer will bear non‐refundable cost responsibility. Where project
developers bear the cost responsibility or must pay an upfront share, transmission upgrade
deferrals, potentially facilitated by the implementation of energy storage technologies, may
become more attractive.
California ISO’s Economic Planning Study represents another aspect of transmission planning
that is pertinent to energy storage valuation. The study will identify congestion in the
25 Ibid., Comments of California Public Utilities Commission.
84
transmission system and compare the economic benefit of a proposed mitigation measure to the
costs, to determine whether the mitigation solution is cost effective (California Independent
System Operator, 2011g).
Since grid operators can place energy storage devices at points of congestion along the
transmission grid, these devices should receive consideration under California ISO’s study.
Moreover, the study should compare the cost effectiveness of transportable versus stationary
energy storage solutions for grid support, a distinction made by EPRI in its White Paper Primer
on Applications, Costs, and Benefits for energy storage technologies (Electric Power Research
Institute, 2010, Table ES‐1).
2.1.3.4 California Public Utilities Commission
The CPUC is engaged in multiple proceedings that relate to energy storage. Foremost is the
rulemaking process pursuant to AB 2514. The AB 2514 rulemaking will enable the “review and
possible adoption of viable, cost‐effective energy storage technologies because they may
provide a strategy for meeting the state’s long‐term clean energy goals while maintaining
system reliability” (California Public Utilities Commission, 2010a, pp. 4‐5). Its purposes
include: (1) review and establish, if appropriate, opportunities for energy storage development
and deployment in California; (2) reduce barriers to such development and deployment; (3)
review and weigh the costs and benefits of such development and deployment; and (4) establish
methodologies that address how those costs and benefits should be distributed. Thus, the AB
2514 rulemaking provides the CPUC with an opportunity to “continue its rational
implementation of advanced sustainable energy technologies and the integration of intermittent
resources in our electricity grid.” (Ibid. p.1).
The CPUC released the scoping memo for the AB 2514 rulemaking, dividing the process into
two distinct phases, to be resolved prior to any determination on procurement targets. Phase 1
will establish policies and guidelines for energy storage in California, based on the energy
storage technologies and applications currently in use, the policies needed to encourage energy
storage for environmental and grid related objectives, barriers to energy storage adoption, and
the relationship between energy storage and other state policy priorities and mechanisms for
meeting state energy goals. Phase 2 will result in a cost‐benefit analysis to measure the value of
energy storage attributes and applications, considering costs, system and other benefits, and
allocation of costs and benefits amongst retail end‐use customers (California Public Utilities
Commission, 2011b).
Based on initial comments submitted by stakeholders, parties generally support the opportunity
to consider these issues in a rulemaking. Opinions were varied as to the readiness of storage
technologies for cost‐effective implementation, the appropriate methodology for evaluating and
85
monetizing the costs and benefits of storage, and whether the scope of the proceeding should
extend to issues such as dynamic pricing and rate recovery. Most comments cited the need for
the CPUC to create a valuation framework for energy storage technologies.26 Several
stakeholders have recommended development of a framework that would enable cost‐benefit
comparisons of storage with different types of resources, as well as with other load shifting and
emissions reducing strategies.27 The California Energy Storage Alliance (CESA) emphasized the
importance of a valuation framework that analyzes energy storage technologies at the
application or use case level, to account for the differences in system size, duration of capacity,
lifecycle and response time. Comments by a number of other stakeholders affirmed the
soundness of an application‐specific approach28 . Comments also encouraged an approach that
quantifies the ancillary and environmental quality benefits of energy storage.29 Other
stakeholders have emphasized that any valuation framework must protect against any “double‐
counting” of benefits due to overlapping value categories.30
Stakeholders have expressed an interest in how this CPUC proceeding will impact other CPUC
and state agency proceedings. A few stakeholders suggested that this proceeding serve as an
“umbrella” to coordinate development of energy storage policy across existing and future
proceedings. As the AB 2514 proceeding moves forward, the CPUC should provide a
transparent analytic framework that outlines its proposed valuation methodology to allow
stakeholders to provide responsive feedback and to offer clarity to participants in parallel
proceedings regarding the reach of the AB 2514 rulemaking.
Other areas under CPUC jurisdiction that may affect or be affected by the outcomes of the AB
2514 proceeding include distributed generation and the Self‐Generation Incentive Program
(SGIP), smart grid, permanent load shifting, demand response, dynamic pricing, and continued
RPS compliance through the LTPP process.
2.1.3.4.1 Smart Grid
26 See California Public Utilities Commission, Docket R10‐12‐007,
http://docs.cpuc.ca.gov/published/proceedings/R1012007.htm. Comments of: A123 Systems;
Beacon Power Corp.; Brookfield Renewable Power Inc.; California ISO; Calpine Corp.;
California Energy Storage Alliance; Environmental Defense Fund; Western Power Trading
Forum.
27 Ibid., Comments of: California ISO; Calpine Corp.; Division of Ratepayer Advocates; Western
Power Trading Forum.
28 Ibid., Comments of California Energy Storage Alliance ; Southern California Edison Co.; San
Diego Gas & Electric; Consumer Federation of California.
29 Ibid., Comments of Environmental Defense Fund.
30 Ibid., Comments of Calpine Corp.
86
Since 2008, the CPUC has been engaged in a rulemaking proceeding to develop Smart Grid
policies and protocols to guide the investor‐owned utilities in efforts to modernize the electric
grid. Federal and state law help direct the CPUC’s activities in this arena. California’s Senate
Bill 17 defined a broad set of characteristics pertaining to a Smart Grid and directed the CPUC
to establish requirements for a smart grid deployment plan that improves “overall efficiency,
reliability, and cost‐effectiveness of electrical system operations, planning and maintenance”
(Senate Bill 17 [Padilla, Chapter 327, Statutes of 2009], Sec. 8360, 8362). Deployment plans must
be consistent with federal standards, which require electric utilities to consider investments in
qualified smart grid measures prior to investing in “nonadvanced grid technologies” and
require state public utility commissions to consider authorizing utilities to obtain ratepayer
recovery for expenditures related to smart grid system deployment (Energy Independence and
Security Act of 2007 [42 U.S.C. Sec. 17001, et seq.], Sec. 1307). In its decision providing guidance
to the IOUs for their deployment plans, the CPUC stated that a purpose of the Smart Grid is to
“enable the integration of higher levels of renewable energy, energy storage, and, eventually,
electric vehicles, at a lower cost to consumers” (California Public Utilities Commission, 2010b,
p.2). The IOU’s first Smart Grid Deployment Plans are due to the CPUC in July 2011 and may
provide insight into how the utilities intend to include or pursue energy storage as part of their
smart grid plans. 31 Additionally, eight California smart grid and energy storage demonstration
projects that received $186 million in stimulus funding through the DOE competitive process
are likely to provide critical data points for current assessments of energy storage performance
potential.
2.1.3.4.2 Self‐Generation Incentive Program
The CPUC’s Self‐Generation Incentive Program (SGIP) provides financial incentives to support
various distributed energy systems installed on the customer side of the meter. In 2008, the
CPUC issued a ruling that energy storage systems meeting certain technical parameters and
coupled with eligible wind and fuel cell technologies would be eligible to receive financial
incentives under the program. The decision outlined operating parameters for “qualified
advanced energy storage” technologies to be eligible and provided for incentive compensation
on a per KW basis (California Public Utilities Commission, 2008, pp. 1, 12‐13, 18). It also stated
that if technologies other than wind or fuel cell were to become eligible under the SGIP
program, then energy storage coupled with those systems would qualify as well (Ibid., p.7).
Thus, future changes to the SGIP qualifying technologies may expand the range of energy
storage devices that can benefit from the incentives. Currently, the SGIP program does not
cover solar photovoltaic technologies, which instead receive financial incentives through the
31 San Diego Gas & Electric (SDG&E) filed its Smart Grid Deployment Plan early, on June 6,
2011. Relevant aspects of SDG&E’s plan are discussed below in Section 2.1.3.6.
87
California Solar Initiative (CSI). The CSI, however, does not provide for separate treatment of
energy storage technologies that may be coupled with distributed generation photovoltaic.
Although legislation enacted in 2009 extended the SGIP program through January 2016, it did
not provide the CPUC with the ability to authorize collection of funds from the electric
corporations for the SGIP program beyond the original sunset date of December 31, 2011. AB
1150 (V. Manuel Pérez, Chapter 310, Statutes of 2011), signed by Governor Brown in September
2011, authorizes funding for SGIP through 2014 and opens the program to energy storage
systems. Changing the SGIP program to include energy storage will provide potentially
significant incentive support to customer‐side energy storage and distributed generation
applications that the ARB finds will achieve overall system reductions in greenhouse gas
emissions.
2.1.3.4.3 Permanent Load Shifting
A CPUC decision in 2009 ordered California’s IOUs to study the means of expanding
Permanent Load Shifting (PLS), defined as “when a customer moves energy usage from one
time period to another on an ongoing basis,” and to study the use of a standard offer program
for PLS (California Public Utilities Commission, 2009c). The resulting study, published in
December 2010, made a number of findings relating to PLS cost effectiveness, value
propositions to the end‐user, market opportunity, and best practices observed in various pilot
programs nationwide. The authors used a technology neutral definition for PLS based on the
impact to the electricity usage profile of routine shifting of energy usage from one time period
of the day to another to help meet peak loads and improve grid operations. The study included
electric energy storage, thermal energy storage, and industrial process shifting practices as
within the scope of the PLS definition. It emphasized the difficulty of creating “a single, simple,
technology neutral PLS program designʺ because of the differences between individual
technologies in terms of cost, commercial readiness, and effect on the electric system. Because
similar challenges exist in creating a uniform program for the deployment of energy storage
technologies, the methods and several conclusions of the PLS study may have useful
implications for energy storage (Energy and Environmental Economics, Inc., 2010).
In evaluating cost effectiveness of ratepayer funded PLS programs, the study used a Total
Resource Cost (TRC) framework, based on avoided cost, comparing the installed costs of a PLS
program with the lifecycle benefits obtained. 32 The results of this evaluation suggested that PLS
program design be divided into two parts: a program for mature technologies and a separate
program for emerging technologies (Ibid., p.109). The goal of the mature technology program
32 For a detailed explanation of the study methodology, assumptions and sensitivity analysis,
see Energy and Environmental Economics, Inc. and Strategen Consulting, 2010, Section 3.
88
would be to maximize the installed capacity of already cost‐effective technologies that would be
likely to pass the TRC test at present.33 The goal of the emerging technology track would be to
develop greater numbers of promising technologies that may be unlikely to pass the TRC test at
current costs.34 The design of the emerging program track would depend upon the CPUC’s
assessment of the usefulness of PLS to manage future load needs, such as for renewable
integration, energy efficiency, and deferral of new generation capacity – needs for which many
of the emerging technologies are especially well suited.
California programs such as the SGIP and CSI may serve as examples for a potential market
transformation program for emerging PLS technologies. The study presents a spectrum of
possible ways to structure standard offers for PLS program design. These fall along two
continuum – one representing incentive level to the customer, which could range from
ratepayer neutral to guaranteeing a certain internal rate of return, and the second representing
the PLS program structure, which could focus on promoting on‐peak energy shifting
performance or on pure capacity. The authors presented several possible incentive structures
and the benefits and pilot results of each, including a PLS‐specific tariff, a standard offer based
on performance ($/kWh shifted), a hybrid standard offer that includes a capacity payment, and
a pure capacity‐based standard offer. Finally, the authors suggested characteristics of a good
PLS program design based on technology data and best practices nationwide. These relate to
system design and baseline technical and economic feasibility study, functional performance
testing of the system once installed, and regular reporting of operational data (Energy and
Environmental Economics, Inc. and Strategen Consulting, 2010, pp.109‐11, 1123‐16).
2.1.3.4.4 Resource Adequacy
The CPUC’s Resource Adequacy (RA) program serves two purposes: (1) to ensure that
California ISO has access to sufficient resources to operate the grid safely and reliably and (2) to
promote the siting and construction of new resources needed for future reliability. All LSEs
within the CPUC’s jurisdiction, including the IOUs, energy service providers, and community
choice aggregators, are subject to the determined RA requirements and must file regular
compliance reports to demonstrate procurement of sufficient resources to meet their
obligations. Because electric power provides can use some energy storage technologies to
reduce peak power demand and to meet certain energy and capacity needs, the CPUC could
promulgate rules to allow certain categories of energy storage procurement to count toward a
LSE’s RA obligations. This option is available for pumped storage and some types of demand
33 The study found that well‐designed, medium and large thermal storage technologies, and
low‐cost process shifting and pre‐cooling, to be the most cost‐effective at present.
34 The study found that that the majority of battery storage systems providing PLS, and small‐
scale thermal solutions, did not pass the TRC at current operating costs.
89
response technologies, as described below. The CPUC and California ISO could also establish a
forward market for energy storage technologies by linking projected regulation and ancillary
service needs to capacity requirements for those resources. The LSEs would need to enter into
procurement contracts to ensure sufficient provision of regulation services, creating an
opportunity for energy storage and especially limited energy storage resources (LESRs) to
secure long‐term contracts for their services. Grid operators could shift excess regulation
capacity to participate in the spot market.
2.1.3.4.5 Demand Response
The CPUC oversees the investor‐owned utilities’ administration of their demand response
programs. Demand response allows end‐use electricity customers to reduce their electricity
usage during a given time period or shift that usage to a different time period, in response to a
price signal, financial incentive, or other predetermined signal. The demand response program
is important to energy storage policy in California in a few ways. First, grid operators can use
energy storage technologies in demand response, especially to shift usage to off‐peak hours
when more supply‐side resources are more cheaply available. Deployment of energy storage
devices on the customer side of the meter might enable increased participation in demand
response by customers whose electricity needs normally would not be compatible. As the
utilities shift to a smart grid and advanced metering, localized storage might make the smart
grid more flexible and responsive to generation‐ and demand‐side needs.
Demand response might also serve as an example for energy storage of a retail program that
can include participation in the wholesale market via the RA program. For example, utilities
are able to count the capacity of their dispatchable demand response toward the total required
capacity that they must maintain under the RA program (California Public Utilities
Commission, 2010c, p.11). To maximize the value of their demand response program, a utility
will try to design it so that it falls within the RA program’s “counting” scheme. Energy storage
policy could follow a similar trajectory so that utility investments in energy storage technologies
could count toward RA requirements to the extent that the energy storage system creates
avoided energy or capacity needs.
2.1.3.4.6 Alternative Fueled Vehicle Proceeding
In August 2009, the CPUC issued an OIR to Consider Alternative‐fueled Vehicle Tariffs,
Infrastructure and Policies, consistent with 2009 legislation directing the CPUC to “evaluate
policies to develop infrastructure sufficient to overcome any barriers to the widespread
deployment and use of plug‐in hybrid and electric vehicles” (California Public Utilities
Commission, 2009b, p.3). While the rulemaking is beyond the scope of this report, changes to
facilitate growth of the plug‐in hybrid and electric vehicle market and increase deployment of
90
electric vehicles are likely to have significant implications for the cost, development, and
deployment of energy storage, especially battery storage technologies. For example, the CPUC
recognizes that electric vehicle charging “could represent a new and substantial increase in
load” and may require changes to electric rate design in order to ensure that increased load
does not compromise grid stability or increase peak load (Ibid., p.8). Impacts on load during off‐
peak hours also may impact the value of energy storage applications for electric energy time‐
shifting and arbitrage. In addition, widespread deployment of electrified transportation has the
potential to serve a variety of grid flexibility needs. Controlled charging can provide
dispatchable demand and demand response at times of high variable generation output,
charging rates can be designed to allow vehicle provision of contingency reserve and frequency
regulation, and vehicle to grid discharge capability can serve as a source of distributed storage,
should it gain public acceptance (Denholm, 2010, pp. 44‐45).
2.1.3.4.7 Long Term Procurement Planning
Two aspects of the CPUC’s semi‐annual Long Term Procurement Planning (LTPP) process are
important to consider in relation to energy storage policy and deployment plans. One is that
the LTPP is the vehicle by which the CPUC evaluates the needs of the IOUs for procurement of
additional fossil‐fuel or renewable generation resources and establishes rules for rate recovery.
The investor‐owned utilities and California ISO are cooperating with the CPUC on planning
scenarios and production simulations to quantify the incremental need for flexible generation
resources in the LTTP process. The CPUC has posted the preliminary results of the IOUs and
California ISO, including estimated regulation and load‐following needs (California Public
Utilities Commission, 2011d). Stakeholders from the renewable energy and consumer and
environmental protection communities also filed testimony with the CPUC on the model
assumptions and planning scenarios.
Under the recently enacted legislation for 33 percent RPS by 2020, the general procurement plan
process must include a renewable energy procurement plan, which will be geared toward
compliance with the 2020 targets and the interim targets set by the statute. The statute also
directs the CPUC to adopt criteria by which to order and select the least‐cost and best‐fit
renewable energy resources to comply with the RPS on a total cost basis. This total cost basis
includes consideration of indirect costs associated with transmission investments and
operational needs for integrating eligible renewables (Senate Bill 2 [Simitian, Chapter 1, Statutes
of 2011] Sec. 399.13(a)(1), (4)). Policy makers should factor the availability of energy storage
technologies to mitigate some of these indirect costs associated with renewables integration into
the utilities’ procurement planning and determine an acceptable level of mitigation for
ratepayers.
91
A second aspect of the LTPP is its importance for considering and implementing the CPUC’s
policies around the California Energy Action Plan loading order. Many stakeholders in the
energy storage proceeding under AB 2514 and in comments on this report have emphasized
that inclusion of energy storage in the loading order in some manner will provide further
guidance to state agencies and utilities to factor storage into long term procurement planning.
The same considerations apply to including energy storage as a potential technology to satisfy
Resource Adequacy requirements.
2.1.3.5 California Air Resources Board
As with the CPUC, the Air Resources Board (ARB) is engaged in a number of regulatory
proceedings that will likely influence related policies on energy storage. As the lead agency
implementing AB 32, the ARB has special responsibility overseeing actions to reduce
greenhouse gas emissions statewide and across all sectors. The applicability of energy storage
technologies to reduce greenhouse emissions in a variety of ways, from peak shaving to deferral
and avoidance of new fossil fuel plant construction for peak generation, ancillary services, and
renewable integration, serves as an additional reason for policy makers to consider an energy
storage role.
2.1.3.5.1 ARB Cap and Trade Proposed Regulations
A carbon cap‐and‐trade program for California’s electricity sector will increase the operating
cost of fossil fuel plants that provide peak generation power, ancillary, and backup services,
potentially driving down the relative cost of alternative and complementary mechanisms,
including electric energy storage. As currently proposed, the regulation to Implement the
California Cap‐and‐Trade Program will connect to the Western Climate Initiative regional
trading program, will cover the major sources of greenhouse gas emissions in the state,
including refineries and power plants, industrial facilities, and transportation fuels, and will
depend upon an open allowance trading market (California Air Resources Board, 2010b, Part I,
Vol. I). The ARB is completing its rulemaking process on schedule for a January 1, 2012 cap‐
and‐trade program start date, although the agency has delayed emissions obligations for
covered entities until 2013.
2.1.3.5.2 ARB Zero Emission Vehicle Program
The ARB’s Zero Emission Vehicle (ZEV) regulations are part of its broader Low Emission
Vehicle (LEV) III regulations, which together focus on reduction of smog and greenhouse gases,
and on increasing penetration of battery electric, fuel cell, and plug‐in hybrid electric vehicles.
In 2008, the ARB adopted modifications to the ZEV regulations, triggering a review process to
account for changes in the vehicle market, and to speed commercialization of zero and near‐
zero greenhouse gas emission producing vehicles. The regulations currently under
92
development will focus on incubating zero and near‐zero emission plug‐in hybrid and electric
vehicle technology for large scale market penetration, thus complementing the CPUC’s
Alternative Fueled Vehicle proceeding, which relates to the regulation of ZEV charging stations
and rates. Depending on the scenarios that the ARB uses to determine vehicle fleet goals for
2050, and the technologies to be promoted in those vehicles, the new regulations have the
potential to jump‐start additional research and incubation of advanced battery storage
technologies.
Based on current Li‐Ion battery technology costs, the ARB predicts that market forces alone will
not be sufficient to drive penetration to the degree needed for long‐term carbon emission
reduction goals.35 The ARB does not expect Li‐Ion battery costs to become commercially
competitive for ZEVs within this or even the next production generation without government
incentives and/or tax credits (California Air Resources Board, 2009, p.15). In a 2009 study of the
vehicle emission reductions needed to meet California’s 2050 clean energy goals, the ARB noted
that it can take 30‐plus years for a new vehicle technology to capture a large portion of existing
vehicle fleets. For example, to achieve a goal for most vehicles on the road by 2050 to be ZEVs
would require immediate escalation of investments in low‐carbon vehicle alternatives now, to
ensure that the markets for those vehicles would develop between 2015 and 2020 (Ibid., p.10).
Thus, the ARB cites the need for three types of action to attain ZEV goals: a regulatory mandate
to provide some degree of certainty to investors, combined efforts of federal, state, and local
governments to adopt ZEV fueling infrastructure, and consumer incentives to purchase ZEVs.
The updated ZEV regulations are scheduled to be released in 2011.
2.1.3.6 Utility Investments in Energy Storage
Even prior to AB 2514, California’s large investor‐owned utilities have taken steps toward
researching the value of energy storage and exploring which applications and types of
technology would be cost effective to deploy. The value or potential value of energy storage for
certain applications, as seen by the IOUs, is apparent from the numerous pilot projects and
ARRA‐funded storage research described in Chapter 4, as well as from the treatment of energy
storage in 2010 rate case application testimony before the CPUC. The following information and
quotes, taken from testimony on behalf of the utility, demonstrate the utility perspective and
investments being made in storage technology.
2.1.3.6.1 Southern California Edison (SCE)
35 In a May 2010 workshop, ARB staff presented current Li‐Ion battery costs as hovering at
$1,000/kWh, possibly dropping to as low as $300/kWh with high production volume. ARB.
2010. ZEV Regulation Workshop. May 3, 2010.
http://www.arb.ca.gov/msprog/zevprog/2011zevreg/05_03_10pres.pdf.
93
SCE’s Project Development Division identified the need to study generation storage
technologies (such as batteries and compressed air storage) (Southern California Edison, 2010a,
p.10). Through rate case testimony, the utility provided several reasons for a planned 1.0 MW
sodium sulfur battery energy storage system on a site adjacent to a generating station, including
that the installation is needed “to enable the optimal loading of the station’s existing diesel
generators for the purpose of minimizing diesel engine air emissions” by providing energy
storage capacity during off‐ peak periods and energy supply capacity during on‐peak periods,
and to provide additional benefits such as “improved system voltage regulation, system
stability, and system reliability” (Southern California Edison, 2010b, pp.15‐16). In the area of
transmission and distribution technologies, SCE cited the use of energy storage systems to
manage the grid and mitigate outages and service interruptions, and “to avoid the economic
losses associated with catastrophic failures and wide‐area blackouts” (Southern California
Edison, 2010c, p.2). Similarly, the utility pointed to dynamic energy storage as providing “an
important capability to support the integration of intermittent renewable energy resources” by
peak‐load shifting and supplying power to the grid when renewable resource generation is
limited (Ibid., p.26 n.34). SCE intends to use distributed energy storage systems to stabilize
variable distributed generation solar and wind resources, to meet the CPUC’s policy goals
under SB 17 and the Smart Grid OIR, and to enable adoption of plug‐in electric vehicles.
Finally, with increased variable renewable energy resources being added to the supply mix,
SCE expects storage to “become a viable option to facilitate and enhance distribution grid
operations” (Ibid., pp.51, 58, 56).
2.1.3.6.2 San Diego Gas and Electric (S D G and E)
Noting the significant changes in the energy system caused by advances in clean energy
technologies, S D G and E consistently included energy storage as an important part of its new
energy portfolio. Importantly, S D G and E is investing in energy storage systems—distributed
as well as utility scale—as one component of its solution to the “intermittency issues created by
the variable output of renewable energy sources” (San Diego Gas and Electric, 2010b, p.12). The
utility considers distributed energy storage in particular “to be one of the key factors in
enabling the proliferation of renewable resources such as wind and solar” (San Diego Gas and
Electric, 2010c, p.68). In discussing the role of storage in its Smart Grid infrastructure and in
solving renewable variability issues through “management and discharge of stored energy in a
controlled and coordinated” manner, one S D G and E representative outlined the multiple
value streams of energy storage as including power quality, power leveling and regulation on
grids connected to renewable energy, peak load shifting and shaving, energy arbitrage,
ancillary services related to California ISO, and T and D capacity deferral (San Diego Gas and
Electric, 2010d, pp. 10, 21). Recent S D G and E rate case testimony also pointed to storage
94
systems as a key part of the future of its storm and outage management and its award‐winning
Sustainable Communities project, which encourages construction of green buildings (San Diego
Gas and Electric, 2010a, p.30). Further, rate case testimony highlights that S D G and E is
working with partners in the automotive sector to develop uses for energy storage systems that
benefit both the transportation and home energy sectors (San Diego Gas and Electric, 2010c, pp.
56, 71).36
S D G and E also has highlighted the role of energy storage in its Smart Grid Deployment Plan,
filed one month early with the CPUC on June 6, 2011. Some of the ways in which S D G and E
intends to use energy storage in its smart grid include: as distributed storage on circuits with a
high penetration of customer photovoltaic systems; at the substation level for areas with
multiple high‐photovoltaic circuits; and to integrate energy efficiency, demand response, and
generation into customer energy usage, including by enabling participation in existing and
future energy markets (San Diego Gas and Electric, 2011b). The utility will use other pilot
programs to test new or unproven technologies, including emerging storage technologies, for
functionality, interoperability, and security, and to identify the most cost‐effective ways to
deliver needed services. Another pilot program will test new distributed energy resources,
including energy storage, in a new community with residential and commercial sectors (Ibid.,
pp. 99, 229).
To support the penetration of intermittent power generation sources, S D G and E is supporting
installation of cost‐effective energy storage, either utility, customer, or third party owned, to
help resolve voltage regulation, voltage flicker, and intermittency of renewable resources (San
Diego Gas and Electric, 2011b, p.48). S D G and E also outlines numerous actions to support SB
17’s objective to accommodate all cost‐effective generation and energy storage options, such as
to enable smart grid communications along distribution lines connecting distributed energy
resources (DER) and energy storage providers with California ISO markets, and develop control
and communications systems to facilitate aggregation of energy storage and DER, including
“vehicle‐to‐grid” applications.
2.1.3.6.3 Pacific Gas and Electric (P G and E)
P G and E’s investments in grid‐scale energy storage projects, in particular its planned
developments of large pumped storage, CAES, and battery facilities, suggest that P G and E
recognizes the potential of various energy storage technologies and applications. P G and E has
focused significant resources on the development of pumped hydro facilities. If constructed,
the utility’s Mokelumne Pumped Storage Project (MPSP) would provide up to 1,200 MW of
36 K. Cordova, “Testimony on Customer Services—Information,” San Diego Gas & Electric, pp.
56, 71.
95
pumped storage capability by 2020. The Department of Energy is partially funding P G and E’s
proposed CAES facility, contributing $50 million to the project through ARRA (Recovery.gov,
“Grants‐Award Summary”). Though the project is still preliminary, P G and E began analysis
on suitable geologic formations for prospective sites early this year. Analysts estimate that the
project—a facility with 300 MW output capacity—will take five years to design, permit, and
build (Marshal, 2009a). Meanwhile, the utility is also planning an advanced sodium‐sulfur
battery project to improve power quality, facilitate integration of renewable energy sources, and
provide backup power to customers in case of power failure (Marshall, 2009b; California Energy
Commission, 2010c). The utility received funding for this demonstration project, which was
planned for a late 2010 installation in Silicon Valley, from the California Energy Commission
and planned it in coordination with the Electric Power Research Institute, NGK Insulators Ltd.,
and S and C Electric. P G and E will add these projects to its existing storage portfolio that
includes the Helms Pumped Storage Facility, a 1.2 GW hydroelectric facility near Fresno,
California.
2.1.3.6.4 Municipal Utilities
Several municipal utilities similarly have energy storage investments underway. In a
presentation as part of the Energy Commission’s 2011 IEPR Committee Workshop on Energy
Storage for Renewable Integration, the Sacramento Municipal Utilities District (SMUD)
explained that greenhouse gas regulations, RPS‐driven additions of renewable energy sources,
and the need for peak load management are driving its interest in storage (Rawson, 2011).
Citing both utility grid benefits and customer side benefits, the utility—which received ARRA
funding for installation of two zinc‐bromine flow battery systems and several different
applications of lithium ion battery systems to firm renewable output, reduce peak load and cost
to serve peak, and improve system reliability—acknowledged that it will need both bulk and
distributed storage in the long run. SMUD is planning to demonstrate sodium nickel chloride
and has begun investigation of compressed air energy storage for similar applications. In
addition, SMUD is advancing the 400 MW Iowa Hill pumped storage project to further
integrate variable energy resources, pending receipt of its new FERC license, favorable results
from geotechnical work to verify suitability of the site to support the project, and acceptable
cost/benefit assessments on the value proposition of the project to SMUD and its customers.
The Department of Water and Power for the City of Los Angeles (LADWP) also has begun to
pursue energy storage options. In September 2010, LADWP announced that it will collaborate
with BYD Ltd. to develop “a grid‐scale battery project for renewable energy storage.” The
anticipated 5 to 10 MW power storage unit will be housed at the utility’s Tehachapi wind farm
and used for power reliability and renewable generation integration. According to the city, the
96
partnership “aims to test, develop and ultimately demonstrate the feasibility of large‐scale
energy storage options for electric utilities across the Country (City of Los Angeles, 2010).
2.1.4 Comparative Cases: Energy Storage Policy Out of State and Abroad
This section presents tariff reforms, energy storage pilot programs, and policy initiatives in
other states and jurisdictions, and highlights storage‐related activities in other countries. This
brief summary does not represent comprehensive treatment of all storage‐related policy and
regulatory initiatives across the country or abroad.
2.1.4.1 Jurisdictions in the United States
2.1.4.1.1 New York Independent Service Operator (NY‐ISO)
The NY‐ISO created a separate resource category for energy storage devices to accommodate
participation in the wholesale market. As in California’s ISO territory, NY‐ISO purchases
regulation services on behalf of all system users. In 2009, NY‐ISO developed new market rules
and software protocols to facilitate the integration of limited energy storage resources (LESRs)
into New York’s regulation services market, citing multiple benefits to the grid system
stemming from the ability of LESRs to respond quickly to rapid control and dispatch signals.
About 80 MW of fast‐response energy storage has been proposed in the form of flywheel and
battery systems. The NY‐ISO expects additional types of energy storage systems to join the mix,
including flow batteries, fuel cells, and supercapacitors. In addition, NY‐ISO views the
economic potential of longer‐duration energy storage systems to provide arbitrage and load‐
shifting, reducing the frequency of dispatch of fossil fuel peaking plants, assisting with wind
energy integration and ramping, and smoothing the variability of increased wind and solar
systems on a minute‐by‐minute and second‐by‐second basis (New York Independent Service
Operator, 2010).
Several changes have been made to the NY‐ISO market to accommodate the addition of energy
storage resources. NY‐ISO defined the LESR as a new class of Regulation Service provider,
characterized by its ability to provide continuous, six‐second changes in output and its inability
to sustain operation at maximum energy withdrawal or injection for an hour (Ibid., p.10). The
NY‐ISO market limits devices classified as LESRs to providing only Regulation Service in the
NY‐ISO market. A market clearing price determines payment, as with traditional providers,
but LESRs only receive payment for the net energy injected into the system, not for a “mileage
payment” based on total energy the technologies withdraw or inject. And unlike traditional
Regulation Service providers, for which grid operators track responsiveness to control signals
and reward based on performance, the LESR currently is not subjected to performance tracking
and does not receive compensation commensurate with the speed and accuracy of its response.
97
FERC approved NY‐ISO’s LESR tariff on May 15, 2009 (Federal Energy Regulatory
Commission, 2009).
NY‐ISO schedules LESRs based on regulation availability and “fast first,” where the Automatic
Generation Control (AGC) dispatches LESR prior to traditional, slower‐moving generators
(New York Independent Service Operator, 2010, p.11). Scheduling fast‐response Regulation
Service devices first increases system efficiency, improves control performance, and reduces
cost. However, if the system requires regulation in a direction that the storage device cannot
provide, the grid operators will dispatch LESR after all other resources (Cutter, 2009, p.9).
According to NY‐ISO, the Regulation Service product is the highest priced ancillary service
product in New York’s wholesale electricity market, creating an attractive market for additional
investment in energy storage. Generally, regulation service is the most attractive product for
energy storage in the ancillary services market due to its higher price, as compared to services
such as spinning or non‐spinning reserve (Ibid., p.2).
Building on this market opportunity, New York has committed to becoming a global leader in
energy storage. The state has set aside $25 million to finance development of energy storage
and battery testing under its New York Battery and Energy Storage Technology Consortium.
An additional $15 million funds advanced vehicle components design and manufacturing,
alongside an initiative to convert the state’s hybrid vehicle fleet into plug‐in electric (New York
Independent Service Operator, 2010, pp.11‐13). Like California, New York is also host to energy
storage demonstration projects supported by federal funds and research at Sandia National
Laboratories.
2.1.4.1.2 Midwest ISO (MISO)
Setting an example for the approach taken in NY‐ISO, MISO was the first to create a separate
tariff class for Storage Energy Resources (SERs), distinct from that applied to generation or
demand response resources (Texas Energy Storage Alliance, 2010, Appendix A). FERC
approved the tariff on December 18, 2008 (Federal Energy Regulatory Commission, 2008b).
MISO then worked with stakeholders to make additional modifications to the dispatch of SERs
for regulation reserve, which FERC approved on December 31, 2009.37
2.1.4.1.3 Independent Service Operator – New England (NE‐ISO)
In 2008, NE‐ISO implemented a pilot program to incorporate 13 MW of “alternative regulation
technologies,” including energy storage devices and demand response. This represents 10
percent of the average hourly regulation requirements in the territory (ISO New England, 2010,
37 See Federal Energy Regulatory Commission. Docket ER09‐1556‐000: Order Approving Multi
Party Facilities Construction Agreements, December 31, 2009.
98
p.8). This program was the first pilot nationwide designed to include technologies not
otherwise able to participate in the regulation market. The program established a range of
technical specifications that it sought for participation, with a minimum of 1 MW capacity and
the ability to respond within 4 seconds to AGC signals. One purpose of the pilot was to identify
the preferred parameters for each participating resource, a necessary step for any ensuing
changes to market and tariff structure. Under the pilot, payment for provision of regulation
services is based on a “mileage” payment, or the sum of the absolute value of energy
withdrawal and injection into the system. Program participants pay for the cost of
interconnection and metering (ISO New England, 2010, pp. 5, 6, 12). As in NY‐ISO service
territory, grid operators dispatch faster resources first. A series of changes to the market rules
that are compatible with energy storage and other non‐traditional regulation technologies have
followed the NE‐ISO pilot, including a recent initiative to consider regulation service payment
adjustment based on a pay‐for‐performance structure.
2.1.4.1.4 PJM Interconnection (PJM)
As in NE‐ISO, PJM has enabled limited energy resources to participate in the regulation service
market as part of an Alternative Technology Resource pilot program (See PJM Interconnection,
2010a, Sec. 3.3.1.2). As in NY‐ISO, PJM established a frequency‐only regulation signal, allowing
fast‐responding LESRs, such as flywheels and batteries, to self‐manage their energy neutral base
points in real‐time under existing market rules. However, the current tariff compensates all
regulation service providers the same, providing no additional compensation for more rapid
and accurate response to control signals (PJM Interconnection, 2010b). Payment for regulation
service is primarily based on foregone energy revenue opportunity cost, not on additional
performance criteria. Additionally, under the current tariff, PJM has limited its definition of
“energy storage resource” devices to new rapid response systems, specifically flywheels and
batteries, so that treatment of the defined energy storage resource as a market seller will not
cause unintended consequences, based on the lack of information about devices and
technologies still likely to emerge. Thus, for the purpose of participation in the energy and
ancillary service markets, flywheel, battery storage, and conventional energy storage such as
pumped hydro are on equal footing (PJM Interconnection, 2010c, pp.6, 2).
In January 2011, PJM initiated a process to consider performance‐based pricing that would
promote more accurate regulation response and enable PJM to expand what is currently limited
use of its dynamic regulation signal. This process is in the early stages and PJM has not
released a timetable for the rulemaking, although it has identified key areas for activity,
including defining bright‐line criteria for performance requirements for incentive payment
eligibility and the design of a mechanism that incentivizes accurate tracking to the regulation
signals. PJM has acknowledged that the alternative resource technology investor community
99
perceives a misalignment of incentives in the PJM regulation market; this process aims to
“properly realign these incentives to attract these new resource types, provide better
operational performance and long‐term cost benefits (PJM Interconnection, 2010c).
2.1.4.1.5 Texas and Electric Reliability Council of Texas (ERCOT)
In January 2011, the Public Utility Commission of Texas reported to the Texas legislature that
“because some storage technologies have the capability to function as both transmission and
generation, the Legislature may wish to clarify the role or roles of storage in the competitive
market, whether as a regulated transmission asset, the cost of which would be recovered
through regulated rates, or as a generation asset that would recover its costs in the various
competitive markets” (Public Utility Commission of Texas, 2011, p.94). The Commission
outlined several benefits of energy storage, including “the flexibility to adjust energy
production or consumption to offset changes in wind or solar power production,” relieve
transmission constraints, and assist with meeting demand during peak periods. It summarized
the barriers to deployment in Texas as related to cost, lack of industry experience with energy
storage and grid characteristics, and insufficient guidance for industry and regulators to define
storage devices and develop operational standards and compensation (Ibid., pp.74‐75).
Texas lawmakers responded with the filing of Senate Bill 943 relating to the “classification and
use of energy storage equipment or facilities,” which passed the Texas legislature and was
signed by the governor on June 17, 2011. SB 943 requires the utility commission to classify
specific energy storage equipment or facilities as generation assets. The law also entitles
facilities in this asset class to be eligible to interconnect to the grid, obtain transmission service,
and sell electricity or ancillary services at wholesale. The legislature amended a previous
version of the bill that had created a separate asset class for energy storage (See Texas
Legislature Online, Senate Bill 943 [Carona]).
In addition, Electric Reliability Council of Texas (ERCOT) is considering similar tariff and
protocol changes as described for other jurisdictions. The Texas Energy Storage Alliance
industry group has recommended several changes to the regulation dispatch to take advantage
of the rapid response and accuracy afforded by energy storage devices (Texas Energy Storage
Alliance, 2010, p.6). Texas stakeholders have recognized the potential for energy storage to
assist with rapid wind and load ramps, ancillary services, and optimizing transmission capacity
and energy arbitrage (Customized Energy Solutions, 2009), due to Texas’ high wind generation
output, expected to exceed 10,000 MW nameplate capacity by 2013, (Electric Reliability Council
of Texas, 2010) and to their transmission limitations. ERCOT recently initiated discussion over
near‐term operational changes proposed by market participants to remove technical barriers to
entry of storage in the energy markets (AES Energy Storage, 2011).
100
2.1.4.2 Foreign Jurisdictions
Many of the same interests driving energy storage domestically exist outside of the United
States.
2.1.4.2.1 Germany
Germany has set ambitious national targets and implemented national policies to achieve
reductions of greenhouse gas emissions from 1990 levels by 55 percent by 2030 and by 70
percent by 2040. The German government has similarly impressive targets for renewable
energy production as a percentage of gross electricity consumption, committing to 35 percent
renewable sources by 2020, 50 percent by 2030, and 65 percent by 2040 (Germany Federal
Ministry of Economics and Technology, 2010, p.5). Germany views energy storage as integral to
its national plan for deployment of intelligent smart grids and demand‐side load management,
to be enabled by implementation of variable electricity pricing and communication
technologies, as well as to the long‐term integration of variable renewables. The short‐term
focus is on maximizing domestically available and cost‐effective pumped storage capacity; in
the long‐term, Germany will focus on expanding to use foreign pumped storage plants and
capitalize on investments in research and development of advanced CAES, hydrogen, and
battery storage. The national government has outlined policy measures to assist the expansion
of Germany’s energy storage capacity. These include extending exemptions from grid access
charges to pumped storage and other electric storage facilities and approving energy storage
systems for participation in the control energy market. To support research and development,
in 2011 the German government will release plans for a “Comprehensive Energy Research
Programme” through 2020 with directed funding initiatives for grids and energy storage (Ibid.,
pp.19, 21. 26).
The German Energy Agency (DENA) recently completed a comprehensive analysis of its
electric grid modifications needed to accommodate 25 to 30 percent renewable energy resources
by 2020 (Germany Energy Agency (DENA), 2010). The analysis included an examination of the
costs associated with expanding transmission capacity, increased energy storage, and demand‐
side management measures, assuming market‐driven deployment. It compared various
integration solutions, using energy storage facilities to provide increased system flexibility, with
the effects of other methods, such as demand‐side management, improved wind forecasting,
and provision of balancing energy by wind turbines or biomass plants. 38 DENA found that in
the current regulatory and market environment, in which Germany’s electricity market is
economically independent from operation of the grid, “the economically optimal behavior of
storage facilities on the electricity market do not necessarily result in behavior which relieves
38 Ibid., p. 16.
101
grid bottlenecks”(Ibid., Sec. 13.1.3). Because market‐driven operation of energy storage affects
the flow of electricity in the grid without necessarily accounting for grid operation constraints,
the use of energy storage could result in further transmission constraints and the need for
additional transmission capacity. To integrate large amounts of wind energy into the grid, the
German study concluded that it would be more economical – at current energy storage
technology costs – to further expand the transmission grid rather than build energy storage
(Ibid., Sec. 13.4.3). The study recommended that policy makers implement incentive systems in
the electricity market to encourage the use of energy storage facilities in a manner that
coordinates with grid operation to relieve transmission bottlenecks and lower total system costs
(Ibid., Sec. 24).
The German study also found that for increased renewable integration scenarios beyond 2020,
some energy storage technologies could be cost‐effectively utilized for system flexibility, load
smoothing/ peak shaving, and for balancing power (Germany Energy Agency (DENA), 2010,
Sec. 23.6). The most cost‐effective uses were for time shifting and smoothing of generated wind
energy at the local level (Ibid., Sec. 23.9.5). Pumped storage represented the only option
currently economically feasible for grid scale applications, but the study found that adiabatic
compressed air storage and hydrogen storage comprised potentially viable, albeit high‐cost,
storage solutions to mitigate wind energy variability (Ibid., 22). The study determined that fast‐
response energy storage technologies such as flywheels and chemical batteries would be
unlikely to contribute to integration of renewable energy sources due to technical limitations in
their storage capacity and service life (Ibid., Sec. 22.5, 22.6).
2.1.4.2.2 China
China has no overarching policy relating to energy storage. However, multiple municipalities
have implemented policies to encourage local development and deployment of storage
technologies, and the national government has allocated resources to numerous demonstration
projects as part of its plan for strong smart grid development in 2011‐2015. Shanghai is at the
forefront of this effort. In April 2010, the municipal government launched the “Development
and Promotion of Smart Grid Technology” program for Shanghai, which aims to transform
Shanghai into a national smart grid and energy storage research and development center. Some
of the local policies include direct funding or subsidies to support development work.
According to the China Energy Storage Alliance , the lack of a direct national policy supporting
the development of the Chinese energy storage industry is a barrier to the establishment of a
strong Chinese market for storage technologies (China Energy Storage Alliance, 2011).
2.1.4.2.3 Australia
102
Australia’s federal government has dedicated $100 million toward the “Smart Grid, Smart City”
initiative – a commercial‐scale, test business case smart grid rollout (Australian Department of
the Environment, Water, Heritage and the Arts, 2009). The project identifies distributed energy
storage as one of the four priority demonstrations, and several aspects of the program
incorporate energy storage. For example, one Australian energy storage manufacturer
implementation proposal included installation of 5 kW/ 10 kWh battery units at 60 households
for such applications of peak load shifting, residential back‐up, deferral of transmission
upgrades, and complement to residential PV systems (China Energy Storage Alliance, 2011,
p.3). In addition, the Australian Department of Resources, Energy and Tourism has a dedicated
“Advanced Electricity Storage Technologies Program,” which funds storage demonstration
projects ranging from hydrogen‐enabled solar storage to zinc‐bromine flow battery (Australian
Department of Resources, Energy and Tourism, Advanced Electricity Storage Program).
2.1.4.2.4 Japan
Japan’s energy goals for 2030 include raising the zero‐emission power source ratio from the
current 34 percent to 70 percent (Japan Ministry of Economy, Trade and Industry, 2010a). Japan
has established a national Renewable Portfolio Standard and is considering implementation of a
Feed‐in Tariff to incentivize necessary investment in renewable energy sources to meet this
goal. A number of Japan’s government‐supported renewable energy and smart grid
demonstration project include integrated energy storage systems. In its “Demonstration of
Next Generation Energy and Social Systems” program, the Ministry of Economy, Trade and
Industry (METI) selected four large‐scale “smart community” demonstration projects, several of
which will pilot electrical battery storage technologies in conjunction with distributed
renewable generation and smart grid networks (Japan Ministry of Economy, Trade and
Industry, 2010b). Energy storage research and development has received significant support
from Japan’s New Energy and Industrial Technology Development Organization (NEDO), an
administrative agency under METI that oversees multiple projects relating to grid‐connected
electric energy storage systems and battery systems for next‐generation electric vehicles. NEDO
research has developed parameters for battery storage‐based voltage control and frequency
balancing in conjunction with clustered and large‐scale photovoltaic power generation. One
purpose of this research is to enable progress toward Japan’s Photovoltaic Roadmap for 2030
goals without restricting PV output (Nakama, 2010).
103
CHAPTER 3: Technology Gaps, Research and
Development Needs, and Policy Challenges
Introduction:
The focus of this chapter is to facilitate informed decision making about the role of energy
storage in California’s transition to a clean energy future by providing a summary of key
findings on technology and cost gaps, research and development needs from Chapter 1, and
key policy barriers that remain to be addressed by the many proceedings outlined in Chapter 2.
The most prominent of the policy barriers creating challenges to energy storage system
deployment in California may be categorized in terms of technology cost and commercial
availability, valuation and cost recovery mechanisms, and/or lack of awareness and
collaboration. A recent U.S. DOE‐sponsored study conducted by Sandia National Laboratory on
industry needs for grid‐scale storage applications categorized these challenges as “insufficient
technical progress, limited demonstration and performance data, lack of standards and models,
weak stakeholder understanding and support, and deficient market structure.”39
3.1 Technology Status Findings: Gaps and R&D Priorities
U.S. Department of Energy Advanced Research Projects Agency ‐ Energy (ARPA‐E) research
has concluded that “significant advances in materials and devices are needed to realize the
potential of energy storage technologies.” (Nexight Group and Sandia National Laboratories,
2010a, p.1). This conclusion may be influenced by ARPA‐E’s objective to pursue high risk, high
payoff basic science that produces order of magnitude performance increase and/or cost
reduction. A related report notes that high cost, technical complexity, efficiency and lifecycle
concerns render many energy storage technologies not yet commercially viable for large‐scale
production and grid‐scale integration (Nexight Group and Sandia National Laboratories, 2010b,
p.24). Although many energy storage technologies are already in commercial use or in
demonstration, there continue to be challenges related to materials constraints and costs, life
cycle and performance uncertainties, lack of demonstration and performance data,
manufacturing requirements, and a need for superior control system and power electronics for
seamless interoperability between storage devices and the grid.
39 U.S. Department of Energy, Office of Electricity Delivery and Energy Reliability and Office of
Energy Efficiency and Renewable Energy, Solar Technologies Program, “Electric Power
Industry Needs for Grid‐Scale Storage Applications,” Sandia National Laboratories, December
2010, p. 23.
104
As demonstrated in Chapter 1, the existing limitations, opportunities, and advances are best
explained in relation to each type of storage technology. However, from a policy perspective, it
may be possible to address some of these limitations through changes to the regulatory
structure, implementation of valuation and cost‐recovery methodologies, development of
uniform standards to measure key energy storage performance functions, including efficiency,
capacity, and duration, and through increased funding support for focused research and
development efforts. Certain storage technology costs, such as for Lithium Ion in particular,
may decrease based upon increased production of battery technologies to meet growing
demand for electric vehicles (Electric Power Research Institute, 2010a, p.5‐2). Many industry
participants acknowledge that production prices of energy storage technologies must decrease
before it will be feasible to achieve significantly greater use under the current market and
regulatory structures (China Energy Storage Alliance, 2011, p.2). Chapter 4 discusses a
framework for California to develop a strategic vision for energy storage for 2020 and reviews
funding incentives, demonstration opportunities, RD and D needs, and policy and regulatory
drivers that can influence the near and long‐term feasibility for increased deployment of energy
storage technologies.
3.2 Policy and Regulatory Findings: Barriers and Challenges
3.2.1 Valuation and Cost Recovery
3.2.1.1 The Need for Better Valuation and Cost Recovery Mechanisms
Energy storage advocates and manufacturers frequently identify the lack of a framework to
valuate or monetize the benefits provided by energy storage technologies, located on either side
of the utility meter, as a primary barrier to energy storage deployment. This barrier falls within
the “lack of standards and models” and “deficient market structures” categories listed in the
Sandia/ DOE report. Yet these deficiencies are not necessarily in existence across‐the‐board.
For example, the wholesale energy market, including regulation, capacity, and ancillary
services, provides opportunities for energy storage to capitalize on already‐monetized value
streams. The existence of these markets, however, does not signify that barriers to participation
or fair valuation for benefits provided have been addressed, as demonstrated by FERC,
California ISO, and the other ISOs/RTOs’ efforts to reevaluate these market mechanisms. Even
with the wholesale energy markets accessible to some energy storage technologies, California
lacks a cohesive framework to evaluate costs and benefits. An example of such a framework is
the 2010 Demand Response Cost Effectiveness Protocols, which provides a standardized method to
evaluate cost effectiveness of demand response programs and includes such inputs as avoided
energy costs, avoided distribution costs, and avoided environmental costs for greenhouse gases
(California Public Utilities Commission, 2010c).
105
Studies also suggest that several recognized benefits of energy storage (such as emissions
reductions or certain avoided‐cost benefits like transmission deferral) represent potential value
streams without clear valuation mechanisms. Furthermore, cost recovery mechanisms, based
upon transparently monetized value streams, are necessary for developers and system owners
to realize or internalize the benefits provided. The high costs of current energy storage
technologies and investment risk will persist without adoption of a valuation framework that
monetizes the independent benefits and that creates opportunities for cost recovery. A key
consideration is the potential market size of the recognized value streams to determine their
impact on the energy storage value proposition overall, or for specific applications.
In its July 2010 white paper assessing barriers and opportunities for energy storage, the CPUC
identified as a primary challenge the need to develop a tool to evaluate cost effectiveness that is
technology‐neutral and that evaluates the benefits and costs of a particular technology in a
particular application (California Public Utilities Commission, 2010d, p.5). A critical feature of
this challenge is how to allocate the costs and benefits of storage across the range of services
that are affected, including generation, transmission, distribution, and regulation. In a July 2010
white paper released by UC Berkeley’s Center for Law, Energy and the Environment (CLEE)
and UCLA’s Environmental Law Center and Emmett Center on Climate Change and the
Environment (UCLA Law), the authors emphasized the need for regulatory proceedings at the
state and federal level that will adopt methodologies to analyze costs and benefits and enable
recapture of the values provided (UC Berkeley/UCLA Law, 2010, p.13). CLEE/UCLA Law
found the lack of a formal mechanism for calculating the value of the benefits and resource
savings brought by energy storage to be a primary barrier to wider implementation of storage
in California. Without such a mechanism, current pricing of electric energy and services does
not enable storage owners to internalize or realize most of energy storage resource benefits (See
Eyer, 2010, p.xvi). As the CPUC contemplates a cost‐benefit methodology for electric energy
storage and potential cost‐recovery mechanisms and ownership models across the range of
service applications, the input provided by a number of industry, utility, and government‐
sponsored studies is critical.
Although selecting or developing a valuation methodology is beyond the scope of this report, a
brief review of the approaches taken in other studies may assist the CPUC and the Energy
Commission in their efforts to create a valuation structure for energy storage technologies.
3.2.1.2 Approaches to Valuating Benefits and Determining Cost Effectiveness
In February 2010, the U.S. Department of Energy sponsored an analytical effort to develop a
high‐level, technology‐neutral framework for assessing the benefits and market potential for
energy storage in the electric grid. The theme driving the resulting “Benefits and Market Guide”
is the potential to aggregate the various, complementary benefits offered by energy storage to
106
create attractive value propositions for the market and for industry (Eyer, 2010, p.xv). The
guide also recognizes that some benefits may not be aggregated, due to institutional barriers or
technical/operational conflicts. The guide’s framework is useful to this analysis insofar as it
characterizes 26 benefits associated with the use of energy storage for electric utility
applications, as well as the financial value, market potential, and estimated economic impact for
various applications. 40 It provides eight examples of potentially cost‐effective value
propositions, such as “electric energy time‐shift plus transmission and distribution upgrade
deferral,” or “renewables energy time‐shift plus electric energy time‐shift plus electric supply
reserve capacity.” The value of individual benefits depends upon a number of assumptions
relating to duration discharge, power rating, financial criteria, and market potential, and may
be meaningless if the stakeholder cannot internalize the benefit. Finally, any particular benefit is
application specific. The guide categorizes 17 electric grid related applications for energy
storage into five categories: (1) electric supply, (2) ancillary services, (3) grid system, (4) end‐
user/ utility‐customer, and (5) renewables integration (Eyer, 2010, p.21).41 It suggests that
lawmakers and regulators recognize a societal storage value proposition that values the benefits
that may accrue to utility customers as a group or to society at large and may be
complementary to value propositions that directly benefit the individual owner/user of a
storage system (Ibid., pp.134‐35).
EPRI also rendered a valuation framework that identifies ten, key near‐term applications for
energy storage across the electrical system and evaluates discrete benefits that can be
aggregated across those categories. The ten categories of near‐term applications are: (1)
wholesale market, (2) renewables integration, (3)/(4) stationary and transportable storage for T
and D, (5) distributed energy storage, (6) residential customer aggregated systems, (7)/(8)
commercial/industrial power quality, reliability, and energy management, and (9)/(10) home
energy backup and management (Electric Power Research Institute. 2010a, p.ES‐9).
Many energy storage technology developers take a similar approach to EPRI. In its comments
to the CPUC under AB 2514, the California Energy Storage Alliance (CESA) suggested that the
40 The Guide’s authors take care to distinguish a benefit, which connotes a value that may be
monetized, or may be qualitative in nature, from an application, which connotes a particular use.
(Eyer, 2010, p. 2 (¶ 1.4.1)).
41 The 17 applications are grouped as follows: (1) Electric supply (electric‐energy time shift,
electric supply capacity); (2) Ancillary services (load following, area regulation, electric supply
reserve capacity, voltage support); (3) Grid system (transmission support, transmission
congestion relief, T&D upgrade deferral, substation on‐site power); (4) End‐user/Utility‐
customer (TOU energy cost management, demand charge management, electric service
reliability, electric service power quality); (5) Renewables Integration (renewables energy time‐
shift, renewables capacity firming, wind generation grid integration).
107
CPUC may need to consider different methodologies to evaluate costs and benefits for energy
storage depending on the application, especially for bulk versus distributed applications, since
one serves primarily at the transmission level and one at the distribution or customer level
(California Public Utilities Commission, Docket R10‐12‐007, Comments of the California Energy
Storage Alliance, p.8). CESA also emphasized the need for the CPUC to develop a methodology
to create a Resource Adequacy (RA) value for energy storage technologies not presently
counted under the program. Some additional (though not all) stakeholders shared this view
and suggested that the CPUC should address RA value in a separate proceeding from the AB
2514 rulemaking. With respect to developing a cost‐benefit methodology, the Western Power
Trading Forum, with membership of numerous energy service providers and investment firms,
suggests that the determination of storage application costs and benefits should be determined
by the end‐user market based upon cost and potential application information that is made
available through the CPUC’s proceeding (Ibid., Comments of the Western Power Trading
Forum).42
Southern California Edison (SCE), in its effort to build upon the work of Sandia, EPRI, and
others and to create a practical cost‐effectiveness and valuation methodology from the utility
perspective, also decided upon an “application‐focused valuation methodology” (SCE White
Paper, 2010, p.4).43 SCE first identified discrete operational uses with independent values or
benefit streams for energy storage and ʺbundledʺ the potential operational uses into
applications, based on both location and operating profile. SCE then matched these applications
with best‐fit technology choices and assessed the pairings from an economic and feasibility
standpoint, using a variety of 2020 scenarios.
SCE noted that the utility of its valuation and application‐technology pairings is limited by
ongoing regulatory uncertainty, especially concerning cost recovery mechanisms and asset
classification (SCE White Paper, 2010, p.63). For each discrete operational use that SCE
identifies, it includes a short list of value metric uncertainties that exacerbate the difficulties of
valuation (See ibid., p. 24, Fig. 3). EPRI’s valuation analysis also recognizes that policy action on
several fronts, including definition of ownership structures and business models that account
for storage flexibility, rules regarding allocation of storage system costs, and tariff and metering
structures that accommodate the unique bi‐directional nature of storage, is required to realize
42 Western Power Trading Forum describes itself as a “broad‐based membership organization
dedicated to encouraging competition in Western states electric markets” and is primarily
concerned with the long‐run cost of electricity to consumers and with maintaining system
reliability. See http://www.wptf.org/.
43 Rittershausen and McDonagh, “Moving Energy Storage from Concept to Reality: Southern
California Edison’s Approach to Evaluating Energy Storage,” Southern California Edison, 2010.
108
the full benefits of energy storage assets (Electric Power Research Institute, 2010a, pp.5‐7). The
Sandia report similarly lists a number of important challenges that affect potentially attractive
value propositions, ranging from the high installed cost per kW of energy storage to the limited
mechanisms available for sharing of risk and reward/benefit (Eyer, 2010, p.139). Chapter 5 will
review potential solutions to these challenges in more detail.
Another approach to valuation would be to follow the avoided cost model used for qualifying
facilities (QFs) under the Public Utility Regulatory Policies Act of 1978 (PURPA). 44 Under
PURPA, a regulated utility is required to buy power from QFs that can produce power for less
than the utility’s “avoided cost,” or the cost to the utility to generate or procure the power itself.
In California, avoided cost is generally based on the cost of natural gas, since it is presumed to
be the fuel that would be used in the marginal generating unit. As with many renewable
generation sources today, applying the same approach to storage might not provide sufficient
revenue to make most or all storage technologies marketable, based on the relatively low cost of
natural gas and the failure of the avoided cost approach to consider the values attached to
storage that go beyond the avoidance of buying a unit of natural gas.
The determination of a valuation methodology requires significant analysis beyond the scope of
this report and likely will receive extensive treatment in the CPUC’s AB 2514 rulemaking.
Generally, the value of energy storage technologies vary depending on the technology, location,
and the market value of services they provide. Policy makers will need to devise a framework
that addresses the multiple values and potential overlapping nature of energy storage’s
benefits. They might consider studying how the various analytical approaches mentioned
above would value the same energy storage projects under the same operating and market
conditions to compare and identify improvements for each approach. And if the CPCU
decides to set a broad mandate for energy storage according to a total optimal deployment
figure, as AB 2514 allows, the agency could use further valuation work to inform the setting of
that figure.
3.2.1.3 Important Considerations for a Valuation Framework
The CPUC has determined to proceed with a two‐phase rulemaking process under AB 2514. In
Phase 1, the CPUC will develop policies and guidelines for energy storage based upon current
uses for storage and potential deployment schemes that can maximize grid and environmental
benefits. This Phase will provide the Commissioners and participating parties with the requisite
information about storage technologies, system benefits, and barriers to deployment. Phase 2,
which will address energy storage application values, costs, and means to determining cost
44 Energy Commission Chairman Weisenmiller raised the question whether energy storage
systems should be valuated based on avoided cost at the 2011 IEPR Committee Workshop on
Energy Storage on April 28, 2011 (See California Energy Commission, 2011b, p. 42).
109
effectiveness, will lay the foundation for a decision on the viability and cost effectiveness of
procurement targets. The CPUC rulemaking ultimately should resolve many of the
uncertainties relating to valuation and cost recovery, providing a means for energy storage
system developers, owners, and users to predict and realize more of the value of benefits
provided by their system.
Considerations that the CPUC will need to address include:
• How to group operational uses and associated benefits, such as by application(s) and
location.
• Whether to prioritize certain applications to meet the objectives set forth by AB 2514.
• Whether procurement targets, if adopted, can help resolve technology cost and
performance challenges given current market and regulatory structures.
• Whether to pursue integrating energy storage into California’s loading order and the
CPUC’s RA and LTPP processes.
• How increased energy storage deployment might impact other policy priorities,
including flexible demand/demand response, distributed generation, and electric vehicle
charging.
• How to compensate energy storage owners/ operators for services not covered by
California ISO markets. An example would be the use of performance‐based contracts
for storage systems installed on the distribution grid that may defer transformer
upgrades, reduce voltage flicker caused by photovoltaic generation, and frequency
regulation and grid stabilization.
• A methodology to determine which services/applications and related value streams may
be aggregated to maximize financial return to a storage system without double‐counting
benefits or committing the same resource to incompatible uses at one time.
In discussing the approach of “stacking” or aggregating energy storage benefits, several studies
have noted that the value of energy storage may only justify deployment when operation of the
device allows it to take advantage of more than one individual source of revenue by combining
services (Denholm, 2010, p.11). It should be noted that these services, which the CPUC
categorizes as economic and operational, are not mutually exclusive and will overlap
(California Public Utilities Commission, 2010d, p.5 fn12). Sandia National Laboratories provides
a detailed explanation of the challenges that accompany aggregating benefits, including the
issue of overlap and the potential for technical and operational conflicts between multiple
applications. Their work cites the difficulty of aggregating benefits that accrue to different
stakeholders as one of the primary challenges faced by otherwise financially viable value
propositions (Eyer, 2010, pp.123‐24). The Public Utility Commission of Texas identified the
challenges associated with storage serving multiple roles or functions at the same time as
110
“cross‐subsidization, competition, and discrimination.” (Public Utility Commission of Texas,
2011, p.74).
Stakeholders to the AB 2514 process have weighed in on this issue. Some emphasize that the
valuation process should quantify ancillary service benefits and environmental quality benefits
to ensure that the complexity of energy storage does not undermine its fair comparison to other
emissions and energy reduction strategies (See California Public Utilities Commission, Docket
R10‐12‐007, Comments of Environmental Defense Fund; California Energy Storage Alliance).
This effort might create a quantifiable benefit that can be stacked alongside the traditionally
monetized value streams, increasing the overall value proposition for energy storage. Others
argue that stacking benefits or quantifying environmental or ancillary benefits may lead to
inflated estimates of the value of storage if non‐mutually exclusive categories are treated as
additive (Ibid., Comments of Calpine Corp.).
Thus, a valuation methodology for energy storage will need to delineate which categories of
uses or applications may or may not be aggregated, as well as how to allocate costs and
resulting benefits, in part based upon the ownership or business model(s) utilized (Electric
Power Research Institute, 2010a, pp.5‐9; Eyer, 2010, p.139). A valuation method must coincide
with the establishment of cost recovery mechanisms that are sufficiently reliable and long‐term
technologies which may require years to provide a return on capital investments can receive
financing. In Chapter 5, this report provides several examples of promising applications for
energy storage and suggestions for viable cost‐recovery and policy or regulatory changes that
might increase the value and viability of storage technologies.
3.2.2 Awareness and Collaboration
Several reports have cited a lack of stakeholder awareness of energy storage benefits, as well as
a lack of coordination and collaboration amongst stakeholders and at the inter‐ and intra‐
agency level, as key barriers to transforming and modernizing the grid using energy storage
technologies. The Sandia report identified “limited familiarity, knowledge, and experience
base” with storage technologies, benefits, and system integration as one of the many challenges
that must be overcome by energy storage proponents (Eyer, 2010, Appendix G‐2). Another
Sandia/U.S. DOE report stated that a lack of understanding by much of the electric power
industry of the value and applications to which storage technologies may be applied,
contributing to a dearth of grid‐level demonstration projects that are necessary to validate the
benefits and obtain industry and public buy‐in (Nexight Group and Sandia National
Laboratories, 2010b, pp.24‐25). Without sufficient proof that a given business model and
technology combination can capture diffuse revenue streams, modernize the grid, and reduce
greenhouse gas emissions, regulators are unlikely to take the time to consider classifying
storage within the generation or distribution market structure or pursuing other methods to
111
deploy storage. The CLEE/UCLA Law white paper similarly identified a lack of awareness of
energy storage benefits as one of three primary barriers to expanding energy storage in
California (UC Berkeley/UCLA Law, 2010, p.3).
The passage of AB 2514 and the increased focus of California lawmakers and energy agencies
on examining the costs and benefits of energy storage may have jumpstarted a process that
counters the lack of awareness reported above. The CPUC rulemaking under AB 2514 and the
Energy Commission’s 2011 IEPR workshops have generated a substantial response from the
regulated community, industry, and environmental and consumer groups. While these
processes may serve to increase awareness of applications for and the benefits of energy
storage, they do not dispel the need for increased coordination and collaboration around study
efforts, performance data sharing, and the development of consistent standards for testing
protocols and storage interconnection and communication. Nationally, the Sandia/ DOE report
suggests that the Department sponsor a clearinghouse of energy storage to “serve as an
authoritative source for the current status of energy storage research and commercialization
(Nexight Group and Sandia National Laboratories, 2010b, p30). The clearinghouse would
comprise an up‐to‐date database of pending and completed storage installations and
demonstration projects, application needs, and technology specifications. It could facilitate
national and state regulatory efforts to evaluate the need for and benefits of energy storage, and
assist technology developers in meeting ascertained needs and specifications.
3.2.3 Competing Policy Priorities
An additional consideration and potential barrier to increased deployment of energy storage
technologies is that of competing policy priorities. Several stakeholders to this report’s
development, as well as to the 2011 IEPR and AB 2414 processes, have raised the question
whether a zealous drive to implement energy storage technologies may detract from or directly
interfere with other competing policy priorities, such as flexible demand/demand response,
low‐cost electric vehicle charging at off‐peak hours, and low electricity prices for ratepayers.
Most of these concerns center on energy storage technology’s cost, specifically whether an
energy storage procurement mandate would force the adoption of more expensive technologies
over other technologies or grid solutions that would cost ratepayers less.
As an example of a competing alternative, GE announced in May 2011 the commercial
availability of a 510 MW combined cycle, gas turbine with base‐load efficiency of 61percent to
ramp up and down at a rate of 51 MW/minute to adjust to wind and solar resources (Shakhtur,
2011). Although the load balancing is done with a fossil fuel, this innovation is significant in
terms of capabilities of modulating large solar and wind ramp rates. GE is introducing its first
units in Europe, China, India and Brazil, while the company waits to see what U.S. policy will
be on renewables and climate change. The one noted U.S. exception is California due to its
112
33percent Renewable Portfolio Standard (RPS). San Diego Gas and Electric (S D G and E) on
May 23, 2011 filed for regulatory approval of 450 MW of GT peaker plants under a power
purchase agreement for precisely the purpose of rapid ramping for wind and solar variability
(San Diego Gas and Electric, 2011a).
The California ISO, in its ʺSummary of Preliminary Results of 33percent Renewable Integration
Study ‐2010 CPUC LTPP Docket No. R.10‐05‐006,ʺ does not call for energy storage to be among
the grid tools necessary to manage the system to meet the 33percent renewable goal (California
ISO. 2011a). While the study does not indicate that the California ISO would not use energy
storage if available as a viable resource, it does make the analytical case that storage would not
be necessary. The California ISO acknowledges that energy storage is a possible viable
resource, however, and recommends ʺrunning additional sensitivities to additional evaluation
[using] storage, pump hydro and demand response.” As a result, various Smart Grid
technologies and the high‐efficiency, simple‐cycle gas turbines will present compelling
alternatives to utility‐scale energy storage in the near term.
Despite the existence of these alternatives, many policies related to them are complementary,
although they may involve tradeoffs (See California Energy Commission. 2011b, pp. 40‐42). The
subject is closely linked to the need for a transparent valuation methodology, discussed above,
which would provide a fair basis for comparing the costs and benefits of energy storage against
other technologies or operational solutions to meet the same grid needs. With proper valuation
and price signals, the consumer market may attain some equilibrium based on response to
properly‐valued price signals (Ibid., p.41).
Since AB 2514 directs the CPUC to base its energy storage determinations on both viability and
cost effectiveness, price signals that emerge from the CPUC rulemakings must use a similar
metric for competing solutions. This comparison feature also relates to the need to evaluate
energy storage costs and benefits at the application level, since the value proposition for the
same technology might vary depending on which application(s) it serves and whether the same
range of applications can be served by the competition. Finally, it presents the CPUC and
Energy Commission with further reason to evaluate whether policy makers should integrate
energy storage into California’s loading order. This integration could clarify the state’s policy
priorities regarding which technologies should be sought first or whether several potential
solutions should be placed on equal footing in the loading order to allow market forces to
determine the ultimate mix of solutions. The CPUC and Energy Commission should evaluate
the risk that promoting energy storage may increase costs to ratepayers or skew a fair
comparison with competing solutions, as they simultaneously analyze the numerous ways in
which energy storage technologies help the state meet its energy and environmental goals.
113
CHAPTER 4: Applications and Scenarios Analyses for
Energy Storage in California – Background, Purpose,
and Approach
Introduction:
As discussed in Chapter 2, California’s energy agencies and the legislature have committed to
examining the benefits of increased deployment of energy storage in California. AB 2514 allows
the CPUC to set broad targets for energy storage procurement (and also allows the agency to
decline to set targets) but emphasizes that any such targets must be consistent with
technological viability and cost effectiveness. For the CPUC to determine appropriate targets, if
any, as a means to increase energy storage deployment, it must decide upon a basic framework
with which to analyze the costs and benefits of and various possibilities for target development.
Through the 2011 IEPR process, the Energy Commission similarly must identify a framework to
evaluate the possible roles for energy storage in California, develop a strategic vision for
deployment, and identify the policy and regulatory needs to achieve the execution of such a
vision. These agency processes should take into account the results of ongoing research by the
California ISO and others on the changes needed to maintain a reliable and stable electric grid
as the state moves toward its 2020 energy goals and beyond.
The remaining sections of this Chapter set forth three important elements to be considered in
the course of developing a vision for energy storage in California for 2020. These are:
• California’s grid needs and applications
• Possible policy, regulatory, and market‐based drivers
• An analytical framework to assess the costs and benefits of storage deployment and to
prioritize applications and technology areas for further study or incentive development.
These elements will serve as the basis for an overarching framework by which to analyze
various scenarios for energy storage deployment in California. Two important considerations
underlie this effort. First, the California Legislature has directed the CPUC to develop targets, if
any, for the deployment of energy storage in California. Although whether and in what form
the CPUC may elect to set targets remains undetermined, targets under AB 2514 are a likely
driver for energy storage technologies. Second, no framework can be effective without an
implementation strategy that, in the case of increasing deployment of energy storage, includes a
cost‐benefit valuation methodology and long‐term, dependable cost‐recovery mechanisms for
storage owners, users, and investors.
114
4.1 California’s Grid Needs and Applications Selection
This report offers three categories of grid applications to serve as a basis for analyzing the
technical and economic feasibility of energy storage solutions in California by 2020. The three
exemplary application categories are:
• Frequency Regulation,
• Renewables Grid Integration, and
• Community Energy Storage / Distributed Energy Storage Systems.
These grid application categories are intended to serve only as representative use cases for
energy storage deployment. They encompass more specific applications and operational uses
that several recent studies and workshops suggest to be strong. A report resulting from a two‐
day stakeholder workshop sponsored by the DOE and organized by Sandia National
Laboratories found that these five applications have “the greatest overall potential to benefit
power system planning and operations” (Nexight Group and Sandia National Laboratories,
2010b , p.16). However, these applications do not encompass all of the possible applications for
energy storage, nor does their selection imply that they are the “best” or even the “priority”
applications for California. Their use in this report is intended to be exemplary in nature and to
highlight the bandwidth of potential value that might be derived from energy storage
technologies in meeting the anticipated needs of the grid for 2020. These applications
correspond with several more specific applications identified by industry groups such as CESA
and analyzed in studies by EPRI, (See Electric Power Research Institute, 2010a), Southern
California Edison, (See SCE White Paper, 2010) Sandia National Laboratories on behalf of the
DOE, (See Eyer, 2010), and NREL (See Denholm, P, 2010).45
The decision to use three exemplary applications rather than prioritize the “best” applications
stems from the timing of this report, which will be completed prior to several significant studies
on California’s grid needs for 2020. Policy makers should look to these sources as they are
published for the most useful information regarding anticipated grid needs to meet the 33
percent RPS by 2020 goal. Although preliminary data are available, they generally do not
account for energy storage needs or impacts.
For example, in May 2011, the California ISO released a Summary of Preliminary Results of
33percent Renewable Integration Study as directed by the CPUC for the 2010 long term
45 These studies present a variety of ways in which energy storage applications or operational
uses can be categorized and provide important baseline information for the feasibility of
assigning storage to these applications. This report will draw on that baseline information in the
Applications and Scenarios Analysis, Chapter 11. See Appendix A of that report for a side‐by‐side
comparison of the different application categories identified in these studies.
115
procurement planning process. The summary presents preliminary findings regarding the
operational requirements for renewable integration in such areas as regulation up and down
and load following, under several different scenarios and based upon key underlying
assumptions. Based on the assumption that California achieves its demand side objectives for
the time period, the preliminary results suggest that “most operational requirements can be
satisfied with potential need for measures to address over‐generation conditions” (California
Independent System Operator, 2011a, Slide 51). California ISO’s preliminary conclusion is that
based upon current load projections, demand‐side management, assumptions for wind and
solar forecast error, and the expected mix of renewable generation sources and retirement of
sources, meeting California’s load and regulation requirements can be achieved without the
need for a significant amount of additional resources.
California ISO recommends updating the analysis in future years as assumptions evolve,
specifically to continue to refine forecast error assumptions using actual data. The ISO also
recommends running additional sensitivities to: (a) assess higher loads, (b) assess changes to
forecast error and requirements, (c) evaluate generation outages, (d) assess resources needed for
local capacity requirements, (e) further evaluate storage, pumped hydro, and demand response,
and (f) assess different assumptions of dynamic transfers (California Independent System
Operator, 2011a, Slide 52). These additional factors, as well as other considerations such as the
environmental and greenhouse gas emission impacts of utilizing existing resources to achieve
the load‐balancing for the future grid, must be weighed alongside California ISO’s preliminary
findings. For example, a study released by the California Council on Science and Technology
found that if electric generation in 2050 is predominantly based on renewable intermittent
resources, firming by natural gas turbines would likely result in emissions that alone would
exceed California’s 2050 targets (California Council on Science and Technology, 2011, p.4).
Thus, it will be critical for California to develop its zero‐emission load balancing technologies
concurrent with renewable resource development.
Although California ISO’s preliminary results do not address the role of energy storage in
meeting the projected grid needs, the grid in 2020 will require regulation and load‐following
resources and may require resources that can serve as a load sink at times of renewable over‐
generation. The remainder of this chapter will (1) outline possible drivers to support the
development and deployment of energy storage devices for those roles, and (2) define the
framework for an application and scenario‐based approach to evaluating the value and
feasibility of storage and to discerning the potential economic and environmental impacts.
116
4.2 Possible Drivers for Energy Storage Deployment in California
Policy makers should base the identification of targets, milestones, and actions necessary for the
development and deployment of energy storage technologies in California on a basic menu of
possible drivers. These drivers, policy and market‐based, can be grouped into five general
areas:
• Targets/ mandates,
• Tariff/ market reform,
• Adjustments to loading order / resource adequacy rules,
• Financial incentives, and
• Funds for R and D.
4.2.1 Procurement Targets
If promulgated under AB 2514, energy storage procurement targets for load‐serving entities
(LSEs) would constitute the most direct path to increased deployment of energy storage in
California. As with the RPS, energy storage procurement targets would require the state’s
utilities to incorporate storage into their procurement of resources to meet a variety of needs
specified in the statute. The targets and any complementary policies to “encourage the cost‐
effective deployment of energy storage systems, including refinement of existing procurement
methods to properly value energy storage systems” must be “technologically viable and cost
effective” (Assembly Bill 2514 [Skinner, Chapter 469, Statutes 2010], Sec. 2837, 2836.2(d)). The
targets would be mandatory; AB 2514 provides dates by which the utilities would have to
achieve the targets and sets out a framework for demonstrating compliance (see ibid., Sec. 2837,
2838).
The notion of energy storage target‐setting has sparked significant controversy. Some energy
industry stakeholders question the effectiveness of procurement targets that may act as a form
of “subsidy” and compete with other potentially lower‐cost services, such as demand response,
or risk interference with other policy priorities, such as low‐cost, off‐peak charging for the
burgeoning electric vehicle market. Other stakeholders believe that targets should not be
necessary and that energy storage technologies should compete with other options for
addressing identified grid needs in the most cost‐effective manner. A few energy storage
advocates, however, note that a broad‐based target set in megawatts or percentages would be
less complicated than a more detailed, application‐based approach and would allow the most
cost‐competitive and feasible technologies to secure contracts under the broad mandate. While
AB 2514 empowers the CPUC to make the final decision on targets, this report seeks to inform
that decision by highlighting the potential benefits and concerns associated with the various
options available to policy makers.
117
AB 2514 provides the CPUC with significant discretion in terms of the form of the targets, if
any. It does not specify that the targets be in percentages or megawatts, nor does it indicate a
preference for a single mandate (for example, 5percent of peak load) or more nuanced targets
that could be application‐ or location‐specific (for example, 5percent of load‐following capacity,
or 50 MW of customer‐side‐of‐the‐meter storage, 100 MW of utility‐side‐of‐the‐meter). The
application and scenario‐based assessment to follow will consider the possible impact of
various forms of targets on energy storage deployment, feasibility, and cost effectiveness, as
well as possible implications of targets for other policy priorities.
4.2.2 Tariff and Market Reform
Tariff and market reform as a policy driver may complement procurement targets or may serve
as an alternate mechanism, in lieu of a regulatory mandate. Chapter 2 provided a detailed
overview of multiple processes underway at FERC and at California ISO to remove barriers and
encourage participation in the electricity markets by energy storage technologies. Most
significant at the state level are California ISO’s modifications to the ancillary services market
rules and implementation of Regulation Energy Management. The California ISO could also
implement shorter market intervals, which would allow more energy storage technologies to be
competitive, as the entity is now considering. These processes could remove barriers to
participation for fast‐responding energy storage resources that could address the state’s
environmental and energy needs in a cost‐effective manner. At the federal level, FERC is
reviewing measures for frequency regulation compensation in the wholesale power market and
is considering whether to classify energy storage technologies so as to allow cost recovery as a
generation or transmission/distribution resource, or potentially as both, or as a separate asset.
Each of these processes is significant for its potential impact on cost‐recovery, and thus on the
economic feasibility of increased energy storage deployment. For example, a decision by FERC
to allow energy storage systems to serve at times transmission functions would allow the
storage system to recover some cost in FERC‐approved transmission rates. A decision to
classify energy storage technologies as a separate asset class would most likely shift jurisdiction
over the development of a cost‐recovery mechanism to the states. A decision to direct
implementation of a “mileage payment” in frequency regulation compensation by the ISOs and
RTOs would increase the attractiveness for storage participation in that market, as would
shorter market intervals.
Other mechanisms that might be considered by the CPUC include retail electricity pricing that
reflects real‐time prices or extends time‐of‐use (TOU) pricing and/or energy and demand
charges beyond certain commercial and industrial users. The CPUC Policy and Planning
Division suggested in its 2010 white paper on energy storage that dynamic rates create energy
arbitrage opportunities for customers, and that those customers subject to energy and demand
118
charges would be able to reduce their maximum demand peak with energy storage, saving on
demand charge costs. The white paper recommended that the CPUC require that all IOU
customers move to dynamic retail rates to incentivize more efficient use of power (California
Public Utilities Commission. 2010d, pp.6, 9).46
Several stakeholders to the AB 2514 and 2011 IEPR process have stated that real‐time/ TOU
pricing and variable pricing for fixed transmission and distribution costs are essential to
enabling deployment of cost‐effective energy storage for such applications as peak‐shaving and
load‐shifting. Without visible forward pricing, energy storage operators will have to estimate
when the largest price differentials occur between buying low and selling high, leading
potentially to less efficient use of the resource. To the extent that these pricing changes provide
incentives for retail or behind‐the‐meter storage, some stakeholders prefer that the customers
who benefit from reduced electric bills shoulder the cost of storage, rather than all ratepayers.
Stakeholders also criticize what they point to as a trend of the past decade at the CPUC to
approve tariffs that reduce the KWh cost difference between on and off peak, narrowing rather
than expanding the cost‐value gap for storage options (See Sch??ss, 2011, p.12). 47 Others argue
that dynamic pricing for residential and small business ratepayers falls outside the scope of the
AB 2514 proceeding and has little bearing on grid scale energy storage (See, example given,
California Public Utilities Commission, Docket R10‐12‐007, Comments of the Division of
Ratepayer Advocates). Still others argue that market rules should allow all grid assets,
including energy storage, to compete for wholesale prices that reflect actual grid conditions as
an approach to ensuring cost effectiveness.
4.2.3 Loading Order and Resource Adequacy Requirements
California’s loading order and the CPUC’s Resource Adequacy (RA) program constitute two
other areas which could be modified to facilitate adoption of energy storage technologies,
although some stakeholders express the belief that inclusion in the loading order may not yield
tangible benefits for energy storage. In its 2010 white paper on energy storage, the CPUC Policy
and Planning Division identified both of these areas as ripe for consideration in a CPUC
rulemaking.
46 The CPUC also points out that state law limits how soon the CPUC can implement default
dynamic rates for residential customers to January 1, 2013 at the earliest (California Public
Utilities Commission. 2010d, p. 9, fn. 19).
47 See Sch??ss, Klauss, “En???y Storage for Renewable Integration: Comments to IEPR
Workshop of April 28, 2011,” KSEngineers,
<http://www.energy.ca.gov/2011_energypolicy/documents/2011‐04‐
28_workshop/comments/TN_60630_04‐28‐
11_KSEngineers_Comments_on_the_April_28_2011_IEPR_Works.pdf>, p.12.
119
As discussed in Chapter 2, the loading order for electricity resources consists of first decreasing
electricity demand through energy efficiency and demand response, and second utilizing
renewable energy and distributed generation resources, then clean fossil‐fuel generation, to
meet new generation needs in a cost‐effective manner (California Energy Commission, 2005a,
p.E‐1). Through several Energy Action Plans (EAPs) and Energy Commission IEPRs, the
loading order has served as the basis for recommended energy policies and decisions. Express
inclusion of electric energy storage in the loading order would demonstrate the importance of
energy storage as a component of meeting California’s electricity needs. For example, if the
state energy agencies inserted “electric energy storage resources” before “clean fossil fuel
generation,” they would send a clear message that state policy prefers the use of clean energy
storage before adding additional fossil‐fuel generation. Policy makers could also include
“electric energy storage” dedicated to shaving peak load with “demand response.” These
policies could benefit energy storage deployment by encouraging decision‐makers to examine
thoroughly the costs and benefits of a particular energy storage technology competing for a
contract.
The RA program exists to ensure that sufficient resources are available for California ISO to
maintain safe and reliable operation of the grid in real time, and to provide appropriate
incentives for the siting and construction of new resources needed to maintain future reliability.
The CPUC enforces RA requirements through required monthly filings by LSEs to demonstrate
sufficient procurement to meet its total load forecast plus 15 percent reserve, and through
annual RA proceedings to review and refine the program. Under AB 2514, energy storage “may
be used to meet the resource adequacy requirements established for a load‐serving entity
pursuant to [Public Utilities Code] Section 380 if it meets applicable standards” (Assembly Bill
2514 [Skinner, Chapter 469, Statutes 2010], Sec. 2836.4(a)). One stakeholder to the AB 2514
proceeding points out that, given this statutory provision, “determining precisely how LSEs
may gain RA credit for [electric energy storage] investments seems to be a productive effort for
the Commission to engage in to further the attractiveness of [electric energy storage]
investments” (California Public Utilities Commission, Docket R10‐12‐007, Comments of
Western Power Trading Forum). Thus, the CPUC may need to develop a methodology to
determine a RA value for applicable energy storage technologies, and refine its standards for
the program, for more energy storage systems to qualify (California Public Utilities
Commission, 2010d, p.8). If the CPUC takes these steps, both IOUs and POUs will have
additional incentive to include more energy storage in their resource procurement strategies.
4.2.4 Financial Incentives
Several types of government‐sponsored financial incentives may be extended or expanded to
drive energy storage development and deployment. These include: loan guarantees, tax credits,
120
cash grants, and state incentive programs such as the Self‐Generation Incentive Program (SGIP)
and California Solar Initiative (CSI). As with procurement targets, some stakeholders criticize
these types of incentives as “subsidies” that skew the market and conflict with other policy
priorities. Other stakeholders want to ensure that any proposed tax incentives are not subject to
frequent legislative renewal that could inject significant uncertainty and discourage investment.
4.2.4.1 Federal Programs
Currently, federal support of renewable energy investments may occur via the renewable
energy production tax credit (PTC), the federal business investment tax credit (ITC), the § 1603
cash grant in‐lieu of tax credit, the manufacturing investment tax credit, and the new Clean
Renewable Energy Bond (CREB) and Qualified Energy Conservation Bond (QECB) programs. A
brief description of each program follows. At this time, none of these provide support for
electric energy storage technologies. The lack of programmatic support thru tax credits is the
reason for the reliance by energy storage developers and investors on either DOE funding,
discussed below, or on private investment.
4.2.4.1.1 Production Tax Credit
A Production Tax Credit PTC) is available to renewable energy producers such as wind,
biomass, geothermal, small irrigation power, municipal solid waste, hydropower, and marine
and hydrokinetic, the PTC spreads credit over a ten year period based on a formula of cents per
kilowatt hour of electricity produced by the facility. (Internal Revenue Code, 26 U.S.C. Sec. 45).
The American Recovery and Reinvestment Act (Recovery Act) extended PTC availability for
certain facilities placed in service by January 2013, and for other placed in service before
January 2014. The PTC is unlikely to apply to energy storage based on the requirement that the
electricity credited be “produced by” the taxpayer, suggesting a pure generation function. (See
ibid., 26 U.S.C. § 45(a)(2)(A)).
4.2.4.1.2 Investment Tax Credit
An Investment Tax Credit (ITC_ allows a qualifying facility to collect a single federal income tax
credit against its tax basis equal to 30 percent of its cost, usually claimed when the facility is
placed into service. (Internal Revenue Code, 26 U.S.C. Sec. 48). Qualifying facilities include
those that provide on‐site and distributed generation, including fuel cell, solar, limited
geothermal, combined heat and power, and small wind energy. Several bills introduced last
year in the 111th Congress would have created a similar ITC for energy storage systems. For
example, Senate Bill 2617, the “STORAGE 2010 ACT,” provided up to $1.5 billion in tax credits
for storage projects connected to the U.S. electric grid, in support of integrating renewable
energy resources and facilitating smart grid development. Similar legislation to the STORAGE
121
ACT, if enacted at the federal level, would provide a substantial injection of up‐front financial
support to energy storage system development.
4.2.4.1.3 § 1603 Cash Grant In‐Lieu of Investment Tax Credit
Section 1603 of the Recovery Act created an attractive alternative to the ITC in the form of a cash
grant administered by the U.S. Treasury Department. Recently extended to include properties
placed in service in 2009, 2010, or 2011, or placed in service after 2011 if construction began
earlier, the cash grant in lieu of the ITC provides payments to qualified applicants in an amount
generally equal to 10 to 30 percent of the basis of the property, depending on the property type
(U.S. Treasury Department, 2011). The Treasury Department has not issued an estimate of the
grants made under Section 1603 in 2010 for renewable energy integrated storage, but it is
believed to be small since generally investor and public owned utilities are not eligible to
benefit from this section of the Internal Revenue Code. Like the storage‐specific ITC, a
legislative act could extend the Section 1603 cash grant timeframe and include stand‐alone
energy storage properties.
4.2.4.1.4 Advanced Energy Manufacturing Investment Tax Credit:
This tax credit, created by the Recovery Act to support domestic manufacturing facilities for
advanced energy projects, applies directly to energy storage systems, although only the limited
subset used in electric or hybrid‐electric vehicles (American Recovery and Reinvestment Act of
2009, Div. B, Tit. I, Subtit. D, Sec. 1302). The 30 percent credit applies toward the cost of the
manufacturing facility itself, in contrast to the PTC and ITC, which credit the energy generation
facility. The application period for the Advanced Energy Manufacturing Investment Tax Credit
has closed and would require an Act of Congress for renewal.
4.2.4.1.5 New Clean Renewable Energy Bond (CREB) and Qualified Energy Conservation Bond
(QECB)
CREBs and QECBs are both forms of tax credit bonds made available to public entities, such as
electric cooperatives, public power systems, and state, local, or tribal governments to finance
renewable energy projects. Like the manufacturing investment tax credit, these programs have
not been renewed by Congress and are not accepting applications. To serve as a driver for
energy storage system deployment by public entities, including POUs, Congress would have to
renew the CREBs and/or QECBs and expand the qualifying technologies to include energy
storage technologies.
122
4.2.4.2 California Programs
Programs that offer financial incentives at the state level include the Self‐Generation Incentive
Program (SGIP) and the California Solar Initiative (CSI).
4.2.4.2.1 Self‐Generation Incentive Program
As discussed in Chapter 2, the CPUC administers the SGIP to promote the installation of
various distributed energy systems on the customer side of the meter. “Qualified advanced
energy storage systems” meeting certain technical parameters and coupled with eligible wind
and fuel cell technologies are eligible to receive financial incentives under the program
(California Public Utilities Commission, 2008, p.1). Assembly Bill 1150 (V. Manuel Pérez,
Chapter 310, Statutes of 2011) authorized funding for the SGIP through 2014 and opens the
program to energy storage systems.48
4.2.4.2.2 California Solar Initiative
The CSI is part of California’s umbrella effort to install 3,000 MW of solar statewide under the
Go Solar California program. With a budget of $2.167 billion over 10 years and a goal of
reaching 1,940 MW of installed solar capacity by the end of 2016 (California Public Utilities
Commission – (c)), the CSI may have an indirect effect on energy storage by heightening the
need for complementary technologies like battery storage that can be coupled with distributed
solar. The CSI does apply to solar thermal technologies that generate electricity or displace
electricity usage (California Public Utilities Commission – (d)). With the passage of AB 1150,
end‐users may have additional incentive to consider an energy storage complement to their
solar system.
4.2.5 Targeted Research and Development Funding
Government financing support for energy storage system deployment in the form of
government loans, grants, or tax credits is demonstrably limited. After years of extremely
limited federal funding, DOE’s Energy Storage Program experienced a significant boost when
ARRA provided $185 million in federal matching funds to support energy storage projects with
a total value of $772 million.
48 See Legislative Analysis, available at: http://www.leginfo.ca.gov/pub/11‐12/bill/asm/ab_1101‐
1150/ab_1150_cfa_20110525_101450_asm_comm.html (created Spring 2011, last accessed July 8,
2011).
123
These projects will generate 537 MW of new storage systems to be added to the grid when
completed. A breakdown of ARRA‐funded projects, organized by project category, is shown in
Table 3 AARA Funded Energy Storage Technology Demonstration Projects.49
The DOE had an energy storage R and D budget of $20M in FY2011, and the House Appropriations
Committee in June 2011 approved $20M in FY2012 which was $37,000,000 below the President’s budget
request.
Table 3: ARRA Funded Energy Storage Technology Demonstration Projects
(SNL ESS 2010)
Energy Storage Projects in California
For various ARRA funded Energy Storage projects in California, the DOE provided $151.5
million in ARRA funds. The Energy Commission, through its PIER program, is providing $6.44
million as cost‐share funds for these projects. The recipients are contributing $477.5 million out
of total cost of $635.46 million for these projects. Also, the PIER program has provided funding
for several Energy Storage projects over many years. In 2010‐2011, the PIER program has
provided $8.97 million to support several non‐ARRA Energy Storage projects. The brief
descriptions of the PIER supported active projects are provided here.
49 DOE Electricity Advisory Committee, Energy Storage Technologies Subcommittee, “Energy
Storage Activities in the United States Electricity Grid”, May 2011, p.3. Available at:
http://www.oe.energy.gov/DocumentsandMedia/FINAL_DOE_Report‐Storage_Activities_5‐1‐
11.pdf
124
American Recovery and Reinvestment Act of 2009 Energy Storage Projects in California
PIER: $6,440,028 Match: $477,496,532 DOE: $151,519,268 Total: $635,455,828
1. Tehachapi Wind Energy Storage Project: Southern California Edison will design and
build a 32 MWh lithium‐ion battery system and smart inverter, and connect it to their Monolith
Substation near the Tehachapi Wind Resource Area. This project will demonstrate the ability of
the battery storage system to enhance grid operations and integrate intermittent wind power in
a remote, transmission‐constrained area. The installed system will help achieve utility load
shifting, increased dispatchability of wind generation, and enhanced ramp rate control to
minimize the need for fossil fuel‐powered back‐up generator operation. The project will
leverage the data and results of an ongoing PIER‐funded study on the Antelope‐Bailey 66 kV
system. The ongoing study aims to determine the ways in which energy storage can address
wind integration issues to help meet Californiaʹs renewable energy goals. One of the key early
findings of this study was that a 32 MWh energy storage device located at the Monolith
Substation can prevent overloaded transmission lines.
PIER: $1,000,000 Total: $53,510,209
2. Wind Firming EnergyFarm: Primus Power Corporation will work with the United
States Department of Energy, Sandia National Laboratory, Pacific Gas and Electric Company (P
G and E), and Modesto Irrigation District to develop, field test, and install and evaluate a 25
MW/75 MWh grid‐connected Zinc‐based flow battery energy storage system. The project will
provide a low‐cost energy storage system with a footprint consistent with or smaller than other
competing technologies and demonstrate primary and secondary applications including:
renewable firming, strategic local peak shaving, automated load shifting, and ancillary services.
PIER: $1,000,000 Total: $46,700,000
3. Solid State Batteries for Grid‐Scale Energy Storage: Seeo Inc. (Seeo) will develop the
first ever large‐scale or grid‐scale prototype of a new class of advanced lithium ion rechargeable
batteries, with unprecedented safety, lifetime, energy density, and cost. The primary focus of
this project will be the development and deployment of a 25kWh prototype battery system
based on Seeoʹs proprietary nanostructured polymer electrolytes. This will validate the
transformational performance advantages of this technology for use in grid‐tied energy storage
applications. In particular, Seeo seeks to address the utility market needs for clean energy
systems, which envision small (<100kW) distributed energy storage systems alongside pad‐
mounted and pole‐mounted transformers, and grid‐connected electric vehicle systems.
PIER: $600,000 Total: $12,392,122
125
4. Fault Current Limiting Superconducting Transformer: Southern California Edison
(SCE), working with Waukesha Electric Systems, Inc. (Waukesha) will design, develop,
fabricate, and install a Smart Grid compatible fault current limiting superconducting
transformer on SCEʹs utility host site. The 28 MVA, three‐phase, medium power utility fault
current limiting transformer twill be placed within SCEʹs MacArthur Substation, which is
located in Newport Beach, California and within the project area of SCEʹs Irvine Smart Grid
Demonstration Project. By incorporating fault current limiting capability, the transformer is
better able to handle fault currents and represents a cost and space efficient means to bring fault
current limiting capability into the substation without adding a separate device, and ultimately
improves the grid performance and reliability.
PIER: $767,134 Total: $21,406,194
5. Premium Power Distributed Energy Storage Systems Demonstration: Sacramento
Municipal Utility District (SMUD) will install and demonstrate a fleet of two Premium Power
Corporation’s Zinc Bromine Flow Battery energy storage systems in Sacramento, California, one
at the SMUD Headquarters (East City Substation) serving the SMUD campus micro‐grid, and
one at a substation serving the nearby Anatolia III SolarSmart Homes community development.
The SMUD Headquarters storage system will explore its utility in improving micro‐grid
operations; emergency operations, including campus islanding; and augmenting peak period
campus operation with non‐peak generated electricity. The storage system at SMUDʹs Anatolia‐
Chrysanthy substation will be integrated with the Anatolia III SolarSmart Homes community,
which will have 600 homes totaling 1.2 MW of photovoltaic (PV) generating capacity. The two
storage systems will be controlled from a common control system at the SMUD headquarters
site to demonstrate fleet control of multiple distributed storage devices. Over a four year period,
this project and the technology validation it supports will demonstrate competitively‐priced,
multi‐megawatt, long‐duration advanced batteries for utility grid applications and validate the
potential penetration of zinc bromine flow batteries, particularly in PV and micro‐grid
applications by demonstrating multiple use cases.
PIER: $227,000 Total: $5,417,123
6. Utility‐Scale Flywheel Energy Storage Demonstration: Amber Kinetics, Inc., will
demonstrate a prototype utility‐scale flywheel energy storage system employing technology
advances in composite flywheel rotor materials, magnetic bearing systems, and high efficiency
motor‐generators. These new technologies, when integrated into a flywheel system, can prove
126
that flywheel energy storage can be competitive with pumped hydro in terms of cost and
efficiency. The goal of the project is to clearly demonstrate the economical and technical
viability of bulk flywheel energy storage and renewable energy integration for the electric grid.
PIER: $369,466 Total: $10,003,015
7. Flow Battery Solution to Smart Grid Renewable Energy Applications: EnerVault
Corporation (EnerVault) will demonstrate the commercial viability of EnerVaultʹs novel iron‐
chromium redox flow Battery Energy Storage System (BESS). This demonstration comprises of
integrating EnerVaultʹs Vault‐20 BESS (250kW, 1MWh) with an intermittent renewable energy
source ‐ a dual‐axis photovoltaic (PV) system. The 36 month project will culminate in the
deployment of a Vault‐20 Beta system in conjunction with a 150kW PV system at a site in
Californiaʹs Central Valley. Additionally, the operating results will be analyzed and compared
to the baseline for final quantification of benefits and operating costs. The capital costs,
operating costs, and benefits will be used to determine a Total Cost of Ownership.
PIER: $476,428 Total: $9,528,567
8. Smart Grid Demonstration Project: The Los Angeles Department of Water and Power
(LADWP) will develop and demonstrate a network of smart grid technologies which also
involves battery energy storage systems for electric vehicles. The project goal is to facilitate the
establishment of protocols and standards in the Smart Grid Demonstration Project that allow
for the measurement and validation of energy savings and fossil fuel emissions reductions
associated with an Electric Vehicle Program by specifying, acquiring, and installing a
sufficiently‐sized Electric Vehicle demonstration network.
PIER: $1,000,000 Total: $120,560,000
9. Advanced Underground Compressed Air Energy Storage Demonstration Project:
Pacific Gas and Electric Company (P G and E) will design, build and demonstrate the world’s
first advanced, “second generation” Compressed Air Energy Storage (CAES) design system that
requires less fuel, utilizes standardized less expensive turbo‐machinery, and captures the waste
heat from the compression cycle. This “second generation” design uses readily available proven
turbo‐machinery that will result in lower capital and operating costs than “first generation
designs”. The plant design will also include the option for future use of thermal storage to test
the potential of adiabatic CAES, a ʺthird generationʺ technology that would completely
eliminate the use of fuel for a CAES plant. The project will use depleted gas fields, located
within P G and E’s service territory, for compressed air energy storage.
PIER: $1,000,000 Total: $355,938,600
127
Other PIER Funded Energy Storage Projects
10. Grid‐Saver Fast Energy Storage Demonstration: Transportation Power, Inc.,
Escondido, CA will demonstrate a new, low cost fast energy storage Lithium‐ion (Li‐ion)
battery technology that can help facilitate acceptance of utility‐scale renewable energy projects
in California and nationwide. Supported by team members Evaira and General Atomics,
Transportation Power will evaluate the feasibility of designing a low cost fast energy storage
system based on innovative design concepts and, following analysis and validation of the
concepts, proceed to prototype demonstration to provide further validation and establish a
basis for widespread commercial adoption of such a system. The system approach, trademarked
as Grid‐Saver, is based on modular building blocks with advanced system integration and
control methods. The Grid‐Saver program aims to demonstrate the viability of the concept by
building, testing, and deploying a 5 MW peak power for periods of up to 10‐15 minutes fast
energy system comprised of interchangeable lithium battery modules and high‐power inverter
modules as building blocks that can be produced at relatively low cost and integrated with
advanced control technology. This also includes validating the performance and establishing a
basis for widespread commercial adoption of such a system.
PIER: $2,000,000 Match: $520,004 Total: $2,520,004
11. Grid‐interactive Photovoltaic System with DC‐link Battery Storage Integration:
Satcon Technology Corporation, Boston, MA will demonstrate how high bandwidth automatic
voltage control can be applied to mitigate intermittency in power plant output and load
induced voltage variations on the utility feeder. It also aims to develop electric energy storage
components and a system specifically designed and optimized for grid‐tied photovoltaic (PV)
applications. To achieve these goals, the project will demonstrate a hybrid electric generation
system comprised of a 500 kW PV array, 500 kW grid‐connected PV inverter, and integration of
500 kWh state‐of‐the‐art battery technology, to further enable utility‐scale renewable energy
sources. The 500 kW system is a standard building block for multi‐megawatt configurations and
scalable to target a utility scale plant of 10 MW. The combined inverter and battery storage
system will help mitigate the destabilizing effects of intermittency, thereby supplementing the
load demand and limiting output power ramp rates to levels that are compatible with other
utility generation resources. The demonstration of advanced inverter and integrated solar PV
storage system will be part of the Sacramento Municipal Utility Districtʹs solar highway project.
PIER: $1,972,211 Match: $1,345,332 Total: $3,317,543
128
12. Evaluation and Optimization of Concentrated Solar Power Coupled with Thermal
Energy Storage: KEMA, Inc. will perform transient thermodynamic modeling of nine different
configurations of concentrated solar power‐thermal energy storage and analyze optimization
and integration of different models for dispatch. This also includes simulation of market
outcomes in market simulation and economics models. Also, KEMA will bring the
methodology and outcomes of the modeling to the market through dedicated technology
transfer activities. The potential benefits include:
• Reduced ancillary services requirements and real time dispatch costs at a system
level as a result of improved and managed CSP generation variability and ramping
behavior, as well as reduced emissions from substituting for stand‐by convention
generation for these services;
• Improved system reliability as measured by system dynamic performance (area
control error, frequency) and system ability to withstand events such as wind
ramping, cloudiness and unit trips;
• Improved system reliability in terms of frequency response to disturbances that are
improved as a result of CSP inertial and governor response. (As compared with the
lack of same from PV and wind resources);
• Enhanced understanding of the economics ‐ cost benefit comparison ‐ of thermal
storage and CSP reconfiguration to avoid gas co‐firing and to allow the provision of
ancillary services.
PIER: $447,642 Match: $173,989 Total: $621,631
13. Energy Storage Demonstration and Compressed Air Energy Storage Study: Pacific
Gas and Electric Company will demonstrate sodium‐sulfur battery in grid application and
study Compressed Air Energy Storage as a viable energy storage system for integrating
renewables in California.
PIER: $2,800,000 Total: $2,800,000
14. Using High Speed Computing to Estimate the Amount of Energy Storage and
Automated Demand Response Needed to Support California’s RPS: Lawrence Livermore
National Laboratory will estimate the amount of energy storage and automated demand
response needed to support California’s Renewable Portfolio Standard using high speed
computing capability of super computer at Lawrence Livermore National Laboratory.
PIER: $1,750,000 Total: $1,750,000
129
4.3 Framework for an Application and Scenario-Based Assessment
Thus far, this report has summarized some of the most important factors for California policy
makers to consider in the course of determining a vision for energy storage in California for
2020: the commercial status and cost of energy storage technologies; the policy framework
guiding the state’s energy and environmental goals; the anticipated needs of the future
electricity grid, and the various policy, regulatory, and market drivers that might be
implemented to facilitate increased deployment of storage technologies by 2020.
The next step – to develop scenarios for how energy storage maybe applied throughout
California’s electric power system and to estimate the costs and benefits of such a vision as
compared to a scenario without deployment of energy storage – requires an analytical
framework with which to assess the costs and benefits of energy storage deployment and
prioritization of applications and technology areas for further study or incentive development.
This framework will serve as a precursor to the development of specific targets and
identification of milestones and actions necessary for development and deployment of energy
storage technologies in California.
The remainder of this report analyzes the deployment of energy storage technologies using an
application and scenario‐based approach. Several major efforts to analyze energy storage in the
U.S. and California, including by the U.S. DOE, Sandia National Laboratories, EPRI/E3, and
SCE, have utilized applications as a starting place to assess storage value. In its white paper,
SCE explains that “focusing on applications defines ‘the problem’ before assessing ‘the
solution’” and enables a “complete strategic assessment of storage” that can reflect “all
operational uses on the electric value chain” (SCE White Paper, 2010, p.15). The Market and
Benefits Guide released by Sandia National Laboratories is careful to distinguish “applications”
from “benefits,” an application being essentially a use while a benefit “connotes a value” which
can be monetary (for example, arbitrage value) or societal (for example reduced air emissions
from generation) (Eyer, 2010, p.2).
Using three exemplary applications, these analyses will attempt to define the benefits that
energy storage technologies bring to the application, associated costs, and likely outcomes for
energy storage deployment under two different scenarios. The first scenario is a baseline or
“business as usual” assessment based on the current regulatory and market rules, current
technical status and fit of storage technologies to the application, and current costs and benefits
of meeting the application with storage. The second scenario is an “accelerated” assessment of
what policy, regulatory, or other changes, including those related to technology cost, would be
needed to spur an increase in energy storage deployment for the application by 2020. When
130
possible, the application and scenario‐based assessment will include a high level comparison of
alternative, non‐storage solutions for the identified applications. These scenarios, based on key
applications, will offer possible pathways for California to pursue, and will identify necessary
targets and milestones that must be met along the way.
131
CHAPTER 5: Applications and Scenarios Analyses for
Energy Storage in California
Introduction:
Chapter 4 described the approach to the applications analyses as well as the reasoning behind
this framework. This chapter examines each of the three exemplary applications, noting that
they do not by themselves represent the full range of the costs, benefits, and value of energy
storage. This chapter defines each application, describes the ongoing technical and policy
barriers facing the various energy storage technologies that may be appropriate for these
applications, and recommends next steps for policy makers and stakeholders.
5.1 Exemplary Application #1: Area and Frequency Regulation
The National Renewable Energy Laboratory (NREL)has identified the emergence of
deregulated energy markets, such as those for high‐value ancillary services operations, to be
one of several primary factors motivating the surge of renewed interest in energy storage
(Denholm, 2010, p.9). These services require fast response time and often limited actual energy
delivery – both needs for which many energy storage technologies are well‐suited. In California
ISO’s service area and nationwide, regulation services are the highest‐compensated energy
market (See Electric Power Research Institute, 2010a, pp.2‐10 – 2‐11; Denholm, 2010, p.12).
Some stakeholders have noted that market prices for frequency regulation can vary significantly
by hour, depending on market conditions, and that grid operators may only have a limited need
for regulation given its function to manage the forecast uncertainty and variability over time
intervals often less than five minutes. In addition, available regulation‐certified resources
would compete with energy storage technologies to provide the service. However, studies
suggest that area and frequency regulation service will continue to be an important factor to
ensure grid reliability under higher renewables penetration. And based on its importance to the
grid and its measurable monetary value in California, area and frequency regulation presents a
use case for assessing the benefits and costs of baseline and accelerated deployment of energy
storage. Note that other studies categorize area or frequency regulation within the broader
application category of wholesale energy services (EPRI, 2010a) or simply as one of many
ancillary service applications (Sandia National Laboratories, 2010). This assessment will
consider area and frequency regulation as a standalone application.
132
5.1.1 Area and Frequency Regulation: Definition and Application
Frequency regulation service is the injection or withdrawal of real power by facilities capable of
responding appropriately to a transmission system’s frequency deviations or interchange power
imbalance. When generation dispatch does not equal actual load and losses on a moment‐by‐
moment basis, the imbalance will result in the grid’s frequency deviating from the standard (60
Hertz). Minor frequency deviations affect energy consuming devices; major deviations cause
generation and transmission equipment to separate from the grid, in the worst case leading to a
cascading blackout. Frequency regulation service can prevent these adverse consequences by
rapidly correcting deviations in the transmission system’s frequency to bring it within the
acceptable range (FERC, 2011).
Energy storage technologies are well suited to resolve momentary differences between supply
and demand, as well as fluctuations in grid frequency. The current frequency regulation
method uses gas or steam turbines to balance constantly shifting load fluctuations by
periodically adjusting generation in response to a signal from the system operator. Electrical
plants use spinning reserves to quickly ramp up to full output when another generator goes
offline. However, these generators must run constantly at reduced load to assure power
quality, which results in increased greenhouse gas and criteria pollutant emissions.
The use of energy storage technology has the capability to be faster than regulation by a gas or
steam turbine. This faster response time can minimize momentary electricity interruptions,
particularly at the distribution level, which are more costly than sustained interruptions
(LaCommare and Eto, 2004). Energy storage technologies can vary output rapidly, changing
from no output to full output within seconds. To optimize efficiency and response time, energy
storage technologies used for frequency regulation must be able to communicate with the grid
quickly and efficiently. Using storage for frequency regulation can reduce the consequences of
rare but costly power interruptions at high‐tech industrial and commercial facilities. Taking
advantage of the capabilities of faster‐ramping resources can improve the operational and
economic efficiency of the transmission system and has the potential to lower costs to
consumers in the organized wholesale energy markets.
The performance targets of energy storage technologies used for frequency regulation
emphasize the importance of discharge duration, response time, and roundtrip efficiency
metrics. If an energy storage technology is able to meet these targets for frequency regulation in
a cost‐effective way, the electric power industry is likely to adopt the technology to assist with
the recovery from momentary disturbances.
133
Table 4: Area and Frequency Regulation Performance Targets for Energy Storage
Metric Target Supporting Information
Size 1 ‐ 100 MW
Discharge Duration 15 minutes to 2 hours Storage technologies should have symmetric
charge and discharge rates for this
application.
Response Time < 1 second Since this application is intended to
reconcile momentary differences, the storage
technology must be able to respond to grid
signals as fast as possible.
Roundtrip
Efficiency
75 ‐ 90% Roundtrip efficiency is the efficiency
measured at the transformer of the energy
output divided by the energy input.
(Nexight 2010a, EPRI 2010a, p.4-24)
In addition to technical performance targets, system cost and lifetime are also key metrics that
energy storage technologies must meet to offer competitive solutions. Similarly, to be
competitive, energy storage technologies must be able to achieve system lifetimes of at least 10
years, based on 4,500 ‐ 7,000 cycles per year (Nexight, 2010b, p. 17).
5.1.2 Area and Frequency Regulation: Technology Status and Market Fit
In a recent study, EPRI evaluated various energy storage technologies for grid applications
(EPRI). The organization identified flywheels, lithium ion batteries, and lead‐acid batteries as
promising candidates to meet frequency regulation performance targets. These systems may
also be applicable for smoothing intermittency of wind and photovoltaic power generation as
well as power quality applications. All of these systems are modular and can be configured in
both smaller and larger sizes.
The figures below represent published ranges estimating the total capital installed cost of
“current” turnkey systems based on 2010 inputs from vendors and system integrators (EPRI
2010a, p. 4‐24). Included are the costs of power electronics if applicable, and all costs for
installation, step‐up transformer, and grid interconnection to utility standards. The report also
assumes that smart‐grid communication and controls are included. In addition, the study noted
that pumped hydro, CAES, and flow battery systems may be applicable for frequency
regulation and other ancillary services. Analysts face difficulty calculating these values,
134
however, due to the overlapping applications for the various technologies and their operational
uses, which may also be subject to different value measures and regulatory rules. Note that
advanced lead acid capital costs are reported on a “rated” MWh delivered per cycle basis.
Table 5: Energy Storage Options for Frequency Regulation
Technology Maturity Capacity
(MWh)
Power
(MW)
Duration
(hrs)
Efficiency
(%)
Total
Cost
($/kW)
Cost
($/kWh)
Flywheel Demo 5 20 0.25 85 – 87 1950‐2200 7800‐8800
Li‐Ion Demo 0.25 – 25 1 – 100 0.25 ‐ 1 87 – 92 1085‐1550 4340‐6200
Advanced
Lead Acid
Demo 0.25 – 50 1 – 100 0.25 ‐ 1 75 – 90 950‐1590 2770‐3800
(EPRI, 2010a, p.4-24)
Typical flywheel applications include power quality and uninterruptible power supply (UPS)
uses. Manufacturers are positioning flywheels to provide frequency regulation services because
they are fast responding and efficient. Analysis of such flywheel services have shown them to
offer system benefits such as reducing carbon dioxide emissions and avoided cycling of large
fossil power systems. Durability and life‐cycle cost data are unavailable at this time.
Purchasers should consider the cost of flywheel replacements over the book life in a life‐cycle
analysis.
In addition to the options identified in Table 5 Energy Storage Options for Frequency
Regulation, , ten adjustable speed pumped storage stations are in operation in Japan and
Europe and are providing frequency regulation services with the following characteristics that
translate to the U.S. site applications presented in Table 6 Adjustable Speed Pumped Storage
Characteristics.
135
Table 6: Adjustable Speed Pumped Storage Characteristics
Technology Maturity Capacity
(MWh)
Power
(MW)
Duration
(hrs)
Efficiency
(%)
Total
Cost
($/kW)
Cost
($/kWh)
Adjustable
Speed
pumped
Storage
Advanced 20,000 1,000 10 80‐82 1,500‐
3,500
150‐350
(HDR, 2010; MWH, 2009)
Given their attractive cycle life and compactness, in addition to high ac‐to‐ac efficiency, electric
power companies are seriously considering Li‐ion batteries for several utility grid‐support
applications, such as community energy storage, transportable systems for grid‐support,
commercial end‐user energy management, home back‐up energy management systems,
frequency regulation, and wind and photovoltaic smoothing. Early system trial applications are
underway using small 5‐ to 10‐kW/20‐kWh distributed systems and large 1‐MW/15‐minute fast‐
responding systems for frequency regulation. In addition, several Li‐ion developers are
demonstrating 1‐MW/250‐kWh and 2‐MW systems. There are several different types of Li‐ion
chemistries, each with their own cost and performance characteristics. Data shown is the
average of currently available systems. Durability and life‐cycle cost data are unavailable at this
time. Battery life‐cycle costs can vary considerably by supplier, depending on the design basis
duty cycle and design life and the associated cost of building or container structures for
housing. Purchasers should consider battery replacement costs, otherwise not shown, as a
variable O and M expense in any life‐cycle analysis. Purchasers should also consider the cost of
battery replacements over the book life in a life‐cycle analysis.
There have been few utility applications for lead‐acid batteries due to their relatively heavy
weight, large bulk, cycle‐life limitations and perceived reliability issues (stemming from
maintenance requirements). Work to improve lead‐acid battery technology and materials
continues. Innovation in materials is improving cycle life and durability, and manufacturers are
developing several advanced lead‐acid technologies that are in the pre‐commercial and early
deployment phase. Researchers are developing these systems for peak shaving, frequency
regulation, wind integration, photovoltaic smoothing, and automotive applications. There are
several advanced lead‐acid technologies. Data shown represent an average of currently
available systems. Battery life‐cycle costs can vary considerably by supplier, depending on the
design basis duty cycle and design life and the associated cost of building/container structures
for housing. Purchasers should consider battery replacement costs, otherwise not shown, as a
136
variable O and M expense in any life‐cycle analysis. This consideration is especially important
with this application, because it involves around a hundred thousand shallow cycles per year.
5.1.3 Area and Frequency Regulation: Technology Potential
Modern mechanical storage applications are capable of providing regulation service, as
flywheels and adjustable speed pumped storage technologies demonstrate. Flywheels
generally produce 100 to 2,000 kW for periods of time ranging from seconds up to 15 minutes
and are thus best suited for high‐power, low‐energy applications such as grid angular stability
and voltage support. Areas for future work include materials development, cost reductions,
and improved manufacture techniques (Gyuk and Eckroad, 2003). In addition, scale‐up of
flywheels could potentially reduce cost through reduction of the required balance‐of‐plant
through the reduction of components (fewer in number but larger in size). Ultimately,
purchasers will need field demonstrations prior to wide‐scale acceptance of flywheel storage for
grid applications. Adjustable speed pumped storage stations are a proven technology, but the
incremental revenues associated with frequency regulation markets must offset the incremental
cost of the stations.
Batteries can store and release power across a broad range of time scales, smoothing rapid
fluctuations in the output of renewable generators or mitigating daily variability. Despite their
operational flexibility and potential to provide a wide range of ancillary services, batteries have
yet to attain high penetration at grid scale. High cost remains a deterrent, as does a lack of
demonstration. The DOE has begun to provide partial funding for storage projects; including a
planned 20‐MW Li‐ion battery system in New York that will provide frequency regulation
services and help integrate increased wind and solar capacity. If the cost of batteries drops due
to technical innovation and increased production volume, use of batteries for grid applications,
such as regulation, could expand substantially (Fertig and Wagner, 2010).
Engineers continue to work to improve lead‐acid battery technology and materials. Innovation
in materials is improving cycle life and durability, and manufacturers are developing several
advanced lead‐acid technologies at the pre‐commercial and early deployment phase.
Manufacturers are developing these systems for frequency regulation, as well as other
applications. Some advanced lead batteries have “supercapacitor‐like” features that give them
fast response similar to flywheels and supercapacitors. Analysts anticipate that advanced lead‐
acid systems from a number of companies will be in early field trial demonstrations by 2011–
2012.
For essentially all energy storage systems, manufacturers will likely reduce future costs after
early demonstrations become proven and validated, products become more standardized, and
initial engineering costs have been removed. For the electrochemical energy storage
137
technologies, including batteries and capacitors, analysts expect high volume production of
cells to lead to cost reductions. The potential growth of electric and hybrid electric vehicles
could result in the demand for high volume production and thus reduced cell costs for all
applications. Power companies and grid operators would still need to integrate these cells, such
as Li‐ion, into systems, complete with thermal management, electronic controls, power
conversion, and inter‐connections, appropriate for frequency regulation and other applications.
5.1.4 Area and Frequency Regulation: Regulatory Status and Policy Changes
Area and Frequency Regulation may be one of the few areas in which the baseline scenario,
which includes changes underway and possible changes under consideration, produces an
optimistic picture for the economic and technological feasibility of energy storage. Yet there are
many additional changes to regulatory policy that could accelerate the deployment of storage
for this application and other ancillary services. These are included in the accelerated scenario.
5.1.4.1 Area and Frequency Regulation: Baseline Characteristics (1)
Grid operators have installed a number of energy storage projects that are providing regulation
services in various places across the country and the globe (See Lin, 2011, pp.6‐8). The baseline
scenario for energy storage thus already includes these demonstrations of viability. In addition,
substantial changes have recently affected the regulation services market, and others are
underway or under consideration. As described in Chapter 2, California ISO recently approved
changes to its tariff for ancillary services, altering the definition of an ancillary service resource
to include non‐generating resources, reducing the continuous energy requirement for day
ahead and real time regulation up and down, and reducing the minimum rated capacity
requirement to 500 kW (California Independent System Operator, 2010c). The creation of a
Regulation Energy Management (REM) option further enables the participation of Limited
Energy Storage Resource (LESR) technologies in California’s regulation market, once the
software becomes operational (See California Independent System Operator, Board of
Governors. 2011d).
Additional changes under consideration include possible implementation of a pay‐for‐
performance or “mileage” payment metric under either California’s ISO’s Integration of
Renewables, Market and Product Review process (California Independent System Operator,
2011e, p.17) or FERC’s proposed rule for Frequency Regulation Compensation in the Wholesale
Power Markets (Federal Energy Regulatory Commission, 2011, p.124), and a possible
determination by FERC to allow classification of energy storage devices within more than one
asset class or as a new asset class for accounting and reporting purposes.
If FERC or California ISO (or both) employ any of these additional changes, the price of energy
storage for regulation is anticipated to increase beyond its current level. Moreover, utility and
138
third‐party investments in energy storage demonstration projects, continued development of
improved electric vehicle technology and secondary use options, and the planned
implementation of California’s cap and trade program, are likely to reduce technology costs and
increase the costs of competing, fossil‐fuel resources that are kept online for regulation needs.
Thus, even if the CPUC determines not to set any type of energy storage procurement target
under AB 2514, energy storage deployment for regulation will experience fewer barriers and
may increase as a result.
5.1.4.2 Area and Frequency Regulation: Accelerated Characteristics (2)
California Energy Storage Alliance (CESA) has identified additional refinements that California
ISO could implement to remove barriers to energy storage participation in the regulation
market and to compensate storage for its additional value‐adds. CESA supports a pay‐for‐
performance or “mileage” compensation method that rewards the accuracy and speed with
which a system responds to the regulation signals (Lin, 2011, p.11). The additional value could
reflect the benefits of increased grid reliability because of faster and more accurate regulation,
as well as the ability of fast‐responding resources to reduce the overall amount of reserves
needed (Denholm, 2010, 12). Higher‐value compensation for fast and accurate regulation
services may emerge as part of the baseline scenario, based on its consideration at FERC and
California ISO, although some stakeholders dispute the need and prefer lower procurement
targets accompanied by reduced planning reserves. This change is critical, however, to any
accelerated deployment of storage for regulation in California.
CESA has identified further changes to rules and procedures at California ISO that should be
made to remove barriers to energy storage participation in regulation. CESA advocates
reducing the continuous energy requirement further from the recently revised 30‐minute real
time and 60‐minute day ahead requirements, as well as modifying California ISO’s Energy
Management System (EMS) to accommodate a negative power dispatch (injection and
withdrawal of energy), and employing a “fast first” dispatch algorithm similar to NY‐ISO’s
regulation tariff (Lin, 2011, pp.11‐12). From a project finance perspective, the extension of long‐
term power purchase agreements (PPAs) to accommodate regulation‐only energy storage
systems, and/or a procurement program similar to resource adequacy at the CPUC, would
equalize the field for regulation‐centric energy storage with conventional generators that
provide regulation services secondary to their PPA‐backed provision of wholesale energy (Ibid.,
p.12). CESA recommends that California ISO work with the CPUC to build regulation capacity
needs into the RA program.
An energy storage procurement target under AB 2514 would strengthen the basis for
accelerated deployment to service California’s regulation needs. As noted in Chapter 2, AB
2514 does not specify what form the targets should take, if any, as a result of the CPUC’s
139
rulemaking. Since LESR for frequency regulation represents the greatest potential value to
ratepayers in a growing market, it would serve as a logical starting place for the regulated
utilities, should the CPUC promulgate a general energy storage procurement target under AB
2514. A more nuanced target, such as some percentage of California ISO’s determined
regulation needs for 33 percent (plus) renewables penetration, or some percentage of utility side
of the meter capacity, would further incentivize energy storage to make a rapid deployment for
this application. However, the CPUC must consider the benefits and costs of such a target,
including potential costs to ratepayers if utilities rate base their energy storage investments.
Another important facet of an accelerated deployment scenario for regulation services would be
for the CPUC to assign a value for environmental or “incidental” benefits of energy storage.
The price assigned to carbon emissions as a result of the anticipated 2012 implementation of the
ARB’s cap‐and‐trade regulations can serve as a baseline for this value. Policy makers can go
further to include a value for other societal benefits, such as reduced reliance on fossil fuel and
increased energy security, reduced criteria air pollutant emissions, and enabling superior
operation of the existing generation fleet (Eyer, 2010, pp.134‐35). Even if policy makers do not
add a price point for these benefits to the valuation equation, they can take these benefits into
account in evaluating whether more accelerated deployment measures will support
environmental and energy‐related policy objectives for the state (and California’s cap‐and‐trade
program may place a price on carbon that could accomplish some of these goals even without
an incidental benefit calculation).
5.1.5 Area and Frequency Regulation: Benefits and Costs
Conventional generators, including hydro, combustion turbine, and combined‐cycle plants, as
well as pumped hydro storage, currently meet most regulation needs (Denholm, 2010, p.12).
According to NREL, as the amount of variable renewable generation increases, so will the need
for “flexibility resources” that can accommodate the increased variability of net load and the
differences between demand and variable supply (Ibid., p.34). California ISO’s ongoing study
of grid needs under 33 percent renewable penetration in 2020 shows up to 1,230 MW of
regulation capacity needs to meet the fall maximum regulation up requirement (California
Independent System Operator, 2011a, Slide 13; California Independent System Operator.
2011h). And a recent study commissioned by the Energy Commission found that energy
storage helped avoid the degradation of the electric system performance without requiring
additional regulation resources beyond the current 400 MW (KEMA, 2010, p.48). While
conventional resources could possibly meet most of California’s regulation needs, the
demonstrated capability of energy storage to outperform conventional resources in this
application merits consideration of the benefits and costs of both scenarios. In addition, the
results of the KEMA study suggest that meeting high‐renewables penetration regulation needs
140
with conventional generation may be problematic, considering that renewable generation will
displace a portion of conventional generation best suited to address ramping rates (See KEMA,
2010, p. 55).
5.1.5.1 Benefits of Energy Storage for Frequency Regulation
Studies have found several distinct benefits to enlisting fast responding energy storage
resources first, or in lieu of, slower, conventional resources for the provision of regulation
services. KEMA, in its research for the Energy Commission, concluded that 3,000 MW of fast‐
acting storage with a two‐hour duration would achieve overall solid frequency performance
under a high‐renewables penetration scenario for 2020, assuming certain control algorithms
were in place (KEMA, 2010, p. 65).50 Strategen Consulting, in an analysis conducted for the
California Energy Storage Alliance, found that “energy storage is a more effective way of
meeting the increasing demand of ancillary services at a lower cost – in both economic and
environmental terms – than [conventional] resources” (Lin, 2011, pp.2‐3). The likely benefits to
utilizing energy storage in this application include reduced overall regulation capacity needs,
reduced reliance on conventional resources and stress on existing generator equipment, and
reduced greenhouse gas emissions. Taking the opposite approach, meeting future regulation
needs without storage would require increased ramping by conventional generators, causing
increased maintenance cost, increased wear and tear and reduced lifetime, postponed de‐
commitment of combustion turbine generators, and increased greenhouse gas and polluting
emissions (KEMA, 2010, p. 72).
The Strategen Consulting white paper presented findings that the fast response rate of a
flywheel, able to quickly increase or decrease its output to match generating output with load,
in comparison to a combined cycle gas turbine, can result in fewer total MW capacity of
regulation that needs to be procured (Lin, 2011, p.3). This is due to the fact that faster ramping
allows a resource to reach its dispatch target sooner and be re‐dispatched more often, as well as
to switch directions quickly. In contrast, slower ramping resources at times can provide
regulation in the wrong direction due to their inability to respond quickly to a signal change in
direction. Faster response time also minimizes costly momentary electricity interruptions,
offering substantial savings to high‐tech industrial and commercial facilities that depend on
uninterrupted power at steady frequency (Nexight Group and Sandia National Laboratories,
2010b, p.16).
The potential to reduce MWs of regulation capacity needs finds support in a 2008 study by
Pacific Northwest National Laboratory (PNNL) for the California Energy Commission, which
50 Performance degraded when KEMA modeled storage systems with a response rate slower
than 5 MW/second. (KEMA, 2010, pp. 63, 76).
141
found that involving fast responding resources in regulation can help reduce California ISO’s
regulation procurement by up to 40 percent (Makarov, 2008, p.37). Specifically, the PNNL
study concluded that an ideal fast responding resource, defined as one with instantaneous
response and unlimited energy, is about 1.7 times more efficient than conventional hydro
power units used for regulation and 2.7 times more efficient than combustion turbines. This
means that 1 MW of fast responding capacity can substitute for 1.7 MW of hydro or 2.7 MW of
combustion turbine. The difference is even greater when comparing fast responding resources
to steam turbines or combined cycle plants. As an example of a fast‐responding resource,
PNNL found that a flywheel is much more efficient than most conventional resources in
California, although its limited energy capability makes it less efficient than the fastest hydro
resources (Ibid., pp.16, 35‐36).
Reducing the capacity need for regulation procurement offers significant environmental and
cost benefits. For conventional units to be ready to respond quickly to regulation signals, a
large number of units must be kept online and partially loaded or “spinning.” These faster and
more flexible units may displace lower‐cost base‐load units, and partial loading may decrease
the efficiency of individual units and cause more to be online at a given time (Denholm, 2010,
p.17). A 2007 study commissioned by Beacon Power compared the emissions of a 20 MW
Beacon flywheel system used for frequency regulation to those of a coal‐fired power plant, a
natural gas combustion turbine, and a pumped hydro storage plant. Over a 20‐year lifecycle in
California ISO service territory, the flywheel provided an estimated savings of 103,455 tons of
CO2, or 53 percent savings over a natural gas base‐load plant. The savings over a natural gas
peaker plant was even higher at 132,917 tons or 59 percent CO2 emissions (Fioravanti, 2007,
p.18). The KEMA study also confirmed that utilization of energy storage avoids emissions
associated with scheduling gas turbines for ramping and regulation (KEMA, 2010, p. 76).51
5.1.5.2 Potential Cost Implications of Energy Storage for Frequency Regulation
The cost effectiveness of energy storage for area and frequency regulation will be determined by
competitive market prices for regulation‐up and regulation‐down, current installed energy
storage system costs, and the ability of the storage system to capture monetized value streams
from compatible operational uses by “stacking” or aggregating benefits. Operational uses
compatible with frequency regulation include provision of spinning/ non‐spinning operating
reserves, ramping or other ancillary services (when treated as a separate market product), and
possibly real‐time energy price arbitrage (SCE White Paper, 2010, pp.41‐42). However, under
current market and regulatory conditions, an energy storage device cannot receive bid awards
51 California Energy Commission and KEMA, Inc., “Research Evaluation of Wind Generation,
Solar Generation, and Storage Impact on the California Grid,” Energy Commission PIER Final
Project Report, June 2010, p. 76.
142
for the same capacity simultaneously in multiple California ISO markets; an energy storage
device can provide only one of these services at any given time.
Several studies have attempted to provide a present value estimate for energy storage in the
regulation services market. As noted in EPRI’s primer on energy storage applications, costs and
benefits, the resulting values are difficult to compare because the studies employ different
metrics, such as $/kW versus $/kW‐h (as opposed to the $/MW and $/MW‐h commonly used for
pricing and quantities associated with ancillary services), and different assumptions relating to
storage system lifecycle, inflation, and discount rates (Electric Power Research Institute, 2010a,
pp.2‐12 – 2‐13). Even so, the range of values reported provides a snapshot of current potential
values and highlights the challenge of estimating value when key factors remain uncertain.
In comparison to the estimated present values for other applications, regulation consistently
falls within the top few. In EPRI’s White Paper Primer on Applications, Costs and Benefits, the
highest ranking applications from the regional or utility perspective were: commercial and
industrial power quality and reliability; both stationary and transportable storage for
transmission and distribution support; and wholesale services with regulation (Electric Power
Research Institute, 2010a, p.2‐11). The highest value applications from the customer perspective
were home energy management and commercial/industrial energy management. In Sandia
National Laboratories’ Benefits and Market Guide, the highest ranking applications for financial
benefit were: substation onsite power; area regulation; transmission and distribution upgrade
deferral; time‐of‐use energy cost management; and load‐following (Nexight Group and Sandia
National Laboratories, 2010b, p.72, Table 11).
For actual estimated values, with the caveat that market prices can vary significantly, EPRI
found one hour of regulation to have a target and high present value52 of $255/kW‐h and
$426/kW‐h, respectively (Electric Power Research Institute, 2010a, p.2‐15, Table 2‐5). Using the
same (converted) metric, Sandia found one hour of area regulation to have a low and high
present value of $785/kW‐h and $2,010/kW‐h, respectively (Ibid., see also Eyer, 2010, pp.79‐80).
When EPRI assumed certain changes to the regulation market, such as shortening of the
minimum energy delivery rules from 60‐minutes to 30‐ or 15‐minutes, dispatching fast‐
responding resources first, compensating for fast response, and/or providing mileage or pay‐
for‐performance payments, the estimated values increased substantially, from less than
52 EPRI explains that the target value represents “an average value in the broader U.S. market
for stakeholders who might consider investing in energy storage,” whereas the high value
represents “the value for premium or niche markets that place a particularly high value on the
benefits provided by an energy storage system. (Electric Power Research Institute, 2010a, p. 2‐
6).
143
$1,000/kW‐h for one hour of fast regulation to over $6,000/kW‐h for fifteen minutes of
regulation (Electric Power Research Institute, 2010a, pp.3‐17 – 3‐18, Figure 3‐16).
The estimated present values alone do not provide much insight into energy storage cost
effectiveness for regulation services. Present value does not account for the breakdown of
benefits between multiple parties, such as the utility, independent third‐party, or customer
(Electric Power Research Institute, 2010a, p.2‐11), nor does it consider system cost.53 Southern
California Edison (SCE), however, developed a valuation methodology for the express purpose
of examining cost effectiveness. SCE conducted a high‐level assessment of economic feasibility
under current market conditions for various applications paired with appropriate energy
storage technologies options (SCE White Paper, 2010, p.5). SCE’s results indicate that even with
a higher market value, energy storage for grid‐level regulation is not yet cost effective. Using
the average of regulation‐up and regulation‐down prices between 2007 and 2010 in California
(years which featured significant price drops in the regulation market), and making other
reasonable assumptions about a storage device’s participation in current A/S markets, SCE
concluded that a 20 MW flywheel with 30 minute energy duration would be cost effective if the
current installed costs were to decrease by 25 percent and the average regulation‐up and
regulation‐down market rents were to increase by 50 percent (Ibid., p.44).54 This result appears
consistent with EPRI’s concluding analysis of the cost gap between application values and
system cost, although the two reports utilized different end metrics and had different scopes.
For frequency regulation, EPRI found that the energy storage technology cost range is in the
same range as or greater than the target and high values (Electric Power Research Institute,
2010a, p.5‐1, Figure 5‐1). But when regulation is provided in addition to other high‐value
services, such as transportable storage for transmission and distribution deferral, the combined
application values may exceed costs (Ibid., p.5‐2).
Several important conclusions flow from these assessments. First, in the baseline scenario for
energy storage deployment for regulation services, California’s regulation market does not offer
significant new opportunities beyond its current status. Although regulation‐up and ‐down is
California’s highest‐valued ancillary service, it does not yet constitute an economically feasible
application, at least when treated independently of other potentially compatible applications.
Second, however, a number of factors could shift frequency regulation into a high‐value and
53 The Sandia Benefits and Market Guide explains that financial benefit is estimated “without
regard to cost” and is “intended to provide a general indication of the price point required for
storage to be financially viable. So if storage can be owned and operated for an amount less than
the estimated benefit, then the value proposition may be financially viable.” Eyer, 2010, p. 71).
54 For a detailed list of the assumptions that underlie SCE’s cost effectiveness conclusions, see
SCE White Paper, 2010, Appendix A, pp. 66‐71.
144
cost‐effective application fairly quickly. These include the frequency regulation and energy
storage proceedings undergoing review at California ISO and FERC, especially introduction of
a mechanism to award fast‐responding resources first, a potential increase in competing natural
gas plant costs resulting from cap and trade program implementation under AB 32, and
anticipated reductions in technology costs, in particular for Li‐Ion batteries. As SCE cautioned
in its valuation study, pending changes to California ISO’s market design as well as the
difficult‐to‐predict effects of complex bidding practices, market liquidity considerations, and the
future behaviors of market participants, make any reliance on current valuation conclusions an
uncertain indicator of future cost effectiveness (See SCE White Paper, 2010, p.44). This concern
may have even more resonance with the current low prices for ancillary services due to
extensive supply, which could increase as the California ISO allows more types of non‐
generation resources to compete in the market.
While it is possible that installed system costs and market rents will shift under the baseline
scenario, assuming regulatory changes by FERC or California ISO, policy makers could
consider several actions to spur a more accelerated deployment, including the potential for
different ownership models (example given utility ownership versus non‐utility or third‐party
ownership) to shift possible risks toward or away from the ratepayer.
5.1.6 Area and Frequency Regulation: Recommended Next Steps and Milestones
Based on the assessment of the technology status, cost, and benefits for energy storage for
regulation services, this section outlines goals for PIER research, funding, focus, and
collaboration with industry and the CPUC and/or California ISO. It also identifies certain
research and/or R and D milestones for policymakers to address in their decision‐making.
5.1.6.1 Milestones
Critical upcoming research and technology milestones for policymakers to consider in making
decisions on energy storage deployment include the following:
• Final results of the California ISO 33percent renewables integration study, which will
forecast the amount of regulation potentially needed
• FERC approval of California ISO’s REM program for LESRs and REM implementation
• Completion of Phase 2 of California ISO renewables integration market products
proceeding
• Data from California‐based ARRA fast‐responding storage demonstration projects
• FERC ruling on Frequency Regulation Compensation NOPR
• Implementation of ARB Cap and Trade regulations and other events with cost
implications for competing fossil fuel technologies. These cost increases may improve
145
the cost effectiveness of energy storage technologies relative to natural gas‐fired power
plants.
5.1.6.1 Next Steps
The Energy Commission’s PIER program could focus on the following activities related to
energy storage penetration of the area and frequency regulation market:
• PIER funding for studies to review the impacts of recent tariff changes at NY‐ISO, ISO‐
NE, PJM, and other grid operators on energy storage technologies attempting to
compete in the area and frequency regulation market. In particular, PIER could
examine the impact of the NY‐ISO’s “fast first” scheduling and ISO‐NE’s “mileage
payment” under its pilot program and consideration of combining mileage payment
with performance based pricing.
• PIER funding to study a review of the technical requirements and cost implications of
intra‐hour and shorter scheduling at California ISO in connection with regional
balancing. Shorter scheduling may be necessary to justify tariff changes that would
compensate providers for speed and accuracy. Some comments received on this report
indicate that compensation for faster or more accurate regulation will not assist energy
storage manufacturers unless California ISO implements shorter scheduling, noting the
multiple initiatives at the Western Electricity Coordinating Council (WECC) related to
intra‐hour scheduling and a within‐hour energy market.
• PIER could work with the California ISO to assess the potential benefits of utilizing
energy storage for needed regulation over existing fossil fuel‐based power plants.
5.2 Exemplary Application #2: Renewables Grid Integration
The challenges associated with the integration of variable energy resources (VER) into the
electricity grid are a key driver behind the growing interest in energy storage technologies.
California ISO presents the renewable grid integration issue as follows:
Although wind and solar average hourly energy production can be forecasted
with reasonable accuracy, they will still produce significant intra‐hour and
minute‐to‐minute variability that is inconsistent with today’s grid operations and
scheduling procedures. In addition, sudden weather changes can lead to
corresponding variations in wind and solar generation output across the service
territory and cause significant energy imbalances on the system.
146
One strategy for mitigating the challenges of intermittent and variable generation
impact on grid operations is to use storage technologies. Energy Storage has the
potential to change the current “just in time” paradigm by absorbing energy
during one period and delivering it within another period based on system
conditions. Storage can address the dilemma of continuously matching supply
and demand. (California Independent System Operator, 2010d, p.4; see also
Denholm, 2010, p.17).
The Energy Commission likewise has identified energy storage as an important focal point in
the 2011 IEPR’s treatment of integrating preferred resources into California’s electricity system
(California Energy Commission, 2011a, p.3). And a wide range of studies reinforce the
proposition that energy storage technologies can serve a key function in the grid integration of
VER. However, these studies note that the actual need for energy storage to “firm” and “shape”
output from VER has yet to be quantified (Denholm, 2010, p.16). NREL proposes that whether
there is a “‘need’ for energy storage to enable renewable integration is actually an economic
question” that requires “comparing the economics of a variety of potentially competing
technologies including demand response, transmission, flexible generation, and improved
operational practices” alongside energy storage (Ibid., p.1). Important to include in the
economic calculation are the expected cost increases stemming from grid integration of VER,
including increased regulation procurement, greater operational demands (for example, more
starts, more ramping) on conventional resources, and increased need for flexible/ dispatchable
capacity (Kristov, 2011, Slide 26).
As noted above with frequency regulation, the economic calculus becomes more difficult when
analysts divide the broader application category of “renewables grid integration” into
component applications or operational uses, which may overlap and be subject to different
value measures and regulatory rules. EPRI’s renewables integration application description
includes ramp and voltage support for wind integration, off‐peak storage for wind integration,
and time shift, voltage sag, and rapid demand support for photovoltaic integration (Electric
Power Research Institute, 2010a, p.2‐5, Table 2‐3). SCE categorizes renewables integration
applications as “off‐to‐on peak intermittent energy shifting and firming at or near generation”
and “intermittent energy smoothing and shaping at or near generation” – one or both of which
may find value in providing dependable operating capacity, energy firming, energy shifting or
wholesale arbitrage, output smoothing, and potentially avoiding “dump” energy and providing
transmission system reliability or deferral (SCE White Paper, 2010, pp.36‐40). In addition,
Sandia National Laboratories classifies short‐ and long‐duration wind generation integration as
separate applications from, for example, renewable electric energy time shift and renewable
capacity firming (Eyer, 2010).
147
The breadth of values possible in a renewables grid integration function for energy storage
suggest that a valuation methodology will need to define the scope of the application, whether
focused on grid‐scale or distributed renewables integration, and the inputs to be considered for
both cost and benefit. Developing such a model is beyond the scope of this report, but the
following analysis seeks to synthesize the relevant technical and policy information that will
affect a final determination of feasibility of energy storage for renewables grid integration in
California.
5.2.1 Renewables Grid Integration: Definition and Application
The increasing integration of renewable energy into California’s energy supply can potentially
reduce reliance on fossil fuels and emissions from electricity generation. However, the
intermittent nature of most renewable energy sources introduces generation variability that can
cause operational and integration issues when connected to the electric grid on a commercial
scale. These issues fall into three major time periods for renewable integration: “smoothing”
(second‐to‐second or minute‐to‐minute), “shaping” (15 to 30 minutes such as for energy
imbalance and ramp‐up/ramp‐down), and “firming” (hour‐to‐hour shifts for firming and
electricity energy time shift to better match renewable production with demand). All three time
periods are crucial to renewable integration and can sometimes overlap. Energy storage
technologies can support the increased penetration of renewables‐generated electricity by
smoothing the power from these sources, thereby easing grid operation where large amounts of
wind and solar generation have been deployed. To optimize the effectiveness of these
operations, storage technologies need the ability to communicate and respond to the grid
through the system operator. For storage technologies used for short‐duration renewables grid
integration, the performance targets shown in Table 8 address roundtrip efficiency, system
lifetime, capacity, and response time metrics.
Energy storage can also play a role in long‐duration renewable energy integration. The cost of
electricity varies along with daily cycles of changing electricity supply and demand. Electricity
prices are higher when electricity is in high demand or when supply is low than they are when
the demand for electricity is lower or when electricity supply is high. Energy storage can take
advantage of lower electricity prices by charging a storage device during times of low price and
then discharging this electricity when electricity prices are high. Often the extreme prices (high
or low) occur not at peak times or at night, but at times of high rates of change in load or
renewable generation. Device charging can also address periods of excess energy supply and
mitigate overgeneration issues.
Electric energy time shift is often referred to as arbitrage, which involves the purchase and sale
of electricity at different times to benefit from a price discrepancy. For example, electricity
generated from wind at night or solar power in the morning can be purchased during these off‐
148
peak times and sold later during on‐peak hours. Storage devices used for electric energy time
shift, including pumped hydro plants, compressed air energy storage facilities, and large
battery installations, can typically store large amounts of electricity to optimize the gain from
electricity price differentials and offset the disadvantages of intermittent renewable energy
sources by shifting this energy to times when it is needed most. Table 7 portrays Renewables
Integration Short Term Metrics by target and associated supporting information.
Table 7: Renewables Integration Short Term Metrics
Metric Target Supporting Information
Roundtrip
Efficiency
75 ‐ 90% Roundtrip efficiency is the efficiency
measured at the transformer of the energy
output divided by the energy input.
System Lifetime 10 years System lifetime will vary by technology and
the number of cycles per year, but 10 years
with high cycling would be a sufficient
technology lifetime. Some energy storage
technologies may require battery pack
replacement more frequently, which should
be acceptable, assuming the overall life cycle
costs are reasonable.
Capacity 1 MW ‐ 500 MW The capacity need of a storage technology
will depend on the size and intermittency of
the renewables operation.
Response Time 1 ‐ 2 seconds to <
60 min.
Fast system response times will allow
storage to respond to changes in renewable
operation to minimize generation
fluctuations.
(Nexight Group, 2010b, p.18)
Performance targets for electric energy time shift focus on system capital cost, operations and
maintenance cost, discharge duration, efficiency, and response time. The environmental impact
of storage devices used for electric energy time shift is also an important factor to consider. If a
storage technology is able to meet these targets, the technology is well positioned to be adopted
149
by the electric power industry as a way to take advantage of the fluctuating price of and
demand for electricity. Table 8 portrays Energy Time Shift Applications – Long Duration
Metrics with associated supporting information.
Table 8: Renewable Energy Time Shift Applications - Long Duration Metrics
Metric Target Supporting Information
Capital Cost $1500/kW ‐
$500/kWh
$250/kW is a utility‐set metric that may not
reflect the full value of storage.
$500/kWh is a sufficient metric to make
storage competitive with gas turbines.
Operations and
Maintenance Costs
$250 ‐ $500/MWh Lower operations and maintenance costs
will allow storage to offer the greatest
economic advantages for electric energy
time shift.
Discharge Duration 2 ‐ 6 hours The price and demand for electricity may
fluctuate over several hours.
Efficiency 70 – 80% 70‐80% is an acceptable baseline efficiency
for energy time shift.
Response Time 5 ‐ 30 minutes The price of electricity will remain low or
high for several hours, which decreases the
need for an instantaneous response.
However, technologies with a fast response
could provide added value with some
frequency response and load following.
(Nexight Group, 2010b, p.19)
5.2.2 Renewables Grid Integration: Technology Status and Fit
5.2.2.1 Short Duration Renewables Integration
The technologies that energy storage experts previously identified as the most promising for
frequency regulation ‐ flywheels, adjustable speed pumped storage, Li‐ion batteries, advanced
lead‐acid batteries, and capacitors – also offer performance that could also meet the metrics for
short duration renewables integration. For further information on flywheel, Li‐ion battery, and
150
lead‐acid battery technologies, see the analysis of technology status, commercial readiness, and
ongoing research needs in the frequency regulation application discussion in section 5.1.2.
5.2.2.2 Renewable Energy Time-Shift Applications
Wind integration applications require both bulk storage capacities of 1 to 400+ MW for 4 to 10+
hours and multiple smaller 1‐ to 20‐MW/15‐ to 60‐min systems for smoothing and balancing.
Technologies that appear compatible with those requirements include pumped hydro,
compressed air energy storage (CAES) primarily with underground storage, large flow batteries
such as zinc‐bromine and vanadium redox systems, advanced lead‐acid batteries, and Li‐ion
battery and flywheel systems for fast response and smoothing (EPRI 2010a, p. 3‐8).
In recent years, the increase of photovoltaic penetration on the distribution grid has presented
operational problems for utilities. Energy storage systems can potentially alleviate voltage
swings in the distribution grid. Large photovoltaic applications may also require high‐power,
low‐energy storage systems that can perform many cycles and are capable of fast response.
Such systems would generally be in the size range of 500 kW to 1 MW or larger with 15 minutes
to 1 hr of storage and could include advanced lead acid batteries and Li‐ion batteries (EPRI
2010a, p. 3‐8). Table 9 presents Bulk Energy Storage for Long Duration Renewables Integration
costs, and other data arranged by type of technology.
151
Table 9: Bulk Energy Storage for Long Duration Renewables Integration
Technology Maturity Capacity
(MWh)
Power
(MW)
Dura‐
tion
(hrs)
Effi‐
ciency
(%)
Total Cost
($/kW)
Cost
($/kWh)
Pumped
Hydro
Mature 1,680‐
5,300
5,400‐
14,000
280‐
530
900‐
1,400
6 ‐ 10 80 ‐ 82 2,500‐4,300
1,500‐2,700
420‐430
250‐270
CAES ‐CT
Underground
Demo 1,440‐
3,600
180 8
20
68 ‐ 75 960
1,150
120
60
CAES
Underground
Com‐
mercial
1,080
2,700
135 8
20
68 ‐ 75 1,000
1,250
125
60
Sodium‐
Sulfur
Com‐
mercial
300 50 6 75 3,100‐3,300 520‐550
Advanced
Lead‐Acid
Com‐
mercial
Com‐
mercial
Demo
200
250
400
50
20‐50
100
4
5
4
85‐90
85‐90
85‐90
1,700‐1,900
4,600‐4,900
2,700
425‐475
920‐980
675
Vanadium
Redox
Demo 250 50 5 65‐75 3,100‐3,700 620‐740
Zn/Br Redox Demo 250 50 5 60 1,450‐1,750 290‐350
Fe/Cr Redox R&D 250 50 5 75 1,800‐1,900 360‐380
Zn/air Redox R&D 250 50 5 75 1,440‐1,700 290‐340
(EPRI, 2010a, p.4‐22)
Pumped hydro storage durations can exceed 10 hours. The pumped storage industry has been
investigating the costs of new projects. Manufacturers presently have data on the equipment
and civil construction costs because these component costs are virtually identical to ongoing
152
existing station modernization and expansion projects. However, the lack of large projects
under construction means that not much publically available data exist. Costs vary significantly
by site, but the values listed above are comparable to industry information and include project
contingencies and substation and interconnection costs (EPRI, 2010a, p. 4‐22).
CAES systems of up to 400 MW to 2000 MW+ are possible, as are underground storage
durations of 20 to 30 hours or longer. Data in the above table are only for the power and
storage duration shown. Estimates include process and project contingency and costs for NOx
(SCR) emission‐control technology. The data assume a storage cavern with salt geology; costs
for other geologies can vary significantly and are site specific. The data do not include costs for
siting, permitting, environmental impact studies and geological assessments. Future system
costs may be lower once developers make standard, pre‐designed systems available. (EPRI,
2010, p. 4‐23).
Advanced lead‐acid battery cost estimates are based on use of advanced industrial‐ grade
batteries from a number of suppliers. Battery life‐cycle costs can vary considerably by supplier
depending on the design basis duty cycle and design life. Purchasers need to consider battery
replacement costs, not shown here, as a variable O and M expense in any life‐cycle analysis.
The data include capital costs on a “rated” MWh delivered per cycle basis. Costs for 50‐MW
systems are based on development of conceptual designs. (EPRI, 2010a, p. 4‐23).
Manufacturers can size redox flow battery systems for a wide range of power and duration of
energy storage. Developers have not yet built technology options for large vanadium, Zn/Br,
Fe/Cr and Zn/air redox to serve large grid‐scale (+10 MW) applications. Analysts base estimates
on conceptual engineering designs, vendor quotes, site layout and grid interconnection
estimates performed by EPRI. Vanadium systems are technically more mature, while Fe/Cr is
still in the laboratory and early R and D stage of development (EPRI, 2010a, p. 4‐23).
5.2.3 Renewables Grid Integration: Technology Potential
Pumped hydroelectric storage provides quick‐ramping large‐scale electricity storage that
currently provides a range of ancillary services, including regulation, reserve, and capacity.
Pumped hydroelectric storage requires two bodies of water at different height, as discussed in
Chapter 1. Geography and concerns over environmental impacts limit acceptable sites, but
developers continue to identify locations and have filed over sixty FERC preliminary permits.
Run of the river‐hydro storage may also be able to support intermittent power generation.
Grid operators still need research on all aspects of this energy storage option to better evaluate
its potential (Fertig and Wagner, 2010).
153
As discussed in Chapter 1, two CAES plants are currently operational (one in Alabama and one
in Germany) and at least four others are in planning or construction stages. The two
operational CAES plants are used for peak shaving, arbitrage, and to supplement the ramp rates
of base‐load generators. CAES stores enough energy to smooth daily variability in wind and
solar power output and can ramp quickly enough to smooth hourly fluctuations. One planned
CAES plant, the 270 MW‐rated Iowa Stored Energy Park in Dallas Center, Iowa, will be used
specifically to smooth wind power output. Current CAES systems use natural gas and have a
heat rate about two‐thirds that of an efficient natural gas combined cycle plant. Adiabatic CAES
systems, which avoid the use of natural gas by storing the heat of compression and using it to
reheat the air that exits the cavern, may be feasible in the near future. CAES is more cost
effective than present battery technology at the GWh scale and can be sited in suitable geology
across much of the United States. Directions for future work include the demonstration of
adiabatic CAES and a more thorough valuation of the services CAES provides in different
electricity markets under varying levels of renewables penetration (Fertig and Wagner, 2010).
As discussed previously for frequency regulation application, batteries can store and release
power across a broad range of time scales, smoothing rapid fluctuations in the output of
renewable generators or mitigating daily variability. Despite their operational flexibility and
potential to provide a wide range of ancillary services, batteries have yet to attain high
penetration at grid scale. If the cost of batteries drops due to technical innovation and increased
production volume, analysts expect the use of batteries for grid applications, such as renewable
integration, to expand (Fertig and Wagner, 2010).
As discussed in Chapter 1, sodium‐sulfur (Na‐S) batteries are a commercial energy storage
technology finding applications in electric utility distribution grid support, wind power
integration, and high‐value service applications on islands. The round‐trip ac‐to‐ac efficiency of
sodium‐sulfur systems is approximately 80percent. The life of a sodium‐sulfur battery is
approximately 15 years after 4500 cycles at 90percent depth of discharge. Na‐S batteries are
currently available in multiples of 1‐MW/6‐MWh units with installations typically in the range
of 2 to 10 MW. The largest single installation is the 34‐MW Rokkasho wind‐stabilization project
in Northern Japan that has been operational since August 1, 2008. At this time, about 316 MW
of Na‐S installations have been deployed globally. Customers in the United States include
American Electric Power (11 MW deployed), P G and E (4 MW, in progress), New York Power
Authority (1 MW, deployed) and Xcel Energy (1 MW, deployed). All together, more than 316
MW are installed globally at 221 sites, representing 1896 MWh. Analysts anticipate that
installed capacity will increase to 606 MW (3,636 MWh) by 2012. (EPRI 2010a, p. 4‐10).
154
Vanadium redox batteries are one of the most mature of all flow battery systems available.
Electric power companies have deployed several vanadium redox systems, including (EPRI
2010a, p. 4‐11):
• A 15‐kW/120‐kWh unit operating over three years in a smart grid application by RISO in
Denmark
• A 250‐kW/2‐MWh unit at Castle Valley, Utah by PacifiCorp, which operated 6 years
before being discontinued when the application need changed
• A 200‐kW/800‐kWh unit at King Island, Tasmania by HydroTasmania
• A 4‐MW/6‐MWh unit at Tomamae, Hokkaido, Japan by JPower
• Smaller 5‐kW units that have been deployed in field trials.
Zn/Br batteries are in an early stage of field deployment and demonstration, and are less
developmentally mature than vanadium redox systems. While field experience is currently
limited, vendors claim estimated lifetimes of 20 years, long cycle lives, and operational ac‐to‐ac
efficiencies of approximately 65percent to 70percent. Module sizes vary by manufacturer but
can range from 5 kW to 500 kW, with variable energy storage duration from 2 to 6 hours,
depending on the application and need. Electric power companies have deployed small
projects comprising 5‐kW/2‐hour systems in rural Australia as an alternative to installing new
power lines. In the United States, electric utilities plan to conduct early trials of 0.5‐MW/2.8‐
MWh transportable systems for grid support and reliability. The first 0.5‐MW systems are
expected to be deployed in early 2011 by EPRI and a consortium of electric utilities. (EPRI
2010a, p. 4‐12).
Fe/Cr redox systems are still in the R and D stage but are rapidly advancing. The low‐cost
structure of these systems also makes them worth evaluating for grid‐storage solutions. (EPRI
2010a, p. 4‐14) Given the considerable uncertainties in performance and cycle life, process and
project contingencies are high, but continued development followed by demonstration appear
warranted.
Thermal energy storage (TES) is commonly associated with concentrated solar thermal power
(CSP). In this system, the heat energy from the sun is transferred to a solid or liquid phase
change medium (molten nitrate salt is currently favored) and stored for later use, when the heat
is transferred to a steam power cycle to drive a turbine. As a result, TES can be considered as
another energy storage system to support renewable integration. This technology was
demonstrated successfully at Solar Two (10 MW) in Barstow, CA from 1996 to 1999 and has
been in commercial operation at Andasol 1 in Spain since 2008. TES is especially applicable to
155
CSP because it smoothes out variation in solar power during the day and allows electricity
generation into nighttime peak demand hours. However, TES raises the already high capital
cost of a CSP plant. Key areas of research for this technology include: 1) reducing system
capital costs, 2) quantifying the long‐term effect of TES on the operation and maintenance of a
CSP plant, 3) evaluating the profitability of different TES configurations, 4) evaluating the effect
of different policy initiatives on the decision to use TES at a CSP plant, and 5) comparing the
economic and environmental implications of TES for power tower and parabolic trough
technologies. These last two areas of research are especially important because TES is more
efficient and cost effective when coupled with power tower than with parabolic trough CSP.
However, parabolic trough CSP is currently the industry leader, and as TES becomes more
important to CSP, policymakers may need to examine policy initiatives to shift industry
preference to power tower (Fertig and Wagner, 2010).
5.2.4 Renewables Grid Integration: Regulatory Status and Policy Changes
As discussed above, there are numerous storage installations and demonstration projects in use
or under construction to assist with renewables grid integration. A prominent example in
California is the grid‐scale Li‐Ion battery storage system procured by SCE for its Tehachapi
wind energy project, scheduled for installation in early 2012 with testing taking place through
the end of 2014 (Edison International, 2010). The growing number of projects directed toward
renewables integration form part of the baseline for energy storage deployment in this
application area. But renewables grid integration is not subject to a single, clear set of
regulatory rules, such as those that govern bidding into the ISO market for frequency
regulation. The value of an energy storage system serving primarily a renewables integration
function will be based upon project‐specific factors such as grid location, and on regulatory and
market factors such as wholesale market energy prices, avoided emissions and avoided
curtailment values, and whether the storage capacity can count toward meeting RA obligations
or long‐term procurement planning needs.
5.2.4.1 Renewables Grid Integration: Baseline Characteristics (1)
At present, the most important characteristic affecting baseline deployment of energy storage
technologies for renewables grid integration is the lack of information available regarding the
amount, locations, and types of energy storage systems that electric power companies need for
system‐level renewables integration. Every major study reviewed for this report noted that
industry experts generally believe that energy storage will be a key asset in integrating grid‐
scale renewable generation and providing additional regulation, ancillary services, and T and D
congestion relief and upgrade deferral. However, not many studies analyze how much energy
storage will be needed and whether energy storage is a cost‐effective solution (See Electric
Power Research Institute, 2010a, pp.3‐4; Denholm, 2010, pp.17, 27). Policy makers and analysts
156
must consider factors such as where grid operators will locate the energy storage asset, who
will operate it, how it will be operated, and how it will fit with the existing portfolio of
resources. The limited information available therefore may not provide an adequate basis for
industry, utility, and regulatory stakeholders to plan for energy storage systems to support this
important set of applications. Nonetheless, the operational and flexibility benefits that energy
storage systems bring to renewables integration, along with the values captured through a
combination of firm energy production, T and D deferral, and participation in the wholesale
energy markets, may continue to drive baseline adoption of energy storage technologies.
Multiple policies currently affect capacity and energy market price signals and long‐term
contract capability. These include resource adequacy and long‐term procurement planning; the
price spread between on‐ and off‐peak hours; T and D deferral values; and revisions to market
and PIRP rules under the California ISO renewable integration market and product review. As
discussed in the following section on cost implications, the impact of these policies on cost is
difficult to gauge due to uncertainty in regulatory proceeding outcomes and the earlier
described lack of a valuation framework for many of the attendant benefits.
Under current market rules, an LSE cannot count investments in most energy storage
technologies toward its RA obligations. As discussed in Chapter 2, the CPUC may respond to
stakeholder input and review RA rules to consider participation by more energy storage
technologies or to address long‐term contracting capability in the context of the AB 2514
proceeding. Without any indication of how the CPUC might decide this issue, the baseline
assumption for this report is that most energy storage technologies do not count toward RA
obligations and are rarely procured in long term Power Purchase Agreements (PPAs).
However, energy storage may still gain leverage in the current LTPP process based on its ability
to fulfill certain identified needs or avoid certain costs. For example, the LTPP preliminary
results for 2020 indicate an increased need for regulation and load‐following capacity and
suggest that some curtailment will be necessary, which could undermine the 33percent
renewable energy goal and potentially jeopardize the economics of some existing renewable
energy projects. The calculation of revenue requirements for different procurement scenarios
includes a cost for incremental transmission additions. The planning methodology includes a
proxy RA contract cost to fill any RA capacity gaps, as well as a market balancing tool that
purchases unspecified generation or sells surplus energy into the market when the capacity
contracted to a service area is out of balance with load– all gaps that could be addressed by
coupling energy storage with VER. In addition, all scenarios calculate the impacts of
greenhouse gas allowance allocation under the ARB’s draft cap and trade regulation, opening
up a possible value for storage in reducing the emissions attributable to IOU ratepayers (See
California Public Utilities Commission, 2011c). Thus, even a baseline scenario may present
157
opportunities for energy storage technologies to achieve value by providing some combination
of services – for example,, reducing need for new regulation resources or spinning reserve,
eliminating/ reducing curtailment, shifting load to fill capacity gaps, reducing Green House
Gasses (GHG) emissions – to the IOUs.
Finally, pending changes to the PIRP program and the market rules affecting VER at the
California ISO may create new opportunities for energy storage under the baseline scenario.
While the particulars of these pending changes remain uncertain, this analysis does not
presume they will have a significant impact. This report therefore addresses suggested
considerations for PIRP under the “accelerated” scenario.
With the 33 percent RPS by 2020 driving adoption of renewable resources, a large percentage of
which are variable, energy storage will continue to attract the attention of market operators,
utilities, other LSEs, and renewable generators. However, more information regarding
renewable integration needs and improved clarity in the application of storage to the market
mechanisms described may be necessary to increase energy storage adoption at a system level,
assuming current baseline policy (including from FERC) and market factors.
5.2.4.2 Renewables Grid Integration: Accelerated Characteristics (2)
In addition to the baseline need for better information and tools that allow energy storage to be
considered and valued in the LTPP process, California can implement a number of policies to
accelerate storage deployment in applications related to integrating variable renewable energy.
All of the changes detailed in Section 5.1.4 on frequency regulation would add value to
renewables grid integration, since storage capacity not needed at a given time for integration
purposes can bid into the regulation or other ancillary services markets. Incorporating storage
in some way into the RA process, or creating a parallel process for guaranteeing a certain
capacity of flexible resources to assist with integration needs, would encourage LSEs to evaluate
energy storage options for this purpose. One potentially valuable mechanism would apply
energy storage as an “adder” toward the RA value for VER, potentially increasing the amount
of dependable operating capacity claimed by renewables. Changes made at California ISO to
the intra‐hour scheduling rules and PIRP should place a high value on forecast and delivery
accuracy, providing an incentive for VER to invest in co‐located energy storage as insurance
against frequent or significant deviations from the forecasted energy delivery. Policy makers
are still grappling with the issue of whether the supplier or the customer should bear the costs
of mitigating these deviations.
An increase in the spread between off‐ and on‐peak prices that control the value of wholesale
energy arbitrage could have a positive effect on the value proposition for storage in integrating
excess renewable energy (such as wind) generated at night or in off‐peak hours. However, an
158
increase in the price spread may be unlikely if increased utilization of energy storage and
demand response narrow the off‐on peak gap between load and generation (SCE White Paper,
2010, p.36), and as the investor‐owned utilities develop and promote strategies for permanent
load shifting under the direction of the CPUC, as described in Chapter 2. Electric vehicle
charging during off‐peak hours may have a similar effect (See Eyer, 2010, p.66).
To address shorter‐term variability and fast ramping requirements, a possible policy measure
could impose ramping limits on renewable generation resources as an interconnection
requirement. This would allow the renewable generator to select what type of resource to use
(for example, storage or combustion turbine) and to decide between several business models
(for example, self‐procure onsite ramping capability, purchase in California ISO market, or
procure directly from a third party). (KEMA, 2010, p.72)
Energy storage procurement targets under AB 2514 could accelerate deployment, particularly in
support of renewables grid integration. Should the CPUC determine targets to be appropriate, it
could structure the targets to support the legislature’s initial emphasis on renewables
integration, as demonstrated by the opening text of AB 2514:
The Legislature finds and declares all of the following:
(a) Expanding the use of energy storage systems can assist electrical
corporations, electric service providers, community choice aggregators, and local
publicly owned electric utilities in integrating increased amounts of renewable
energy resources into the electrical transmission and distribution grid in a
manner that minimizes emissions of greenhouse gases.
(b) Additional energy storage systems can optimize the use of the significant
additional amounts of variable, intermittent, and off‐peak electrical generation
from wind and solar energy that will be entering the California power mix on an
accelerated basis. (Assembly Bill 2514 (Skinner, Chapter 469, Statutes 2010). Sec.
1(a)).
One possibility would be to tie any procurement targets under AB 2514 to the annual
procurement planning for 33 percent renewables, mandated under the April 12, 2011 enacted
California Renewable Energy Resources Act (SB 2, Simitian), as a means to avoid forced
curtailment of renewable resources and increased greenhouse gas emissions from natural gas‐
fired power plants used to integrate renewables. This approach would require the subject
utilities to include storage in their renewable procurement planning, possibly to meet a certain
percentage of the additional capacity or to create a certain capacity “buffer” against variable
delivery by the renewable resources that are procured. An alternate or complementary
159
approach, directed toward deployment on the customer‐side of the meter, would identify an
overall target to encourage energy storage system installation in conjunction with utility smart
grid deployment plans and with distributed generation systems supported by statewide
programs like the SGIP and CSI. Energy storage systems at the distribution‐level and system‐
level would require a different set of regulatory changes to remove barriers to participation in
ISO markets.
As with any other type of regulatory mandate, the cost effectiveness of accelerated deployment
based on these kinds of strategies requires careful evaluation. In addition, the same
consideration of environmental and “societal” benefits discussed earlier, including the
identification of a value that goes beyond the price point to be set by cap‐and‐trade, would
provide additional leverage for an accelerated vision of storage deployment for renewables grid
integration.
5.2.5 Renewables Grid Integration: Benefits and Costs
The benefits and costs of utilizing energy storage to integrate large amounts of variable
renewable energy into the grid are relative to those of utilizing the alternative, which are
primarily conventional generators. The uncertainties relating to potential market and
regulatory changes and the lack of information regarding the necessary scale of integration
services combine to constrain efforts to quantify the benefits and costs. The following
assessment therefore weighs the known or presumed benefits and costs in the context of
pending or possible regulatory and market changes.
5.2.5.1 Benefits of Energy Storage for Renewables Grid Integration
In its analysis of the role of energy storage with renewable electricity generation, NREL
emphasizes an understanding of how current resources address electricity demand variability,
primarily by the up and down cycling of conventional generators. (Denholm, 2010, p.17).
Partially loaded generators and responsive load (for example, demand response) provide
contingency reserves and some regulation services. NREL points to four significant impacts of
increased variable generation: (1) increased need for frequency regulation (discussed above), (2)
increase in ramping rate, (3) uncertainty in wind resource and net load55, and (4) increase in
overall ramping range (Ibid., p.18). Other studies have also identified increase in ramping as a
primary challenge resulting from growing VER penetration. (See KEMA, 2010, pp.50‐51).
55 The NREL report treats VER as a source of demand reduction, such that wind and solar
generation reduces load, leaving conventional generators to meet the “residual” demand of
normal demand minus the electricity produced by wind/solar. Thus, uncertainty in wind
production combined with uncertainty in load can increase net load uncertainty. See Denholm,
2010, p. 18, Fig. 4.1.
160
Sandia National Laboratories frames renewables grid integration as requiring dual solutions:
short‐duration to address ramping needs, and long‐duration for electric energy time‐shifting
and load following to balance and align production with demand (Nexight Group and Sandia
National Laboratories, 2010b, p.17).
Although “ramping” is not currently a market product, California ISO may decide to create
such a product to better address renewables grid integration (SCE White Paper, 2010, p.19).
KEMA found that lower than optimal amounts of storage might be sufficient to maintain
system frequency during normal operations or between ramp periods (10A.M.‐4P.M.) but
would result in degraded ramping performance during the daily morning and afternoon ramps
(KEMA, 2010, pp.67‐68). By contrast, co‐locating energy storage with renewable generation
enables off‐to‐on peak intermittent energy shifting, providing firm power that reduces the need
for increased ramping as well as frequency regulation and additional operating capacity (SCE
White Paper, 2010, p.36). Where grid operators require ramping, energy storage technologies
offer important benefits, based on response speed and the avoided emissions and cost of
retaining combustion turbine gas plants online for the same purpose (KEMA, 2010, p.55; Electric
Power Research Institute, 2010a, p.3‐7).
Another significant benefit for renewables integration may be the potential for energy storage to
reduce the need for or amount of curtailment. Wind integration studies have shown that
without some type of “outlet” for “excess generation” during periods of high production and
low demand, conventional generators must reduce output or wind energy must be curtailed
(Denholm, 2010, pp.23‐24). This challenge is likely to increase as VER displaces load‐following
conventional units that might have the flexibility to reduce output quickly. According to NREL,
the cycling down of conventional units has technical and economic consequences, just as
curtailment has adverse economic consequences for VER. (Ibid, p.25‐26, 31). These
consequences may be avoided by utilizing storage for electric energy time shifting and to meet
ramping requirements. In addition, storage can improve the capacity factor of wind generation
resources, especially when considering curtailment due to over‐generation or transmission
constraints (Electric Power Research Institute, 2010a, and p.3‐5). However, these studies
emphasize that regional system operators need further research into where and how much
energy storage is needed to address wind variability and system balancing.
Increasing amounts of large‐scale photovoltaic generation facilities in California similarly will
require solutions to mitigate voltage sag and rapid shifts based on cloud cover, the effects of
which may be more severe than wind ramp because of the rapid rate of change (Electric Power
Research Institute, 2010a, p.3‐8). Fast‐responding energy storage systems may be able to
mitigate such voltage swings quickly to minimize integration costs. Compared to wind, fewer
161
studies have focused on large‐scale photovoltaic integration needs, creating a need for further
research in this area.
The benefits of energy storage for both wind and photovoltaic grid integration may extend into
savings in avoided transmission and distribution infrastructure investments. Of the anticipated
$400 billion and $200 billion that utilities will have to spend on distribution and transmission
infrastructure, respectively, in the next 15 years, a sizeable portion of that infrastructure is
needed less than 400 hours per year during peak demand highs (Electric Power Research
Institute, 2010a, p.3‐9). Most studies rightly classify transmission and distribution deferral as a
separate application from renewables integration based on different performance and locational
needs.56 However, substantial benefits can accrue to the grid from electric energy shifting and
peak shaving at or near the generation site. Sandia National Laboratories notes that developers
can locate a project using energy storage to time‐shift energy from intermittent renewables so
that it also time‐shifts wholesale electric energy from the grid, possibly providing transmission
support or a T and D upgrade deferral benefit (Eyer, 2010, pp.128‐129). Stated differently, the
ability of energy storage for renewables grid integration to complement and serve multiple
other uses, including transmission and distribution upgrade deferral (Eyer, 2009, pp.51‐52, is
itself a valuable benefit.
5.2.5.2 Potential Cost Implications of Energy Storage for Renewables Grid Integration
As noted above, the cost effectiveness of energy storage for renewables integration will be based
upon project‐specific factors, especially operating cost and storage efficiency and regulatory and
market factors affecting the wholesale energy price and the other potentially complementary
value streams, such as avoided emissions, capacity, and T and D upgrade costs. Because
“renewables grid integration” can span multiple uses and value streams, providing a single
present value estimate is difficult. In their respective valuation analyses, Sandia National
Laboratories and EPRI provide high and low value estimates for the applications shown in
Table 11Application value Estimates.
56 For a complete assessment of T&D deferral benefits from modular energy storage, see Eyer,
2009.
162
Table 10: Application Value Estimates
Application Sandia National
Laboratories
EPRI
Electric energy time shift/
Price Arbitrage
$400/kW ‐ $700/kW $134/KW ‐ $800/kW
Renewables energy time
shift
$233/kW ‐ $389/kW $201/kW ‐ $500/kW
Electric supply capacity $359/kW ‐ $710/kW $220/kW ‐ $726/kW
Renewable capacity firming $709/kW ‐ $915/kW $88/kW ‐ $484/kW
Load following $600/kW ‐ $1,000/kW NA
Electric supply reserve
capacity
$57/kW ‐ $225/kW $80/kW ‐ $220/kW
Renewable grid integration –
short duration
$500/kW ‐ $1,000/kW $104/kW ‐ $311/kW
Renewable grid integration –
long duration
$100/kW ‐ $782/kW $104/kW ‐ $1,866/kW
(Eyer, 2010, p.72, Table 11; EPRI, 2010a, p. 2-15, Table 2-5)
Depending on the type of energy storage installation, its performance characteristics, and its
location on the grid, energy storage for renewables grid integration may be able to access some
combination of these values at different times, as well as the value for other services like area
regulation and transmission upgrade deferral. After reviewing twelve basic possible energy
storage application‐technology pairs, SCE concluded that “applications with the greatest
potential directly address the longer duration decoupling of supply and demand” because of
the “aggregated benefit streams associated with deferring or displacing peak‐related costs over
several hour durations” (SCE White Paper, 2010, p.9). Still, energy storage applications for
renewables grid integration do not fall within the top‐tier value propositions in the leading cost
analyses, although SCE found long‐duration integration uses to be promising although not yet
cost effective.
In its cost effectiveness valuation,57 SCE modeled both a 300 MW pumped hydro facility and a
20 MW, 6‐hour Na‐S battery for off‐to‐on peak intermittent energy shifting and firming at or
near generation. It also modeled a 10 MW, 30‐minute flywheel for intermittent energy
smoothing. SCE found that based on current net present values, none of these technologies
57 For a detailed list of the assumptions that underlie SCE’s cost effectiveness conclusions, see
SCE White Paper, 2010, Appendix A, pp. 66‐69.
163
were cost effective. To be cost effective by 2020 for energy shifting and firming, the pumped
hydro storage facility would require installed cost to fall by 35percent, transmission avoided
cost to increase by 25percent, and market rents from energy arbitrage to increase by 50percent.
The Na‐S battery would require technology installed cost to fall by 50percent and market rents
from energy arbitrage to increase by 75 percent to be cost effective. For intermittent energy
smoothing, SCE found that the flywheel installed technology cost would need to drop by
50percent and market rents would need to increase by a factor of 100 (SCE White Paper, 2010,
pp.37‐38, 41).
Notwithstanding the significant gaps between the current and cost‐effective value estimates,
SCE concluded that energy shifting and firming holds promise because of falling battery
technology costs and the potential for off‐peak renewable energy production, which
theoretically could increase the arbitrage value of energy shifting (Ibid., p.39). This effect is far
from certain; Sandia National Laboratories’ notes that even modest deployment of electric and
plug‐in electric vehicles “could lead to a non‐trivial decrease in that all‐important difference
between on‐peak and off‐peak energy prices,” threatening to decrease the value of electric
energy time shifting (Eyer, 2010, p.75). In addition, SCE and other studies have noted that as
energy storage and other electric energy shifting and load shifting technologies gain market
share, the high‐and‐low peak price differential may decrease further (see SCE White Paper,
2010, pp.75‐77).
For shorter‐duration energy smoothing applications, SCE concluded a low likelihood of cost
effectiveness in the near or medium‐term future but also cited the lack of metrics to guide a
valuation based primarily on avoided renewable integration costs (SCE White Paper, 2010, p.
41). Addressing this point in its 2010 evaluation of renewable generation and energy storage
impacts on the California grid, KEMA noted that while energy storage systems are more
expensive than combustion turbines for managing ramping and VER volatility, “the avoided
emissions and wear and tear may make the incremental cost worthwhile” (KEMA, 2010, p.72).
KEMA also recommended additional research on the cost value of both of these sets of benefits.
Policymakers and purchasers need research to address the considerable uncertainty
surrounding how wholesale energy prices and price differentials will change over the course of
the next ten years and under different scenarios for off‐peak renewable production, electric
vehicle off‐peak charging, and electric energy time shifting and load shifting. Without further
research, the near and mid‐term prospects for cost‐effective energy shifting uses in connection
with renewables grid integration remain unclear. Similarly, without a clear valuation metric
that places a price on avoided renewable integration costs, energy storage technologies serving
integration needs may not be able to recapture the cost of the benefits they provide. California
ISO’s consideration of cost allocation mechanisms for the integration of VER, including the
164
possible use of incentives for VER to manage their own variability and reduce impacts on grid
operations (Kristov, L. 2011, Slide 26), may ultimately provide clarification regarding the value
of avoided VER integration costs.
This type of metric is easier to create for some products than others. For electric supply
capacity, purchasers can calculate the value offered by energy storage based on the avoided cost
of installing new generation or on the price of procuring generation capacity in a regulated
wholesale electricity market (Eyer, 2010, p.76). If the California ISO one day creates a capacity
product, its price can be set for the wholesale level, possibly accounting for the avoided
emissions and pollution benefits of utilizing storage for capacity as opposed to combustion
turbines. The cost of avoided emissions will be easier to calculate once the ARB implements cap
and trade in accordance with its AB 32 Scoping Plan in 2012.
5.2.6 Milestones and Next Steps for Energy Storage for Renewables Grid Integration
As with Section 5.1.6 above, this section suggests goals for PIER research, funding, focus, and
collaboration with industry and the CPUC and/or California ISO. It also identifies certain
milestones for policymakers at the CPUC and the Energy Commission to consider.
5.2.6.1 Milestones and Next Steps for Energy Storage for Renewables Grid Integration
Critical upcoming research and technology milestones for policymakers to consider include the
following:
• Final results of the California ISO 33percent renewables integration study
• Completion of Phase 2 of California ISO renewables integration market products
proceeding
• Implementation of legislation to extend SGIP funding for energy storage with the
potential to finance standalone and solar‐coupled energy storage systems (AB 1150 [V.
Manuel Pérez, Chapter 310, Statutes of 2011])
• Data from California‐based ARRA grid storage demonstration projects
• FERC determination on the Request for Comments Regarding Rates, Accounting, and
Financial Reporting, as it particularly relates to asset classification and whether energy
storage technologies can qualify as transmission asset and how operators of these
technologies can be compensated for providing multiple services (for example,
transmission and generation or ancillary).
• Implementation of the ARB cap and trade regulations and other events with cost
implications for competing fossil fuel technologies. These cost increases may improve
the cost effectiveness of energy storage technologies relative to natural gas‐fired power
plants.
165
• The filing of proposed changes to the California ISO PIRP, which could increase the
incentives for PIRP resources to couple with energy storage technologies
5.2.6.2 Next Steps
The Energy Commission’s PIER program could focus on the following activities related to
energy storage and renewables integration:
• Funding for studies on how energy storage could operate in conjunction with demand
response technologies. Can energy storage increase the value, responsiveness, and
accuracy of demand response? Can energy storage increase the value by making
participation in demand response programs more palatable to end users with specific
electricity needs?
• Funding for a study on value of energy storage depending on location. How
dramatically does the value of the device change if it is sited with generation versus on
the grid closer to load? These questions are relevant to determine whether policy should
encourage co‐locating energy storage with generation assets, such as by changing the
PIRP program or by the CPUC placing ramp limits on VERs.
• Funding a cost study on new, fast, and efficient natural gas plants, such as GE’s recent
technology discussed in Chapter 3, Section 3.2.3, and how these options compare for
integrating renewables versus energy storage devices.
• PIER could work with the CPUC to define which applications fall within the “grid
renewables integration” category and what their valuation (such as frequency regulation
for grid renewables) should be.
• PIER could partner with the CPUC to develop a valuation of energy storage
technologies for renewables grid integration. The CPUC could look to SBX1‐2 for
direction on how energy storage may add value by reducing indirect costs associated
with transmission investments and operational needs for integrating eligible
renewables, both of which are to be considered in LTPP process for 33percent
renewables (Sec. 399.13(a)(1), (4)).
• PIER and the CPUC could review how demand response can apply toward RA
obligations, perhaps as a model for energy storage. The review could focus on the
possibility that utility procurement of energy storage could increase the capacity rating
of variable renewables for RA purposes.
• PIER could work with California ISO to account for the environmental and greenhouse
gas costs of utilizing natural gas power plants and current fossil fuel plants for increased
regulation and integration needs to meet the 33percent goal.
166
5.3 Exemplary Application #3: Community Energy Storage /
Distributed Energy Storage Systems (DESS)
Community energy storage or distributed energy storage systems (DESS) offer electric utilities a
different approach to grid management from applications related to regulation or system‐level
renewables grid integration. Locating DESS near load expands the range of potential uses and
applications to address grid‐level, distribution‐level, and community end‐user needs. Although
this analysis focuses on utility‐side DESS, distributed storage technologies placed on the
customer side of the meter would offer other services for commercial/industrial and residential
electricity end users. As California implements its ambitious clean energy agenda, locally sited
storage may play a key role in various other applications, such as integrating variable
distributed generation and implementing smart grid and electrified transportation programs.
5.3.1 Community Energy Storage: Definition and Application
Community energy storage, or distributed energy storage systems (DESS), involve small energy
storage systems sited on the utility side of the meter, typically next to a pad‐mounted
transformer serving four to eight residences, a business park, a campus, or multi‐family units.
Operators can remotely control individual DESS units to manage their individual charge and
discharge activity in response to regional need at the circuit, substation, or system level.
Analysts envision such units to support grid peak loads in the summer months and to provide
backup support as needed. Applications for DESS include:
• Distribution Deferral; Peak Shaving,
• Reliability, and
• Dual‐Mode Frequency Regulation.
167
Figure 17: Concept of a Neighborhood Using DESS
(SCE & EPRI, EPRI Smart Grid Demonstration Initiative, 2010, p. 9)
While the definition and specification of DESS systems are still evolving, (EPRI 2010a, p. 3‐11)
individual DESS units might typically have nominal capacities of 25 to 50 kW with 2 to 4 hours
of storage capability. Electric utilities could control, manage, and aggregate the DESS units
using an integration platform to provide large‐scale grid support management. However, other
business models may also emerge in which cities or third parties own the systems. To employ
DESS, utilities will require storage devices with low cost, long life (15 years or more), low
maintenance, and a small footprint. Technology options include advanced lead‐acid, Li‐ion, and
flow type batteries.
5.3.2 Community Energy Storage: Technology Status and Fit
This application is primarily for utility‐side‐of‐the‐meter grid‐support applications at the end of
the line, near pad‐mounted transformers. These systems also could be located near end‐use
customers or on the customer side of the meter for energy management, power quality and
reliability. In those applications, electric utilities may gain from distribution grid support and
peak load management.
All systems are modular, and operators can configure them in both smaller and larger sizes not
represented. Ideally, systems with 3 to 4 hours of energy duration may be of most value for
grid peak management. Figures in the table below represent estimated ranges for the total
capital installed cost of “current” systems, based on 2010 inputs from vendors and system
integrators. The data include costs of power electronics (if applicable) and all costs for
168
installation, step‐up transformer, and grid interconnection to utility standards. The data also
assumes that smart grid communication and controls will be included. The data do not include
siting and permitting costs and base installation costs on aboveground configurations.
Table 11: Distributed Energy Storage System Metrics
Technology Maturity Capacity
(kWh)
Power
(kW)
Dura‐
tion
(hrs)
Efficiency
(%)
Total
Cost
($/kW)
Cost
($/kWh)
Advanced
Lead Acid
Demo‐
Commer‐
cial
100‐250 25‐50 2 ‐ 5 85‐90 1600‐
3,725
400‐950
Zn/Br Flow Demo 100 50 2 60 1450‐
3,900
725‐1,950
Li‐ion Demo 25‐50 25‐50 1 ‐ 4 80‐93 2,800 950‐3,600
(EPRI 2010a, p. 4-28)
For batteries, values are reported at rated conditions based on reported depth of discharge.
Costs include both process and project contingency depending on technical maturity of the
system. This chapter previously presented the status of battery technologies for frequency
regulation and renewable integration applications, with key findings summarized here for
community energy storage applications.
Cost estimates for advanced lead acid batteries are based on use of advanced industrial‐ grade
batteries from a number of suppliers. Battery life‐cycle costs can vary considerably by supplier
depending on the design basis duty cycle and design life. Purchasers need to consider battery
replacement costs, not shown, as a variable O and M expense in any life‐cycle analysis. Capital
costs are reported on a “rated” kWh delivered per cycle basis.
Flow batteries have limited deployment in this application. Redox battery systems can be sized
for a wide range of power and hours of energy storage. While the chart only shows the Zn/Br
option, other redox chemistries may also find application in this setting but may need to include
additional siting and permitting costs depending on the technology. The estimates are based on
engineering designs and vendor quotes and include site layout and grid interconnection
estimates performed by EPRI.
Electric utilities are actively considering Li‐ion battery systems for this application. As
discussed in Chapter 1, there are several different types of Li‐ion chemistries, each with their
169
own cost and performance characteristics. Data shown is the average of currently available
systems. Each chemistry has its own cost structure, so actual selected system costs may vary.
Durability and life‐cycle cost data are unavailable at this time, and a life cycle analysis should
include battery replacements costs over the book life. Developers are testing numerous Li‐ion
systems for this application in 2011.
5.3.3 Community Energy Storage: Technology Potential
As discussed in Section 5.1.3, for essentially all energy storage systems, future costs can be
reduced once early demonstrations are proven and validated, products become more
standardized, and initial engineering costs have been removed. For the electrochemical energy
storage technologies, including batteries and capacitors, significant cost reductions may come
from high‐volume production of cells. The potential growth of electric and hybrid electric
vehicles could increase demand for high‐volume production and thereby reduce cell costs for
all applications. Operators would still need to integrate these cells, such as Li‐ion, into systems,
complete with thermal management, electronic controls, power conversion, and inter‐
connections, appropriate for frequency regulation and other applications.
Researchers continue to improve lead‐acid battery technology and materials, as discussed in
Section 5.1.2. Some advanced lead batteries have “supercapacitor‐like” features that give them
fast response similar to flywheels and supercapacitors. Advanced lead‐acid systems from a
number of companies are anticipated to be in early field trial demonstrations by 2011–2012.
Zinc‐bromine is a type of redox flow battery that uses zinc and bromine in solution to store
energy as charged ions in tanks of electrolytes, as discussed in Section 5.2.3. Small projects
comprising 5 kW/2‐hour systems are being deployed in rural Australia as an alternative to
installing new power lines. In the United States, electric utilities plan to conduct early trials of
0.5 MW/2.8 MWh transportable systems for grid support and reliability.
The first 0.5 MW systems are expected to be deployed in early 2011 by EPRI and a consortium
of electric utilities. (EPRI 2010a, p. 4‐14).
Li‐ion batteries are being seriously considered for several utility grid‐support applications such
as DESS (community energy storage). Both electric utilities and Li‐ion vendors are interested in
selecting one or two high value grid‐support applications that offer a combination of large
market size and high value to accelerate the volume production of PHEV batteries. Many
analysts believe stationary markets for Li‐ion batteries could exceed those for transportation.
Early system trial applications are underway using small 5 to 10 kW/20 kWh distributed
systems. Several electric utilities are also planning to deploy DESS systems in the 25 to 50 kW
size range. In addition, Li‐ion developer Altair Nanotechnologies has 1‐MW/250‐kWh trailer‐
170
mounted Li‐ion battery systems in service with both AES and PJM, while A123 Systems has a 2
MW unit serving the California ISO and another 12 MW installed by AES Gener at a substation
in Chile. In total, vendors have deployed approximately 18 MW of grid‐connected advanced
Li‐ion battery systems for demonstration and commercial service. (EPRI, 2010a, p. 4‐18).
5.3.4 Community Energy Storage: Regulatory Status and Policy Changes
No single set of policies, regulations, or market rules is likely to impact deployment of DESS for
utility‐side applications. The CPUC regulates distributed generation programs on both the
utility and customer‐side of the meter. Because these programs may increase local reliability
concerns, they may serve as a driver for more distribution‐level storage. In addition, the
California ISO monitors local capacity needs on a local/regional level and identifies areas with
local reliability problems and limited import capability. These areas may be attractive sites for
DESS deployment (California Independent System Operator, 2009, p.5). Some uses for DESS on
the utility side of the meter, such as peak shaving and frequency regulation, may require access
to the wholesale energy markets. Other benefits of DESS may rely on a valuation methodology
to calculate avoided cost, such as for variable distributed generation integration and for
distribution system upgrade deferral. Still other benefits may only become available with
changes to the RA regulatory process.
5.3.4.1 Community Energy Storage: Baseline Characteristics
At present, a number of programs exist to expand the contribution of renewable distributed
generation to California’s electricity generation portfolio. The SGIP and CSI programs provide
incentives for utility customer installation of distributed generation, including solar, wind
turbine, fuel cell, and corresponding energy storage systems when applicable, all of which affect
deployment of distributed energy storage systems. Other CPUC programs support utility‐side
or wholesale distributed generation (See California Public Utilities Commission ‐ (a)).
In addition, the extension of tradable Renewable Energy Credits to distributed generation
owners creates additional financial incentive for distributed energy resources.
The CPUC regulates the interconnection of these facilities and is engaged in a working group
process on renewable distributed energy resources interconnection rules reform (See California
Public Utilities Commission – (b)). A report commissioned by the CPUC in 2010 found that
distributed generation deployment has not had any noticeable impacts yet on transmission and
distribution infrastructure. However, projected distributed generation growth and the
possibility of a high‐penetration distributed generation scenario raises potential reliability,
power quality, and transmission/distribution‐related complications (Itron, 2010, pp.3‐10, 4‐1 ‐ 4‐
171
3). The continual increase in variable distributed generation – and potentially in the cost of
integration – may set a baseline need for distribution‐level DESS.
The investor‐owned utility Smart Grid Deployment plans may provide an additional incentive
for development, demonstration, and deployment of DESS. For example, S D G and E
emphasizes a vision for its smart grid based on “smart” customers, a “smart market,” and a
“smart utility.” The goal of the “smart market” is to empower customers to eventually
participate in demand response and possibly even ancillary services markets, maximizing the
efficiency of customer‐sited distributed energy resources (San Diego Gas and Electric, 2011b,
p.28). This vision requires the utility to ensure sufficient communication and balancing systems
for California ISO to maintain reliable service in the face of substantial changes in supply and
demand at the local or regional level. Although not likely to be a factor in the near‐term,
implementation of this type of smart grid plan will require flexible resources like DESS that the
utility can control to mitigate local variability at the distribution level.
Also present in a baseline scenario is the value of distribution upgrade deferral that may be
enabled by an energy storage device installed to reduce overloads on distribution lines and
improve power quality, reducing the need for infrastructure replacement or upgrade. This
value stream depends upon a clear and consistent valuation methodology for calculating
deferral value, creating a separate avoided cost calculation for this particular use. Depending
on the location of the DESS on the distribution grid, additional value could come from avoided
transmission fees, where larger DESS modules can provide on‐peak energy closer to load,
relieving some of the on‐peak energy flow in congested transmission lines (SCE White Paper,
2010, p. 49).
Participation in the wholesale energy market via price arbitrage may be necessary for DESS to
realize the value from peak shaving, depending on the ownership model. Although the energy
is stored downstream of the transmission system, a utility can calculate the difference between
charge and discharge prices (minus efficiency losses) to figure the avoided cost of procuring
that energy at peak hours from a grid‐scale generator (SCE White Paper, 2010, pp. 46‐47). As
with price arbitrage for renewables grid integration, multiple factors might affect the economic
viability of this value stream, making it an uncertain source of income in mid and long‐term
planning.
DESS may be able to find more certain value by offering dependable operating capacity for
eventual participation in the RA process. SCE terms this function “distribution‐level RA
deliverability” (SCE White Paper, 2010, p. 46). As with renewables grid integration, this option
is not available under current baseline RA rules. However this report considers it an option for
accelerated deployment.
172
5.3.4.2 Community Energy Storage: Accelerated Characteristics
Activities at the CPUC relating to distributed generation can be geared toward ensuring that
increasing amounts of variable distributed generation have a net positive impact on the
performance of the distribution grid. This goal might require consideration of DESS for
inclusion in customer‐side incentive programs, as well as in the context of any revised
interconnection standards that come out of the CPUC’s Renewable Distributed Energy
Collaborative working group.
In addition, the CPUC’s development of a valuation methodology (or more, depending on the
application) for energy storage pursuant to the AB 2514 rulemaking should include tools to
measure the value of distribution‐level DESS based on location and grid characteristics. This
methodology in turn depends upon a transparent process for valuing distribution upgrade
deferral benefits and avoided costs for variable distributed generation integration. In its smart
grid deployment plan, S D G and E suggests that accurate price signals are necessary to
encourage desired consumer behaviors. For example, by assigning the cost borne by the utility
of “storing” consumer‐generated renewable energy to those consumers participating in
residential net‐energy metering, these consumers might then decide to invest in energy storage
coupled to their systems (San Diego Gas and Electric, 2011b, pp.28‐29).
As with other applications for energy storage, accelerated deployment of DESS would benefit
from a change in RA regulations such that the utility can count investments in or procurement
of dispatchable DESS toward its RA obligations. Finally, targets promulgated under AB 2514, if
any, can be used to accelerate deployment specifically of DESS by linking procurement targets
to identified needs, such as distribution deferral, peak shaving, and distribution‐level frequency
regulation, for which the optimal storage resource would be located closer to load. Without
further research into the benefits and costs of DESS for these purposes, the CPUC may not be
able to determine viability and cost effectiveness of any targets in accordance with AB 2514.
5.3.5 Community Energy Storage: Benefits and Costs
As with grid integration of VER, there is insufficient information available on variable
distributed generation resources, the likely extent and rate of penetration, and the anticipated
effects on the grid and benefits and costs of utilizing energy storage to assist with integration.
This assessment reviews the known or presumed benefits and costs of DESS, installed on the
utility side of the meter, in the context of pending or possible regulatory and market changes.
5.3.5.1 Benefits of Energy Storage for Community Energy Storage
As discussed above, community energy storage may provide numerous benefits to the local
grid and may mitigate complications associated with increased amounts of variable distributed
generation – both customer‐side and utility‐side. Recent energy storage cost‐benefit analyses
173
classify distributed energy storage differently but identify many overlapping benefits.
Depending on the location and applications for the DESS, specific benefits may include
improved power quality and reliability, distribution upgrade deferral, distribution‐level outage
mitigation, variable distributed generation integration, area regulation or ancillary services, and
local reserve capacity. The EPRI, Sandia National Laboratories, and SCE valuation efforts focus
on the following DESS uses and benefits:
Table 12: DESS Uses and Benefits
Application Description Potential Benefits
EPRI Distributed Energy
Storage Systems ‐
commercial or
residential
Centrally managed
modular system (on
utility side of the
meter)
Increased customer reliability;
voltage support; grid
support/ distribution deferral;
reduce distribution losses;
possible ancillary services.
Sandia
National
Laboratories
Locational
Distributed
Storage
Applications
Applications best
served or only served
by storage that is
distributed where
needed/ close to load
Voltage support; transmission
congestion relief &
distribution upgrade deferral;
electric service reliability &
power quality; distributed
renewables integration/
capacity firming
Non‐Locational
Distributed
Storage
Applications
Applications that may
be served by
distributed storage,
located anywhere that
its operation does not
cause technical
problems for the grid
Electric energy time‐shift;
electric supply capacity; load
following; area regulation;
reserve capacity; renewables
energy time‐shift
Southern
California
Edison
Primary
distribution level
peak shaving and
outage mitigation
Transportable storage
located downstream of
an outage or overload;
provides power to
end‐users during
short‐term
distribution overload
RA and dependable operating
capacity; energy shifting/
price arbitrage; outage
mitigation; avoid
transmission‐level dump
energy/ curtailment;
distribution upgrade deferral;
power quality; end‐user
backup power
Peak shaving
below “secondary”
Storage located
downstream of
RA and dependable operating
capacity; energy shifting/
174
distribution system
level
secondary distribution
system to alter end‐
user load shape by
charging off‐peak and
discharging at peak
price arbitrage; distribution
upgrade deferral; avoid
transmission‐level dump
energy/ curtailment; local/ in‐
basin generation;
transmission congestion fee
avoidance; power quality;
outage mitigation
Intermittent “DG”
output smoothing
and integration
Storage located
between distributed
generation sources
and the distribution
system; devices
charges when
generation exceeds
power consumption
Variable distributed
generation integration/
minimize backflow of energy
onto the grid; avoid energy
dumping; local/ in‐basin
generation; variable resource
smoothing & shaping; power
quality; outage mitigation
(Eyer, 2010, pp.56-59; EPRI, 2010a, pp.2-7, Table 2-4 and 3-10 to 3-12; SCE White Paper, 2010, pp. 45-54)
As the penetration of variable distributed generation, especially photovoltaic, increases, the
challenges associated with managing voltage and intermittent variation in loads are expected to
increase. For variable distributed generation “output smoothing” and integration, the DESS can
be located to minimize or avoid potential “backflow” of unused or excess energy from the
generation site onto the grid. A small percentage of energy backflow at a given time may not be
problematic. However, unmitigated backflow from increased variable distributed generation
may overstress circuits, requiring upgrades or causing outages (See SCE White Paper, 2010,
pp.22‐23). In addition, sudden momentary drops in output from distributed generation can
increase variable energy integration costs. Carefully sited DESS have the potential to avoid line
upgrades at the point that will provide the most value to the system, serving both to smooth
output and to absorb and store backflow. Another potentially compatible benefit is the
avoidance of transmission congestion fees, where the DESS store and release energy to reduce
demand downstream of a congested transmission area (Eyer, 2010, p.57).
According to EPRI, many of these benefits will have value where solar generation is
concentrated on the distribution system, but the value is difficult to quantify as alternative
strategies for managing concentrated photovoltaics are still being developed (Electric Power
Research Institute, 2010a, p. 3‐19). Researchers struggle to quantify other benefits. For example,
the cost value of outage mitigation and backup power are based on the individual
circumstances of the affected end‐users, while electric energy shifting depends upon highly
variable wholesale prices and possibly diminishing price spreads. (See SCE White Paper, 2010,
pp.46‐47; Eyer, 2010, p.66). Still other applications, such as the use of transportable storage for
175
distribution level peak shaving, offer the benefits of avoided generation procurement and
distribution deferral – both of which require valuation that considers what would be the
generation and/or distribution upgrade needs of the particular location for the deployment of
DESS.
5.3.5.2 Potential Cost Implications of Energy Storage for Community Energy Storage
The cost effectiveness of utility side of the meter DESS will vary based on the needs of the
location and distribution area served, including the value to the end users of electric power
quality and reliability and the avoided cost values attached to distribution upgrade deferral and
avoided generation procurement for variable distributed generation integration. Sandia and
EPRI provided present value estimates for key DESS benefits, including voltage support and
distribution upgrade deferral, renewables integration avoided cost, and price arbitrage. Alone,
these benefit values are slight compared to the technology and operating costs. But when
aggregated across several categories, DESS may possess economically viable value propositions.
EPRI illustrated this aggregation value with an example of stationary T and D support – an
example which could work similarly for transportable or modular storage for DESS
applications. In the California ISO market, the aggregated value of such services as deferred
transmission investment, system capacity, and voltage support amounted to less than $500/kWh
of present value of energy storage. However, when the same system was modeled to serve
regulation in the ancillary services market and potentially defer transmission investments
(which would most likely occur by co‐locating energy storage and renewable generation), the
value more than doubled (Electric Power Research Institute, 2010a, pp. 2‐8 – 2‐9). EPRI cautions
that the locations in which all of these benefits can be realized together are limited, as is the
market size of the higher‐value value propositions. Even so, the modular configuration of DESS
allows them to be sited to maximize the services and benefits that can be aggregated. The range
of those services is broader than what central or bulk storage could address (Eyer, 2010, p.56),
suggesting that proper planning and assessment of location and compatible applications could
create opportunities for cost‐effective DESS deployment.
For example, the use of transportable storage for DESS applications may increase the value over
time. Sandia provides an example of a transportable storage device that, over the course of 10
years, is used either for T and D upgrade deferral or to serve electric service power quality and
reliability needs. Based on Sandia’s present value estimate for both of those applications, a
modular DESS that could be owned and operated for less than $1,700/kW for 10 years could be
cost effective (Eyer, 2010, p.129).
176
SCE’s analysis of distributed energy storage applications found that several hold promise and
may even be hypothetically cost effective for limited niche applications.58 For intermittent
distributed generation “output smoothing” and integration, SCE modeled a 500 kW, 15‐minute
Li‐Ion battery, such as could be installed to balance a 1 MW photovoltaic system.59 The SCE
report stated, “Initial valuations could in fact be cost effective at current technology costs (see
figure 14 and Appendix A for additional information). However, in the near term (next few
years), large DG sites requiring costly infrastructure upgrades will be passed over for those
which do not. In other words, given the current availability of suitable sites, DG deployments
will naturally gravitate first towards projects which will not require distribution upgrades.
While a potentially promising use of storage, at best SCE believes actual deployments will be
limited to extremely specific projects which do not currently exist but might in the future” (SCE
White Paper, 2010, pp.53‐54).
For peak shaving below the secondary distribution system level, SCE again modeled a Li‐Ion
battery, noting that commercially available sodium‐sulfur batteries are too large (in capacity
and physical size) for the application. To be cost effective, a 25 KW, 4‐hour, high‐energy Li‐Ion
battery would require technology costs to fall by 50 percent, T and D deferral costs to increase
by 10 percent, and market rents from arbitrage to increase by 50 percent. SCE found these
figures to be promising based on anticipated technology cost reductions and assumptions that
developers could meet certain operational practices (SCE White Paper, 2010, p.51). Moreover,
extending the location for this DESS further “downstream,” closer to the customer or end user,
can increase distribution deferral value. By contrast, SCE found that a transportable, 2MW, 4‐
hour Li‐Ion battery used for primary distribution level peak shaving and outage mitigation did
not carry the same cost‐effective potential based on technology costs and the difficulty of
evaluating distribution‐level benefits without location‐specific field trials (Ibid., pp.48‐49).
5.3.6 Milestones and Next Steps for Community Energy Storage
As with Section 5.1.6 and 5.2.6 above, this section suggests goals for PIER research, funding,
focus, and collaboration with industry and the CPUC and/or California ISO. It also identifies
certain milestones for policymakers at the CPUC and the Energy Commission to consider.
58 For a detailed list of the assumptions that underlie SCE’s cost effectiveness conclusions, see
SCE, 2010a, Appendix A, pp. 71‐74.
59 This is an example of DESS tied to a particular end‐use, and most likely installed on the
customer side of the meter, rather than utility‐side. Nonetheless, it demonstrates the cost
effective potential of DESS and could be extrapolated to apply to a utility‐side system sited to
serve a neighborhood or community.
177
5.3.6.1 Milestones and Next Steps for Community Energy Storage
Critical upcoming research and technology milestones for policymakers to consider include the
following:
• IOU filing of Smart Grid plans,
• Final results of the California ISO 33percent renewables integration study,
• Implementation of legislation to extend SGIP funding for energy storage with the
potential to finance standalone and solar‐coupled energy storage systems (AB 1150 [V.
Manuel Pérez, Chapter 310, Statutes of 2011]),
• Data from California‐based ARRA grid storage demonstration projects,
• Implementation of TOU/dynamic rates under the smart grid legislation timeline, and
• California meeting the CSI and distributed generation goals.
5.3.6.2 Next Steps
The Energy Commission’s PIER program could focus on the following activities related to
community energy storage:
• PIER funding for research into communication technologies that enable participation
and lower cost of participation by non‐utility and end user owners with grid connected
energy storage technologies. This research should seek to understand the integration
implications of high‐density solar PV areas on the distribution and transmission grid. It
should also determine a value for DESS in conjunction with PHEV/EV charging stations
that rely on variable renewable resources. The study should contemplate second use of
PHEV/EV batteries for modular DESS short duration applications. Finally, it could
assess the impact on the cost effectiveness of DESS if it is based on second use
technology, including an accounting of the societal and environmental benefits.
• PIER should work with the CPUC to develop a valuation methodology for DESS, both
utility‐side and end‐user‐side.
• PIER should partner with IOUs and public utilities to fund and support smart grid plans
that use distributed generation and energy storage, especially with utility devices for
communication and scheduling.
• PIER should work with distribution system owners to develop additional research
projects, as needed.
178
CHAPTER 6: 2020 Energy Storage Vision and
Recommendations
The AB 2514 process presents California with an opportunity to be a leader in incorporating
energy storage technologies to address vital grid needs and environmental and energy goals.
This concluding section seeks to inform policymakers at the CPUC and Energy Commission as
they help determine the role that energy storage can play in the state’s electricity system and
offers a vision for how energy storage can address pressing grid needs in 2020. With myriad
technologies and applications at issue, coupled with the fast pace of technology developments
and the changing nature of the state’s electricity generation portfolio, few subjects are as far‐
reaching and volatile as the future of energy storage in the electricity system.
6.1 Energy Storage Vision for 2020
As policy makers anticipate future grid needs and how energy storage can address them, they
should consider a number of key developments that are likely to transpire between now and
2020 that will influence the role of energy storage. Given the technology and policy factors
examined in this report, this section identifies the following key developments as critical
determinants in envisioning optimal energy storage deployment by 2020.
6.1.1 33 percent or More Renewable Energy Generation Under the 2020 Renewables
Portfolio Standard
The California Renewable Energy Resources Act (Senate Bill 2, Simitian), signed by Governor
Brown into law in April 2011, requires California to procure 33percent of its electricity from
renewable sources by 2020. However, integrating these variable and intermittent sources, such
as solar and wind, will require an array of policy mechanisms. Increased energy efficiency,
demand response, aggregating variable resources across large areas, forced or paid curtailment
of the renewable sources, increased deployment of fossil fuel‐based generators, such as natural
gas‐fired turbines, and the emerging Smart Grid, can help meet the integration challenge.
However, energy storage offers a beneficial option to integrate this variable energy without
additional greenhouse gas emissions, while providing other services for the grid, such as peak
load shaving and back‐up power in the event of blackouts. While policy makers should explore
all cost‐effective options, given California’s environmental and greenhouse gas goals and need
to avoid curtailment of renewable resources, California may require between 3,000 to 4,000
megawatts of fast‐acting energy storage by 2020 to integrate the projected increase in renewable
energy. Studies by KEMA and the National Renewable Energy Laboratory, among others,
provide support for this estimate, which could increase or decrease by 2020, depending on a
number of policy and technology developments.
179
6.1.2 Implementation of the Smart Grid
The deployment of “Smart Grid” systems in California may enhance the market for customer‐
side of the meter energy storage systems. The Smart Grid concept is based on actively
managing resources and loads by taking advantage of real‐time data monitoring, analysis, and
control enabled by digital sensor, information, and communication networks. The current
power grid is designed to have generation sources respond on‐demand to user needs, while a
smart grid can be designed so that usage varies on‐demand with production availability from
intermittent power sources such as wind and solar. The utility can actively shed end‐user loads
during peak usage periods, or the cost per kilowatt can dynamically vary between peak and
non‐peak periods to encourage turning off non‐essential high power loads.
The emergence of a “smarter grid” may both improve and lessen the need for energy storage.
Because utilities may be able to reduce end‐user load with the new, smarter technology, they
may not require energy storage to shave peak load or maximize the efficiency of renewable
energy electricity production. However, as customers, particularly from the business and
industrial sectors, access real‐time pricing information and identification of periods of peak
electricity rates, they may consider purchasing on‐site distributed energy storage systems,
particularly batteries, to avoid peak charges. These purchases will in turn drive down costs for
the technologies both due to economies of scale with greater production and the potential to
spur innovations that make the technologies more efficient.
In addition, residential and business consumers can use the smart grid to increase their
purchases of thermal energy storage technologies for heating, ventilation, and air conditioning
(HVAC) systems. Utilities can also coordinate and aggregate these systems on the utility‐side of
the meter, resulting in expanded markets for thermal energy storage technologies. Closely
related to demand response technologies, these systems have the potential to accommodate
intermittent renewable generation and reduce peak load.
California will benefit in 2020 from federal investments in smart grid demonstration projects
and the associated energy storage projects involved in these efforts. The American Recovery
and Reinvestment Act invested $685 million in smart grid and energy storage demonstration
projects, laying a foundation for utilities and industry to partner with the government and
enable the real‐world testing and development necessary for energy storage to gain ground.
These projects should yield important data on cost and performance to spur further research,
investment, and deployment by 2020.
6.1.3 Proliferation of Microgrids
A new manifestation of a utility‐based Smart Grid is the emergence a rapidly growing market of
“microgrids” for industrial/commercial‐type customers that require extremely high reliability
180
and availability of electricity supply. Leading examples of microgrids include military bases,
data centers, some manufacturing processes, universities, flight control centers, and
governmental facilities. The military represents the largest growth customers for microgrids,
given their energy mandates from legislation and executive order to make dramatic
improvements to their installations’ resiliency and endurance as a military force.
Electric energy storage is a mandatory element in the design and architecture of a microgrid
that does not rely on fossil fueled electricity generation. Energy storage is central to achieving
the higher reliability and availability of energy supply, as well as maintaining the power quality
on distribution circuits. As with military applications, only energy storage can prevent a loss of
access to power, provide efficient access to alternative and renewable energy sources available
to the installation, and offer adequate and resilient power for critical missions.
In essence, microgrids are a recent microcosm of larger‐scale Smart Grid (transmission and
distribution) utility in the digital energy age. They are likely to become more common by 2020.
Because microgrids represent early adopters of electricity energy storage by necessity and its
high value, their deployment will be an excellent tool to evaluate new and emerging distributed
energy storage technologies and to address their integration issues, particularly as the Smart
Grid continues to deploy at utility scale. Thus, greater micro‐grid deployment in California will
lead to better data and a more robust market for energy storage technologies.
6.1.4 Increased Residential Needs for On-Site Solar, Electric Vehicles, and Home Area
Networks
The proliferation of residential PV systems and the potential large‐scale deployment of electric
vehicles (as well as an increase in zero net energy buildings) have the capacity to fundamentally
change the market for energy storage in 2020. Customers can potentially use behind‐the‐meter
energy storage systems, particularly batteries, to capture and balance electricity from on‐site
renewables. They may also be able to use their battery systems to participate in power markets
by providing grid services when not in use. However, most residential customers will still need
to access electricity from utilities as backup or supplemental power. Barring technological
advances, this application may ultimately be more effective for commercial customers to offset
high demand charges while still relying generally on grid‐supplied power.
The deployment of electric vehicles (EV) may create other opportunities for energy storage.
Large‐scale adoption of EVs will likely drive down battery costs for lithium ion batteries as
manufacturers increase production to meet demand. Engineers are developing fast charging
systems, referred to as “Level 3,” to recharge EVs in a matter of minutes rather than hours.
The delivery of so much electrical power in a short period of time could stress the local
distribution network, so the addition of energy storage between the grid and Level 3 chargers
181
could provide needed buffering.
In addition, vehicle owners in 2020 and sooner may be able to sell or otherwise recycle used
batteries for second‐use applications, such as providing distributed storage opportunities for
the grid. Meanwhile, vehicle owners may be able use their batteries in “vehicle‐to‐grid”
applications that, when plugged in to specially designed bi‐directional charging systems, could
offer ancillary services to utilities, such as frequency regulation. The utilities could compensate
them for these services, providing vehicle owners with a revenue stream from their EVs.
Vehicle‐to‐grid research is ongoing to understand the impacts of these services on battery life
and how consumer behavior may affect the potential for this application.
6.1.5 Opening of Electricity Markets to Competition From Energy Storage and
Incentive Programs
The opening of various electricity markets to energy storage competition through Federal
Energy Regulatory Commission (FERC) and the California Independent System Operator (Cal
ISO) rulings will likely spur greater deployment of these technologies that can compete better in
a less restrictive market. For example, current markets for regulation services often disfavor
fast‐acting energy storage technologies, such as flywheels, that could otherwise provide
competitive services with fossil fuel‐based providers. Energy storage will also play increasing
roles in providing other operational and market benefits for the grid, such as congestion
management, oscillation mitigation, and infrastructure/asset management for optimal
utilization, among others. As market rules change to allow the monetization of the benefits
provided by specific energy storage technologies, investors and developers will fund more
projects and thereby help decrease the costs and provide more data on performance.
In addition, regulatory entities like FERC and the California Public Utilities Commission can
provide greater certainty of cost recovery for investments in energy storage technologies by
ensuring that these asset investments have greater certainty of cost recovery and by developing
a valuation methodology to help monetize the benefits provided by energy storage. Additional
changes, such as allowing more energy storage technologies to participate in the Resource
Adequacy program, would provide utilities with incentives to procure more energy storage
capacity. These regulatory changes will likely stimulate greater investment and deployment of
energy storage across the grid.
Complementing these regulatory initiatives, the state and federal governments may develop
incentive programs, through new programs or the expansion of existing ones, to help finance
energy storage projects. From federal loan guarantees and funding for demonstration programs
to expansion of the self‐generating incentive program, these measures will likely bring down
costs and help manufacturers provide better data to spur further investment by 2020.
182
6.1.6 Natural and Human-Caused Disasters
Disasters over the next decade may spur greater investment in energy storage. Possible
scenarios range from costly blackouts that jeopardize business functions and perishable goods
storage to nuclear meltdowns, earthquakes, terrorist attacks, and other catastrophes. One of the
results of these disasters could be regulations that require on‐site backup energy storage to
provide uninterrupted and continuous electricity and to help the grid restart as quickly as
possible, among other benefits.
6.1.7 Likely Price Decreases for Various Energy Storage Technologies
The result of the potential increase in investment and deployment of energy storage
technologies, in turn driven by the aforementioned policy changes and technological advances
that may occur over the next decade, means that California and the nation will experience
potentially significant price drops in the cost of various energy storage technologies by 2020.
For example, the proliferation of electric vehicles may drive down battery costs, while market
rules that allow greater penetration of flywheels for specific grid applications may drive down
costs and increase deployment. Increased penetration of renewables will likely increase the
need to balance the intermittent load with non‐carbon based fuel sources, spurring a market for
up‐and‐coming technologies like those related to batteries. In addition, advances in existing
technologies, such as creative siting for pumped hydro facilities in abandoned mine shafts, may
eliminate barriers that could lead to wider adoption.
While not all technologies will benefit from economies of scale and will still require
technological advancements to lower costs and ensure performance, California will likely
experience cost benefits from the coming grid developments, while the technologies will also
benefit from increased investment that could spur the necessary technological innovation. In
addition, demonstration projects will provide critical data and certainty and could provide
more assurance to investors.
The likely competitor for energy storage over the next decade may be inexpensive natural gas.
However, natural gas prices are not likely to stay at current prices and may increase by 2020,
leading to greater market opportunities for energy storage. In addition, the ability of energy
storage technologies to harness and dispatch carbon‐free renewable energy will become more
important as the state implements carbon pricing and other legal and regulatory mechanisms to
reach its ambitious greenhouse gas reduction goals. As a result, by 2020, the need for energy
storage will be significantly greater than today.
6.1.8 Future Research Needs to Implement the Vision
To meet the challenges and vision outlined in this section, this report cites a number of technical
and policy research needs that the state should undertake to help implement the vision of
183
energy storage in 2020. These needs include:
• Bulk energy storage demonstrations for variable renewable energy integration (for
example pumped hydro, CAES, and solar thermal).
• Field demonstration of modular energy storage technologies (for example batteries,
flywheels) in various grid applications.
• Evaluation/demonstration of aggregated storage, for example TES HVAC or electric
vehicle (EV) batteries, especially in a Smart Grid scenario.
• Develop simulations, analytical tools, and intelligent control systems for planning,
designing and dispatching energy storage devices for multiple applications and benefits.
• Quantification of costs and benefits of energy storage in grid applications.
• Models of the impact of 33percent renewable energy on California’s electricity grid to
determine needs for energy storage to support the grid, including sensitivity analysis to
address cost variables of storage and other needed energy resources, environmental
impacts, and emerging Smart Grid performance enhancements.
• A valuation methodology for energy storage that accounts for its costs and benefits to
ratepayers and the public, given the state’s energy and environmental goals. This
method could value technologies based on the most promising applications, ensuring
that benefits to additional applications are appropriately counted. This report identifies
three potentially promising applications for energy storage: frequency regulation,
integrating variable renewable energy, and developing community or distributed
energy storage systems. In addition, the methodology could analyze how the location of
the energy storage technologies may change the value of the resource. The data would
help determine whether policy should encourage co‐locating energy storage with
generation assets.
• Research into communication technologies that enable participation and lower cost of
participation by non‐utility and end‐user owners with grid connected energy storage
technologies. This research should seek to understand the integration implications of
high‐density solar PV areas on the distribution and transmission grid. It should also
determine a value for distributed energy storage systems in conjunction with PHEV/EV
charging stations that rely on variable renewable resources. The study should
contemplate second use of PHEV/EV batteries for modular distributed energy storage
systems short duration applications. Finally, it could assess the impact on the cost
effectiveness of DESS if it is based on second‐use technology, including an accounting of
the societal and environmental benefits.
• Studies on how energy storage could operate in conjunction with demand response
technologies. Can energy storage increase the value, responsiveness, and accuracy of
184
demand response? Can energy storage increase the value by making participation in
demand response programs more palatable to end users with specific electricity needs?
• Studies to review the impacts of recent tariff changes at NY‐ISO, ISO‐NE, PJM, and other
grid operators on energy storage technologies attempting to compete in the area and
frequency regulation market. In particular, the study could examine the impact of the
NY‐ISO’s “fast first” scheduling and ISO‐NE’s “mileage payment” under its pilot
program and consideration of combining mileage payment with performance based
pricing.
• Studies to review the technical requirements and cost implications of intra‐hour and
shorter scheduling at California ISO in connection with regional balancing. Shorter
scheduling may be necessary to justify tariff changes that would compensate providers
for speed and accuracy.
• Research into the potential impacts of utilizing energy storage for needed regulation
over existing fossil fuel‐based power plants. This effort could include a cost study on
new, fast, and efficient natural gas plants and how these options compare for integrating
renewables versus energy storage devices.
By undertaking these research projects, the California Energy Commission and other entities
can help the state prepare for the future needs that energy storage technologies can help
address.
6.2 Policy Drivers for Energy Storage Deployment
The CPUC and the Energy Commission should consider developing and implementing policies
to eliminate barriers to energy storage deployment that prevent cost‐effective deployment.
6.2.1 CPUC and AB 2514 Target Potential
The setting of energy storage targets engenders significant controversy. The state’s investor‐
owned utilities oppose any targets for energy storage, believing that such targets may result in
increased costs for ratepayers and the selection of energy storage technologies that may not be
as cost effective as competing solutions. They also cite the need for more pilot and
demonstration projects to provide additional data and certainty regarding investments in
energy storage. These and other stakeholders fear that the setting of targets may cannibalize
options that may be more cost effective and promising, such as automated demand response
technologies.
Conversely, proponents of energy storage targets believe that only a strong mandate from the
state will overcome the inertia and resistance to adopting energy storage technologies. These
proponents cite a complex and overlapping regulatory structure that serves to disfavor energy
storage technologies that would provide critical benefits for the state that are not currently
185
monetized under this system. In addition, they cite institutional barriers at the regulatory
agencies and utilities that make energy storage less competitive compared to traditional assets.
A broad mandate, based on an assessment of the grid needs that energy storage could meet,
could force changes on this byzantine system at a faster rate than the business‐as‐usual scenario.
Regardless of the merits of each side, the CPUC has a legislative mandate to evaluate the
feasibility and effectiveness of setting energy storage targets, if any. Should the agency decide
to set these targets, this report seeks to inform how such targets might function. As a general
recommendation, the CPUC should look to other proceedings for guidance on target setting.
For example, several proceedings have taken a two‐tier or two‐track approach to goal setting,
which may provide an example for energy storage target structuring under AB 2514. The
CPUC should consider a two‐track target with a more ambitious, shorter‐term target, possibly
based on or informed by applications that have already proven to be cost effective and suitable
to energy storage, and a longer‐term target, oriented toward applications that only can be cost
effective with certain emerging technologies that require substantial new R and D or time for
data collection. Examples of this approach include the permanent load shifting study
referenced in Chapter 2 that recommends near‐term and longer‐term targets based on
technology stage. In addition, the California ISO Renewable Integration: Market and Product
Review phases are broken into near‐term and longer‐term solutions. The key to these solutions
is the link to the market design changes being considered in California ISO phase 2 and the
potential positive impacts on the cost effectiveness of affected energy storage technologies.
These design changes include pay‐for‐performance regulation, load‐following reserve
requirement, a methodology to allocate the costs of integrating increased amounts of variable
renewable energy resources, and modifications to intra‐day market settlements.
A two (or more) track approach could also function well with application‐based targets, in
which the CPUC 1) makes necessary changes to remove barriers to competition for the
procurement of energy storage under a variety of business models and 2) selects a valuation
methodology that allows LSEs to procure cost‐effective resources for system need. To jumpstart
the valuation process, the CPUC should engage early with the authors of existing cost‐benefit
and valuation studies (such as EPRI, Sandia, and SCE) and other stakeholders to begin the
valuation phase of the AB 2514 proceeding as soon as possible, if not concurrently with the
ongoing first stage. In addition, since AB 2514 emphasizes the environmental benefits of energy
storage (such as through reduced greenhouse gas emissions and reduced peak power plant
reliance), the CPUC should consider a determination of cost effectiveness under the statute as
including the value of various societal and environmental benefits. Inclusion of these factors
could alter the cost effectiveness conclusions, given that none of the studies cited in this report
quantified those benefits.
186
6.2.2 Additional Policies for Energy Storage Deployment
These steps could include adding energy storage to the loading order as a preferred resource
over fossil fuel (potentially depending on their specific application), allowing energy storage
technologies to participate in the RA program, providing long‐term contract capability
(especially for capacity‐based applications but possibly for regulation, which would require
cooperation from California ISO over current scheduling practices), encouraging tariff changes
at California ISO and FERC, and continuing R and D funding and demonstration projects.
In addition, as discussed in Chapter 2, policymakers should consider allowing energy storage
technologies to play a larger role in transmission planning, congestion mitigation, and serve as
an alternative or deferral option to transmission upgrades. FERC may need to classify energy
storage as a transmission asset to help resolve this issue. If FERC decides against allowing
energy storage to qualify as a transmission asset, transportable or modular energy storage and
DESS on the distribution grid may serve a similar function that could be monetized.
Finally, policymakers should be aware of the important role that federal tax credits have played
in stimulating renewable energy development. These incentives should be renewed with the
potential for expansion to energy storage through federal legislation. Overall, continued R and
D investment in energy storage is important to lower technology costs and to better inform
decision‐makers about the future of various technologies.
6.2.3 Agency Coordination
Various agencies involved in the electricity system should ensure they are engaged in ongoing
collaboration to coordinate energy storage policies. The Energy Commission, through the IEPR,
can recommend ways that California energy agencies can work together toward common goals.
Stakeholders providing comments to this report have repeatedly indicated that agencies
occasionally work at cross‐purposes, stunting opportunity for growth and progress in the
energy storage field. To the extent this sentiment is accurate, agencies should therefore
collaborate to align market design and policy changes to the direction set forth in the 2011 IEPR.
For example, if the IEPR concludes that low‐high peak differentials and price volatility increases
the opportunity for participation from non‐generating technologies in the market, then
deployment of energy storage, implementation of dynamic pricing, regulations for EV charging,
and other relevant policies should align to that goal.
6.3 Conclusion
187
The subject of energy storage involves numerous technologies, applications, regulatory entities,
and grid processes. Given the state’s long‐term goals for integrating variable renewable energy
into the grid and decreasing greenhouse gas emissions, policy makers must determine how
energy storage can best fit with these future grid needs. These leaders should begin by
identifying the critical grid needs that energy storage technologies could address, developing a
method for valuing the various energy storage technologies in these applications, and
evaluating the policy options available to them to increase deployment of energy storage where
the technologies, market value, and locational and environmental benefits offer the most value
compared to the alternatives. With this framework, policy makers can determine how
California can achieve an appropriate and cost‐effective deployment of energy storage that
meets all of the state’s energy and environmental goals.
GLOSSARY
AA Advanced Adiabatic
AFC Alkaline Fuel Cells
ARB California Air Resources Board
ARPA‐E Energy Advanced Research Projects Agency
ARRA American Recovery And Reinvestment Act
BASE Beta‐Alumina Solid Electrolyte
CAES Compressed Air Energy Storage
CCSE California Center For Sustainable Energy
CESA California Energy Storage Alliance
CPUC California Public Utilities Commission
CSI California Solar Initiative
CREB Clean Renewable Energy Bond
CLPS Closed Loop Pumped Storage Projects
CSP Concentrating Solar Power
DENA German Energy Agency
DESS Distributed Energy Storage Systems
DOE Department Of Energy
EDLC Electric Double‐Layer Capacitor
EPRI Electric Power Research Institute
ERCOT Electric Reliability Council Of Texas
EES Electrochemical Energy Storage
EAP Energy Action Plans
FERC Federal Energy Regulatory Commission
FES Flywheel Energy Storage
188
GE General Electric
GHG Green House Gasses
GIP Generator Interconnection Procedures
GPM Gravity Power Moduletm
HVAC Heating, Ventilation, And Air Conditioning
HTF Heat‐Transfer Fluid
HTS High Temperature Superconductor
ISO Independent System Operator
IEPR Integrated Energy Policy Report
IOU Investor Owned Utilities
ITC Investment Tax Credit:
LADWP Los Angeles Depʹt Of Power And Water
LESR Limited Energy Storage Resource
LSE Load‐Serving Entities
LTPP Long Term Procurement Planning
LEV Low Emission Vehicle
MISO Midwest ISO
METI Ministry Of Economy, Trade And Industry (Japan)
MCFC Molten Carbonate Fuel Cells
NREL National Renewable Energy Laboratory
NEDO New Energy And Industrial Technology Development Organization (Japan)
NOPR Notice Of Proposed Rulemaking
O and M Operating And Maintenance
OIR Order Instituting Rulemaking
OEM Original Equipment Manufacturer
P G and E Pacific Gas And Electric Company
PNNL Pacific Northwest National Laboratory
PIRP Participating Intermittent Renewables Program
PLS Permanent Load Shifting
PAFC Phosphoric Acid Fuel Cells
PJM Pjm Interconnection
PHEV Plug‐In Hybrid Electric Vehicle
PEM Polymer Electrolyte Membrane
PPA Power Purchase Agreements
PTC Production Tax Credit
POU Publicly Owned Utilities
PV Photovoltaic
QECB Qualified Energy Conservation Bond
QF Qualifying Facilities
189
RA Resource Adequacy
R and D Research And Development
RAES Regenerative Air Energy Storage
REM Regulation Energy Management
RPS Renewable Portfolio Standard
RA Resource Adequacy
SMUD Sacramento Municipal Utilities District
S D G and E San Diego Gas And Electric Co.
SGIP Self Generation Incentive Program
STRC Societal Total Resource Cost
SOFC Solid Oxide Fuel Cells
SCE Southern California Edison
SMB Superconducting Magnetic Bearings
T and D Transmission And Distribution
TES Thermal Energy Storage
TMO Transition Metal Oxides
TOU Time Of Use
TRC Total Resource Cost
UPS Uninterruptible Power Supply
VER Variable Energy Resources
WECC Western Electricity Coordinating Council
ZEV Zero Emission Vehicle
190
REFERENCES
Abraham, K.M., Z. Jiang. 1996. ʺA Polymer Electrolyte‐Based Rechargeable Lithium/Oxygen
Battery.ʺ Journal of the Electrochemical Society, 1996: 1‐5.
AES Energy Storage. 2011. “ERCOT Energy Storage Registration/Testing Issues.” Presented at
ERCOT Emerging Technologies Working Group Meeting. January 2011.
http://www.ercot.com/content/meetings/etwg/keydocs/2011/0105/AES_ES_ERCOT_Storage_Re
gistration_and_Testing.pdf.
American Recovery and Reinvestment Act of 2009. Pub. Law 115‐5. 123 Stat. 115 at 345‐47.
February 17, 2009.
Armand, M., J.M. Tarascon. 2008. ʺBuilding Better Batteries.ʺ Nature, 2008: 652‐657.
Assembly Bill 2514 (Skinner, Chapter 469, Statutes of 2010).
Assembly Bill 1150 (V. Manuel Pérez, Chapter 310, Statutes of 2011).
Assembly Bill 32 (Núñez, Chapter 488, Statutes of 2006).
Australian Department of the Environment, Water, Heritage and the Arts. 2009. Smart Grid,
Smart City: A New Direction for a New Energy Era.
http://www.ret.gov.au/energy/Documents/smart‐grid/smartgrid‐newdirection.pdf.
Australian Department of Resources, Energy and Tourism, Advanced Electricity Storage
Program.
http://www.ret.gov.au/energy/clean/cei/advanced_electricity_storage_technologies_program/Pa
ges/AdvancedElectricityStorageTechnologies.aspx.
Ballard Power Systems, Inc. “CLEARgenTM Specification Sheet.” SPC5103076‐0B.
Ballard Power Systems Inc. 2009. “Notice of Annual Meeting, Management Proxy Circular and
2009 Annual Report.”
Ballard Power Systems, Inc. 2011. “Smarter Solutions for a Clean Energy Future.” March 8, 2011.
Beacon Power. http://www.beaconpower.com/ (accessed: March 2011).
Beck, F., P. Rüetschi. 2000. ʺRechargeable Batteries with Aqueous Electrolytes.ʺ Electrochimica
Acta, 2000.
Bitterly JG. 1998. Flywheel technology. IEEE Aerospace and Electronic Systems Magazine.
191
Buenoa, C. and J.A. Cartab. 2006. “Wind Powered Pumped Hydro Storage System: A Means of
Increasing the Penetration of Renewable Energy in the Canary Islands.” Renewable and
Sustainable Energy Reviews, 10: 312–340.
Boeing Phantom Works: DOE Energy Storage Program, 2004 Peer Review.
Bolund, B., H. Bernhoff, M. Leijon. 2007. “Flywheel energy and power storage systems.”
Renewable and Sustainable Energy Reviews 11: 235‐258.
Bullough, C. 2004. ʺAA‐CAES: Integration of Wind Energy.ʺ Alstom Power Technology Centre,
United Kingdom.
Bullough, C., Gatzen, C., Jakiel, C., Koller, M., Nowi, A., Zunft, S. 2004. ʺAdvanced Adiabatic
Compressed Air Energy Storage for the Integration of Wind Energy.ʺ Proceedings of the European
Wind Energy Conference. London, United Kingdom.
California Energy Storage Alliance. 2010. Comments of the California Energy Storage Alliance to
Order Instituting Rulemaking Pursuant to AB 2514 to Consider the Adoption of Procurement Targets
for Viable and Cost‐Effective Energy Storage Systems. January 21, 2010. Appendix A.
http://www.storagealliance.org/workfilings/2011‐01‐21‐
CESA_Storage_Scoping_Ruling_Opening_Comments_FINAL.pdf.
California Air Resources Board. 2008. Climate Change Scoping Plan: a Framework for Change.
December 2008.
California Air Resources Board. 2009. White Paper: Summary of Staff’s Preliminary Assessment of
the Need for Revisions to the Zero Emission Vehicle Regulation. November 25, 2009.
http://www.arb.ca.gov/msprog/zevprog/2009zevreview/zevwhitepaper.pdf.
California Air Resources Board. 2010a. Proposed Regulation for California Renewable Electricity
Standard: Initial Statement of Reasons. June 2010.
California Air Resources Board. 2010b. Staff Report: Initial Statement of Reasons, Proposed
Regulation to Implement the California Cap‐and‐Trade Program, Part I, Vol. I. October 28, 2010.
California Air Resources Board. 2010c. “ZEV Regulation Workshop.” Sacramento, California.
May 3, 2010. http://www.arb.ca.gov/msprog/zevprog/2011zevreg/05_03_10pres.pdf.
California Council on Science and Technology. 2011. California’s Energy Future: the View to 2050.
Summary Report. May 2011.
California Energy Commission, Public Utilities Commission. 2005. Energy Action Plan II.
September 21, 2005.
192
California Energy Commission. 2005a. Implementing California’s Loading Order for Electricity
Resources. Staff Report. July 2005.
California Energy Commission, Public Utilities Commission. 2008. 2008 Energy Action Plan
Update. February 2008.
California Energy Commission. 2009. 2009 Integrated Energy Policy Report. CEC‐100‐2009‐003‐
CMF. December 2009.
California Energy Commission. 2010a. 2010 Integrated Energy Policy Report Update.
CEC‐100‐2001‐001‐CMF. January 2011.
California Energy Commission. 2010b. Solar Thermal Projects Under Review or Announced.
http://www.energy.ca.gov/siting/solar/index.html (accessed May 15, 2011).
California Energy Commission. 2010c. Energy Commission Awards $2 Million to PG&E for Battery
Storage Research. February 9, 2010. http://www.energy.ca.gov/releases/2010_releases/2010‐02‐
09_battery_storage.html.
California Energy Commission. 2011a. 2011 Integrated Energy Policy Report Scoping Order. Docket
# 11‐IEP‐1. August 31, 2010.
California Energy Commission. 2011b. Transcript: IEPR Committee Workshop on Energy Storage for
Renewable Integration. http://www.energy.ca.gov/2011_energypolicy/documents/2011‐04‐
28_workshop/2011‐04‐28_Transcript.pdf.
California Environmental Protection Agency, Air Resources Board, Energy Commission, Public
Utilities Commission, California Independent System Operator. 2006. California’s Clean Energy
Future – Collaborative Vision for 2020. September 27, 2006.
California Independent System Operator. 2006. PIRP FAQs – Part 1. July 26, 2006.
http://www.caiso.com/17d0/17d0c24717130.html.
California Independent System Operator. 2009. 2012‐2014 Local Capacity Technical Analysis:
Report and Study Results. December 31, 2009. http://www.caiso.com/2495/2495c63b23450.pdf.
California Independent System Operator. 2010a. Discussion Paper, Renewable Integration: Market
and Product Review. July 8, 2010.
California Independent System Operator. 2010b. Attachment B, Non Generator Resources; Fourth
Replacement California ISO Tariff. May 18, 2010.
193
California Independent System Operator. 2010c. Revised Draft Final Proposal for Participation of
Non‐Generator Resources in California ISO Ancillary Services Markets. March 8, 2010.
California Independent System Operator. 2010d. Smart Grid Roadmap: IS‐1 ISO Uses Energy
Storage for Grid Operations and Control. November 12, 2010.
California Independent System Operator. 2011a. “Summary of Preliminary Results of 33percent
Renewable Integration Study.” Presentation for the 2010 CPUC Long Term Procurement Planning.
Docket R‐10‐05‐006. May 10, 2011. http://www.caiso.com/2b73/2b73796015b90.pdf.
California Independent System Operator. 2011b. Revised Straw Proposal, Renewable Integration:
Market and Product Phase 1. February 17, 2011. http://www.caiso.com/2b27/2b27e9ea6d680.pdf.
California Independent System Operator. 2011c. Discussion & Scoping Paper on Renewable
Integration Phase 2. April 5, 2011. http://www.caiso.com/2b3d/2b3da2e92470.html.
California Independent System Operator, Board of Governors. 2011d. Decision on Regulation
Energy Management. February 3, 2011. http://www.caiso.com/2b1a/2b1acd6d20610.pdf.
California Independent System Operator. 2011e. Regulation Energy Management Draft Final
Proposal. January 13, 2011.
California Independent System Operator. 2011f. Draft Final Proposal: Generator Interconnection
Procedures Phase 2. May 27, 2011. http://www.caiso.com/2b8b/2b8be21d30d90.pdf.
California Independent System Operator. 2011g. “Economic Planning Study.” Presented at
2011/2012 ISO Transmission Plan Stakeholder Meeting of Unified Planning Assumptions and Study
Plan. April 7, 2011. Folsom, CA. http://www.caiso.com/2b57/2b57e58541f30.pdf.
California Independent System Operator. 2011h. “33 percent High Load Case – Preliminary
New Scenarios: Monthly Regulation and Load Following Requirements – High Load Case.
April 5, 2011. http://www.caiso.com/23bb/23bbc01d7bd0.html.
California Public Utilities Commission. Docket R10‐12‐007.
http://docs.cpuc.ca.gov/published/proceedings/R1012007.htm.
California Public Utilities Commission ‐ (a). “Distributed Generation in California.”
http://www.cpuc.ca.gov/PUC/energy/DistGen/.
California Public Utilities Commission – (b). “Renewable Distributed Energy Collaborative (Re‐
DEC).” http://www.cpuc.ca.gov/PUC/energy/Renewables/Re‐DEC.htm.
194
California Public Utilities Commission – (c). “About the California Solar Initiative.”
http://www.cpuc.ca.gov/PUC/energy/Solar/aboutsolar.htm.
California Public Utilities Commission – (d). “CSI Program Application for Other Solar
Technologies.” http://www.cpuc.ca.gov/PUC/energy/Solar/othersolar.htm.
California Public Utilities Commission. 2008. Decision Addressing Eligible Technologies Under the
Self‐Generation Incentive Program (SGIP) and Modifying the Process for Evaluating SGIP Program
Change Requests. November 24, 2008.
California Public Utilities Commission. 2009a. RPS Quarterly Report (Q4 2009).
California Public Utilities Commission. 2009b. Order Instituting Rulemaking to Consider
Alternative‐fueled Vehicle Tariffs, Infrastructure and Policies to Support California’s Greenhouse Gas
Emissions Reduction Goals. August 20, 2009.
California Public Utilities Commission. 2009c. Decision Adopting Demand Response Activities and
Budgets for 2009 through 2011. D09‐08‐027. August 20, 2009.
http://docs.cpuc.ca.gov/word_pdf/FINAL_DECISION/106008.pdf.
California Public Utilities Commission. 2010a. Order Instituting Rulemaking Pursuant to AB 2514
to Consider Adoption of Procurement Targets for Viable and Cost‐Effective Energy Storage Systems.
December 16, 2010.
California Public Utilities Commission. 2010b. Proposed Decision of Comm’r Ryan, Decision
Adopting Requirements for Smart Grid Deployment Plans Pursuant to Senate Bill 17 (Padilla, Chapter
327, Statutes of 2009). July 24, 2010.
California Public Utilities Commission. 2010c. Attachment 1: 2010 Demand Response Cost
Effectiveness Protocols. Docket R07‐01‐041. December 16, 2010.
California Public Utilities Commission. 2010d. Electric Energy Storage: an Assessment of Potential
Barriers and Opportunities. July 9, 2010.
California Public Utilities Commission. 2011a. “Progress Toward California’s Renewables
Portfolio Standard Goals,” Presented at Senate Oversight Hearing. February 1, 2011.
http://www.cpuc.ca.gov/NR/rdonlyres/B5320B26‐82A8‐45EB‐AD27‐
266FD4C09FF2/0/PresentationforOversightHearings_RPS_2111CPUC.pdf
California Public Utilities Commission. 2011b. Scoping Memo and Assigned Ruling of Commissioner
and Administrative Law Judge. Docket R10‐12‐007. May 31, 2011.
195
California Public Utilities Commission. 2011c. 2010 Long‐Term Procurement Plan System Analysis
Preliminary Results. Docket R10‐05‐006. May 9, 2011.
http://www.cpuc.ca.gov/NR/rdonlyres/DC4E126B‐B8D0‐41C5‐9CB5‐
081CCF98689E/0/LTPP_Apr29_Filing_E3_final_5911.pdf.
California Public Utilities Commission. 2011d. 2010 Long‐Term Procurement Plan System Analysis
Preliminary Results. Docket R10‐05‐006. May 9, 201.
http://www.cpuc.ca.gov/NR/rdonlyres/DC4E126B‐B8D0‐41C5‐9CB5‐
081CCF98689E/0/LTPP_Apr29_Filing_E3_final_5911.pdf.
Carta, J.A., J. Gonzalez, V. Subiela. 2003. “Operational analysis of an innovative wind powered
reverse osmosis system installed in the Canary Islands.” Solar Energy 75(2):153–168.
Cavallo, A. 2007. ʺControllable and affordable utility‐scale electricity from intermittent wind
resources and compressed air energy storage (CAES).” Energy 32. pp. 120‐127.
Chen, H., T.N. Cong, W. Yang, C. Tan, et al. 2009. ʺProgress in Electrical Energy Storage System:
a Critical Review.ʺ Progress in Natural Science. March 2009.
China Energy Storage Alliance. 2011. Industry Trends Newsletter. April 2011.
http://www.chinaenergystorage.org/do/bencandy.php?fid=57&id=311.
City of Los Angeles. 2010. Increasing Reliable Renewable Energy in Los Angeles. September 13, 2010,
http://mayor.lacity.org/PressRoom/PressReleases/LACITYP_011469.
Consumer Power and Conservation Financing Authority, Public Utilities Commission, Energy
Resources Conservation and Development Commission. 2003. Energy Action Plan. May 8, 2003.
Customized Energy Solutions. 2009. “Electric Energy Storage Technologies in ERCOT.”
Presented at the ERCOT Renewables Technology Working Group. December 7, 2009.
http://www.ercot.com/content/meetings/rtwg/keydocs/2009/20091207‐RTWG/Walawalkar‐
EnergyStorage‐ERCOT‐RTWG‐12‐2009.pdf.
Cutter, Eric, et al. 2009. “Impact of Market Rules on Energy Storage Economics.” Energy and
Environmental Economics.
Dell, R.M. and D.A.J. Rand, ʺClean Energy,ʺ Cambridge : Royal Society of Chemistry, c2004.
ISBN 0‐85404‐546‐5 (p. 187 ‐ 191).
Demachi, K., K. Miya, R. Takahata, et al. 2002. “Numerical evaluation of rotation speed
degradation of superconducting magnetic bearing caused by the electromagnetic phenomena,”
Physica C 378–381: 858–863.
196
Denholm, P., et al. 2010. The Role of Energy Storage with Renewable Electricity Generation. National
Renewable Energy Laboratory. Technical Report NREL/TP‐682‐47187. January 2010.
DeTeresa SJ. Materials for advanced flywheel energy storage devices. MRS Bull 1999.
Division of Ratepayer Advocates. 2011. “Comments of the Division of Ratepayer Advocates on
Order Instituting Rulemaking Pursuant to Assembly Bill 2514.”California Public Utilities
Commission. January 21, 2011. http://docs.cpuc.ca.gov/published/proceedings/R1012007.htm.
Douche, M.L., F. Novel‐Cattin, R. Durand, J.J. Rameau. 1997. ʺCharacterization of Different
Grades of Aluminum Anodes for Aluminum/Air Batteries.ʺ Journal of Power Sources, 1997: 197‐
205.
ECN abstract. 1994. Large Hydrogen Underground Storage. http://www.ika.rwth‐
aachen.de/r2h/index.php/Large_Hydrogen_Underground_Storage.
Edison International. 2010. Southern California Edison Finalizes Stimulus Grant to Start Large‐Scale
Energy Storage Demonstration. October 20, 2010.
http://www.edison.com/pressroom/pr.asp?id=7482.
Electric Power Research Institute. 2010a. Electrical Energy Storage Technology Options: A White
Paper Primer on Applications, Costs and Benefits. EPRI. Palo Alto, CA 1020676.
Electric Power Research Institute. 2010b. Functional Requirements for Electric Energy Storage
Applications on the Power System Grid. Palo Alto, California: 1020075.
Electric Power Research Institute and Southern California Edison. 2010. Smart Grid
Demonstration Initiative.
http://www.smartgrid.epri.com/doc/EPRI%20Smart%20Grid%20Demonstration%202‐
Year%20Update_final.pdf (accessed August 3, 2011).
Electric Reliability Council of Texas (ERCOT). 2010. Emerging Technologies Integration Plan
Quarterly Update for the Period Ending December 31, 2010.
http://www.ercot.com/content/meetings/tac/keydocs/2011/0303/11._ETIP_Quarterly_Report‐
4th_Quarter2010_Final_Revised_III_.doc.
Electric Reliability Council of Texas (ERCOT). 2011. ERCOT Wholesale Market Subcommittee,
Meeting Minutes. February 8, 9, 2011.
http://www.ercot.com/content/meetings/wms/keydocs/2010/0922/approved_minutes_wms_100
922.doc.
197
Electricity Storage Association (ESA).
http://www.electricitystorage.org/ESA/technologies/pumped_hydro/ (accessed April 2011).
Ellis, J. 2011. “Re: Docket #11‐IEP‐1N, Questions for Panelists at the April 28, 2011 Workshop,”
Comments to the April 28, 2011 IEPR Workshop. April 26, 2011.
http://www.energy.ca.gov/2011_energypolicy/documents/2011‐04‐
28_workshop/comments/Jack_Ellis_Comments_Energy_Storage.pdf.
Energy and Environmental Economics, Inc. and Strategen Consulting. 2010. Statewide Joint IOU
Study of Permanent Load Shifting. December 1, 2010.
Energy Independence and Security Act of 2007 (42 U.S.C. Sec. 17001, et seq.).
Executive Office of the President, National Science and Technology Council. 2011. A Policy
Framework for the 21st Century Grid: Enabling Our Secure Energy Future. June 13, 2011.
Executive Order No. S‐14‐08. November 17, 2008.
Executive Order No. S‐21‐09. September 15, 2009.
Eyer, J. 2009. Electric Utility Transmission and Distribution Upgrade Deferral Benefits from Modular
Electricity Storage. Sandia National Laboratories for U.S. Department of Energy. June 2009.
Eyer, J., G. Corey. 2010. Energy Storage for the Electricity Grid: Benefits and Market Potential
Assessment Guide. Sandia National Laboratories for U.S. Department of Energy. SAND2010‐
0815. February 2010.
Federal Energy Regulatory Commission. Docket AD10‐13.
http://elibrary.ferc.gov/idmws/docket_sheet.asp.
Federal Energy Regulatory Commission. 2007. FERC Order 890. 72 FR ¶ 12,266. May 15, 2007.
Federal Energy Regulatory Commission. 2008a. FERC Order 719. 125 FERC ¶ 61, 071 (issued
October 17, 2008).
Federal Energy Regulatory Commission. 2008b. Order on Compliance. 125 FERC ¶ 61, 319 (issued
December 18, 2008).
Federal Energy Regulatory Commission. 2009. Order Accepting Tariff Revisions. 127 FERC ¶ 61,
135 (issued May 15, 2009).
Federal Energy Regulatory Commission. 2010. Request for Comments Regarding Rates, Accounting,
and Financial Reporting for New Electric Storage Technologies. 75 FR 122, 36381. June 25, 2010.
198
Federal Energy Regulatory Commission. 2011. “Frequency Regulation Compensation in the
Organized Wholesale Power Markets,” 134 FERC ¶ 61, 124. February 17, 2011.
Fertig, E. , S. Wagner. 2010. The Role of Energy Storage in Renewable Power Integration. Integration
and Policy Workshop for RenewElec Project. Carnegie Mellon University. October 21‐22, 2010.
Finkenrath, M., Pazzi, S., DʹErcole, M., Marquardt, R., Moser, P., Klafki, M., Zunft, S. 2009.
ʺStatus and Technical Challenges of Advanced Compressed Air Energy Storage (CAES)
Technology.” International Workshop on Environment and Alternative Energy. Munich, Germany.
Fioravanti, R., J. Enslin. 2007. Emissions Comparison for a 20 MW Flywheel‐based Frequency
Regulation Power Plant. KEMA‐Inc. May 18, 2007.
Florida Solar Energy Center. 2007. Hydrogen Basics – Production.
Fthenakis, V. 2008. Compressed Air Energy Storage (CAES) Scoping Workshop: Enabling Solar and
Wind Energy Technologies on a Grand Scale.
Fujiwara, N., M. Yao, Z. Siroma, H. Senoh, et al. 2011. ʺReversible air electrodes integrated with
an anion‐exchange membrane for secondary air batteries.ʺ Journal of Power Sources, 196: 808‐813.
Gabrys CW. 2001. High Performance Composite Flywheel. US patent Pub. No.: US 2001/0054856 A1.
December 27, 2001.
General Electric. 2010. GE Durathon Batteries. May 5, 2010 (accessed May 4, 2011).
Genta G. 1985. Kinetic energy storage: theory and practice of advanced flywheel systems. London,
Butterworths.
Germany Energy Agency (DENA). 2010. DENA Grid Study II, Integration of Renewable Energy
Sources into the German Power Supply System from 2015‐2020 with Outlook to 2025. November 2010.
http://www.dena.de/fileadmin/user_upload/Download/Dokumente/Studien___Umfragen/dena
_Grid_Study_II_‐_final_report.pdf.
Germany Federal Ministry of Economics and Technology, Federal Ministry for the
Environment, Nature Conservation, and Nuclear Safety. 2010. Energy Concept for an
Environmentally Sound, Reliable, and Affordable Energy Supply. September 28, 2010.
http://www.bmwi.de/English/Redaktion/Pdf/energy‐
concept,property=pdf,bereich=bmwi,sprache=en,rwb=true.pdf.
Gordes, J. N., A. L. Weisbrich, D. L. Rainey, P. W. Olson. 2000. “Offshore WARP Wind Power
and Integral H2—Gas Turbines or Fuel Cells.” Proceedings of the European Seminar on Offshore
199
Wind Energy in Mediterranean and Other European Seas (OWEMES) Conference. Syracusa, Sicily,
Italy. April 2000.
Greenblatt, J. B., Succar, S., Denkenberger, D. C., Williams, R. H., Socolow, R. H. 2007. ʺBaseload
wind energy: modeling the competition between gas turbines and compressed air energy
storage for supplemental generation.” Energy 35. pp. 1474‐1492.
Gunther, Erich. 2009. “Ice Bear Energy Storage System – 90 out of 100.”Smart Grid News.com.
February 2, 2009.
http://www.smartgridnews.com/artman/publish/reviews/Ice_Bear_Energy_Storage_System_90_
out_of_100.html.
Gyuk, I., S. Eckroad. 2003. EPRI‐DOE Handbook of Energy Storage for Transmission and Distribution
Applications. Electric Power Research Institute and U.S. Department of Energy.
Hassenzahl, W.V. 2001. IEEE Trans. Appl. Supercond. 11: 1653.
HDR Engineering, Inc. 2010. Hydroelectric Pumped Storage for Enabling Variable Energy Resources
within the Federal Columbia River Power System. Prepared for Bonneville Power Administration.
September 30, 2010
Huggins, R.A., 2010. Energy Storage. Springer‐Verlag GmbH. July 2010.
Hydrogenics. HySTATTM Hydrogen Station, Rev 1004.
Ibrahim, H., A. Ilinca, and J. Perron, ʺEnergy storage systems—Characteristics and
comparisons,ʺ Renewable and Sustainable Energy Reviews (2008), Volume: 12, Issue: 5,
Publisher: Elsevier, Pages: 1221‐1250.
Internal Revenue Code, Section 45. 26 U.S.C. Sec. 45.
Internal Revenue Code, Section 48. 26 U.S.C. Sec. 48.
ISO New England, Inc. 2010. Markets and Services Tariff, Appendix J, Alternative Technologies
Regulation Pilot Program. August 30, 2010.
Itron. 2010. Impacts of Distributed Generation: Final Report. Prepared for the California Public
Utilities Commission. January 2010.
Itron. 2011. Cost‐Effectiveness of Distributed Generation Technologies. February 9, 2011.
http://www.cpuc.ca.gov/PUC/energy/DistGen/sgip/proposal_workshops.htm.
200
Japan Ministry of Economy, Trade and Industry. 2010a. The Strategic Energy Plan of Japan:
Summary. Revised June 2010. http://www.meti.go.jp/english/press/data/pdf/20100618_08a.pdf.
Japan Ministry of Economy, Trade and Industry. 2010b. Proposals for Next Generation Energy and
Social System Demonstration Areas. April 2010. http://www.meti.go.jp/english/press/data/pdf/N‐
G%20System.pdf.
KEMA. 2007. Energy Time‐shift from Wind Generation Emissions Comparison for a 20 MW Flywheel‐
based Frequency Regulation Power Plant. Beacon Power Corp. KEMA Project: BPCC.0003.001.
January 8, 2007.
KEMA, Inc. 2010. Research Evaluation of Wind Generation, Solar Generation, and Storage Impact on
the California Grid. Prepared for the California Energy Commission.CEC‐500‐2010‐010. June
2010.
Kohari, Z., I. Vajda. 2005. “Losses of flywheel energy storages and joint operation with solar
cells.” Journal of Materials Processing Technology 161(1–2): 62–65.
Koshizuka N., et al. 2002. “Present status of R&D on superconducting magnetic bearing
technologies for flywheel energy storage system.” Physica C 378‐381: 11‐17.
Kristov, L. 2011. “Allocation of Integration Costs.” California ISO. Presented at the Renewable
Integration Phase 2 Stakeholder Meeting. April 12, 2011. Slides 25‐29.
http://www.caiso.com/2b5a/2b5ae96271cf0.pdf.
Kruse, B., Grinna, S., Buch, C. 2002. Hydrogen Status og Muligheter. Bellona.
LaCommare, K.H., J.H. Eto. 2004. Understanding the Cost of Power Interruptions to U.S. Electricity
Consumers. Ernest Orlando Lawrence Berkeley National Laboratory. September 2004.
http://certs.lbl.gov/pdf/55718.pdf.
Largeot, C., C. Portet, J. Chmiola, P. Taberna, et al. 2008. ʺRelation Between Ion Size and Pore
Size for an Electrical Double Layer Capacitor.ʺ Journal of the American Chemical Society, 130: 2730‐
2731.
Lin, J., et al. 2011. Energy Storage – a Cheaper, Faster & Cleaner Alternative to Conventional
Frequency Regulation. Strategen Consulting for the California Energy Storage Alliance. February
16, 2011.
Lipman, T.E., R. Ramos, D.M. Kammen. 2005. An Assessment of Battery and Hydrogen Energy
Storage Systems Integrated with Wind Energy Resources in California. California Energy
Commission, PIER Research. CEC‐500‐2005‐136. September 2005.
201
Liu, H., J. Jiang. 2007. “Flywheel energy storage—an upswing technology for sustainability.”
Energy and Buildings 39: 599‐604.
Long, T., D. Narayan, et al. 2000. “Simulation of the interaction between flywheel energy
storage and battery energy storage on the international space station.” Proceedings of the
Intersociety Energy Conversion Engineering Conference. Las Vegas, NV. July 2000. pp. 848–854.
MacKay, David J.C. 2007. Enhancing Electrical Supply by Pumped Storage in Tidal Lagoons.
University of Cambridge. March 2007.
Makansi, J. 2005. ʺEnergy Storage: Why so few pay attention and what we can do about it.ʺ
EESAT. San Francisco, CA.
Makarov, Y.V., et al. 2008. Assessing the Value of Regulation Resources Based on Their Time Response
Characteristics. Pacific Northwest National Laboratory. June 2008.
Marshal, J. 2009a. “PG&E Opts for Energy Storage.”NEXT100. August 26, 2009.
http://www.next100.com/2009/08/pge‐opts‐for‐energy‐storage.php.
Marshal, J. 2009b. “PG&E’s Big Battery.”NEXT100. February 16, 2010.
http://www.next100.com/2010/02/pges‐big‐battery.php.
MWH. 2009. Technical Analysis of Pumped Storage and Integration with Wind Power in the Pacific
Northwest. Prepared for U.S. Army Corps of Engineers, Northwest Division, Hydroelectric
Design Center. August 2009.
Miller, M. 2011. “Modifications to Intra‐Hour Scheduling & Other Market Design
Enhancements,” California Independent System Operator. April 12, 2011. pp. 21‐24.
http://www.caiso.com/2b5a/2b5ae96271cf0.pdf.
Mulcahy, T.M., J.R. Hull, K.L. Uherka, R.G. Abboud, J.J. Juna. 2011. IEEE Trans. Appl. Supercond
11: 1729.
Nair, N.C., N. Garimella. 2010. ʺBattery energy storage systems: Assessment for small‐scale
renewable energy integration.ʺ Energy and Buildings. November 2010.
Nagaya, S., N. Kashima, M. Minami, H. Kawashima, S. Unisuga. 2001. IEEE Trans. Appl.
Supercond 11: 1649.
Nakama, H. 2010. “NEDO Research Related to Large‐scale PV‐related Grid‐connection
Projects.” New Energy and Industrial Technology Development Organization (NEDO).
http://www.sandia.gov/regis/presentations/H_Nakama‐NEDO.pdf.
202
Nakhamkin, M. 2008. ʺSecond Generation of the CAES Technology.ʺ Compressed Air Energy
Storage Scoping Workshop. Columbia University.
Nakhamkin, M., Chiruvolu, M., Patel, M., Byrd, S., Schainker, R., Marean, J. 2009. ʺSecond
Generation of CAES Technology ‐ Performance, Operations, Economics, Renewable Load
Management, Green Energy.ʺ Power‐Gen International. Las Vegas, NV.
Nakai, H., A. Matsuda, M. Suzuki. 2001. “Development and testing of the suspension system for
a flywheel battery.” Control Engineering Practice 9: 1039–1046.
National Renewable Energy Laboratory. Parabolic Trough Thermal Energy Storage Technology.
http://www.nrel.gov/csp/troughnet/thermal_energy_storage.html (accessed June 20, 2011).
Neburchilov, V., H. Wang, J. Martin, W. Qu. 2010. ʺA Review on Air Cathodes for Zinc‐Air Fuel
Cells.ʺ Journal of Power Sources, 2010: 1271‐1291.
New Energy and Industrial Technology Development Organization (NEDO). “Energy and
Environmental Technology Projects.”
http://www.nedo.go.jp/english/activities_nedoprojects_energy.html#sb (accessed June 30, 2011).
New York Independent Service Operator. 2010. Energy Storage in the New York Electricity
Markets. March 2010.
Nexight Group and Sandia National Laboratories. 2010a. ʺAdvanced Materials and Devices for
Stationary Electrical Energy Storage Applications.ʺ U.S. Department of Energy. December 2010.
Nexight Group and Sandia National Laboratories. 2010b. “Electric Power Industry Needs for
Grid‐Scale Storage Applications.” U.S. Department of Energy. December 2010.
Nagaya, S., N. Kashima, M. Minami, H. Kawashima, S. Unisuga. 2001. IEEE Trans. Appl.
Supercond 11: 1649.
NGK Insulators. NaS Batteries. http://www.ngk.co.jp/english/products/power/nas/index.html
(accessed May 5, 2011).
Nguyen, T., R. Savinell. 2010. ʺFlow Batteries.ʺ The Electrochemical Society Interface. Fall 2010.
NYSERDA/DOE Joint Energy Storage Initiative Demonstration Projects. March 16, 2010.
http://www.storagemonitoring.com/nyserda‐doe/storage‐home.shtml (accessed: March 2011).
Pacific Gas and Electric Company. 1997. Thermal Energy Storage Strategies for Commercial HVAC
Systems.
203
Pdc Machines. “Diaphragm Compressors.”
http://www.pdcmachines.com/data/pdf/Compressors%20Formated.pdf.
Pieronek, Thomas, J., D. Decker, A.K. Spector Victor. 1997. “Spacecraft flywheel systems—
benefits and issues, in: National Aerospace and Electronics Conference.” Proceedings of the IEEE.
Dayton, USA. July 1997. pp. 589–593.
PJM Interconnection, LLC. 2010a. Alternative Technology Resource Pilot Program, Project
Guidelines. April 2010. Section 3.3.1.2. http://www.pjm.com/~/media/markets‐ops/atr/atr‐
guidelines.ashx.
PJM Interconnection, LLC. 2010b. Regulation Market Incentive Payment: Problem Statement.
http://www.pjm.com/~/media/committees‐groups/committees/mrc/20110119/20110119‐item‐05‐
regulation‐market‐performance‐incentive‐problem‐statement‐clean.ashx.
PJM Interconnection, LLC. 2010c. New Energy Storage Resource and Revised Station Power
Definition. July 6, 2010.
Popp, M. “Storage for secure power supply from wind and sun.” www.poppware.de.
Assembly Bill 1150 (V. Manuel Pérez, Chapter 310, Statutes of 2011). Bill Analysis.
http://www.leginfo.ca.gov/pub/11‐12/bill/asm/ab_1101‐
1150/ab_1150_cfa_20110525_101450_asm_comm.html (accessed June 30, 2011).
Proton Energy Systems. HOGEN Hydrogen Generating Systems, H Series Hydrogen
Generators Technical Specifications. PD‐0600‐0062 Rev. B, 01.09.
Public Utility Commission of Texas. 2011. Report to the 82nd Texas Legislature: Scope of Competition
in Electric Markets in Texas. January 2011.
Purtz, J. 2010. Submarine Adiabatic and Isobaric Compressed Air Energy Storage.
Rawson, M. 2011. “Utilities’ Perspective on Energy Storage.” Sacramento Municipal Utility
District. Presented at California Energy Commission 2011 IEPR Workshop. Sacramento, California.
April 28, 2011. http://www.energy.ca.gov/2011_energypolicy/documents/2011‐04‐
28_workshop/presentations/20_SMUD_Rawson_Panel_3.pdf.
Recovery.gov . “Grants‐Award Summary – Pacific Gas & Electric.”
http://www.recovery.gov/Transparency/RecipientReportedData/pages/RecipientProjectSummar
y508.aspx?AwardIdSur=75561&AwardType=Grants (accessed May 9, 2011).
REN21. 2008. Renewables 2007 Global Status Report. Paris: REN21 Secretariat, and Washington,
DC: Worldwatch Institute.
204
Ries, G., H.W. Neumueller. 2001. “Comparison of energy storage in flywheels and SMES.”
Physica C 357–360: 1306–1310.
Rittershausen, J., M. McDonagh. 2010. Moving Energy Storage from Concept to Reality: Southern
California Edison’s Approach to Evaluating Energy Storage. Southern California Edison.
RWE. 2011. http://www.rwe.com/web/cms/en/365478/rwe/innovations/power‐
generation/energy‐storage/compressed‐air‐energy‐storage/project‐adele/.
RWE. 2010. ADELE ‐ Adiabatic Compressed‐Air Energy Storage for Electricity Supply.
Sandia National Laboratories. National Solar Thermal Test Facility.
http://www.sandia.gov/Renewable_Energy/solarthermal/NSTTF/salt.htm (accessed April 2011).
Sandia National Laboratories. 2010. Energy Storage Systems Projects. March 16, 2010.
http://www.sandia.gov/ess/About/projects.html (accessed: March 2011).
San Diego Gas & Electric. 2010a. “Alan B. Marcher – Testimony on Electricity Distribution
Capital Costs.” Application of SDG&E for authority to update its gas and electric revenue requirement
and base rates effective on January 1, 2012. December 2010. SDG&E Doc #249150.
San Diego Gas & Electric. 2010b. “Lee. S Krevat – Testimony on Smart Grid Policy.” Application
of SDG&E for authority to update its gas and electric revenue requirement and base rates effective on
January 1, 2012. December 2010. SDG&E Doc #249827.
San Diego Gas & Electric. 2010c. “Kathleen H. Cordova – Testimony on Customer Services—
Information.” Application of SDG&E for authority to update its gas and electric revenue requirement
and base rates effective on January 1, 2012. December 2010. SDG&E Doc #249538.
San Diego Gas & Electric. 2010d. “Thomas Bialek – Testimony on Customer Services—
Information.” Application of SDG&E for authority to update its gas and electric revenue requirement
and base rates effective on January 1, 2012. December 2010. SDG&E Doc #249440.
San Diego Gas & Electric Co. 2011a. SDG&E Proposes Adding 450MW of Local “Peaking” Power.
May 23, 2011.
http://public.sempra.com/newsreleases/viewpr.cfm?PR_ID=2607&Co_Short_Nm=SDGE.
San Diego Gas & Electric. 2011b. Smart Grid Deployment Plan, 2011‐2020. June 6, 2011.
http://www.sdge.com/regulatory/A11‐06‐006.shtml.
Schiess, K. 2009. “TES Testimony.” Presented to California Public Utility Commission.
205
Schiess, K. 2011. “Energy Storage for Renewable Integration: Comments to IEPR Workshop of
April 28, 2011.” Comments to the April 28, 2011 IEPR Workshop. KSEngineers.
http://www.energy.ca.gov/2011_energypolicy/documents/2011‐04‐
28_workshop/comments/TN_60630_04‐28‐
11_KSEngineers_Comments_on_the_April_28_2011_IEPR_Works.pdf.
Schoenung, S. 2011. “Economic Analysis of Large‐Scale Hydrogen Storage for Renewable Utility
Applications.” International Colloquium on Environmentally Preferred Advanced Power Generation.
Costa Mesa, California. February 8‐10, 2011.
Schueneman, T. 2011. “Gravity Power Module Revolutionizes Pumped Hydro Energy Storage,”
Triple Pundit. March 11, 2011. http://www.triplepundit.com/2011/03/gravity‐power‐module‐
aims‐revolutionize‐pumped‐hydro‐energy‐storage/.
Scrosati, B., J. Garche. 2010. ʺLithium Batteries: Status Prospects and Future.ʺ Journal of Power
Sources.
Senate Bill 2 (Simitian, Chapter 1, Statutes of 2011).
Senate Bill 107 (Simitian, Chapter 464, Statutes of 2006).
Senate Bill 17 (Padilla, Chapter 327, Statutes of 2009).
Sioshansi, R., P. Denholm. 2010. “The Value of Concentrating Solar Power and Thermal Energy
Storage.” National Renewable Energy Laboratory. NREL‐TP‐6A2‐45833. February 2010.
Shakhtur, A. 2011. “GE Builds a High‐Tech Fossil Fuel Power Plant for the Renewable Energy
Age.” Fundamentalmente Energia. May 29, 2011. http://alishakhtur.com/2011/05/29/ge‐builds‐a‐
high‐tech‐fossil‐fuel‐power‐plant‐for‐the‐renewable‐energy‐age/.
Siems, S.O., W.‐R. Canders, H. Walter, J. Bock. 2004.. Supercond. Sci. Technol. 17: S229.
Simon, P., Y. Gogotsi. 2008. ʺMaterials for Electrochemical Capacitors.ʺ Nature Materials, 2008:
845‐854.
Southern California Edison. 2010a. “Generation‐related Project Development Division (PDD) ‐
Testimony of R. Harding and M. Nelson.” 2012 General Rate Case, Vol. 10. November 2010.
Southern California Edison. 2010b. “Generation‐Catalina Operation and Maintenance (O&M)
Expenses And Capital Expenditures ‐ Testimony of R. Hite.” 2012 General Rate Case, Vol. 11.
November 2010.
206
Southern California Edison. 2010c. “T&D Business Unit—Advanced Technology ‐ Testimony of
D. Kim.” 2012 General Rate Case, Vol. 2. November 2010.
Sparta Township, NJ. 2004. “Riverbank Project ‐ Frequently Asked Questionsʺ. Sparta
Township, NJ – Official Website. http://www.spartanj.org/cit‐e‐
access/webpage.cfm?TID=30&TPID=9809 (accessed June 20, 2011).
Statoil. “Atmospheric Electrolysers: 10 Nm3/h ‐ 485Nm3/h.”
http://www3.statoil.com/hydrogentechnologies/svg03816.nsf?OpenDatabase.
Steward, D., Ramsden, T., Harrison, K. 2010. “Hydrogen for Energy Storage Analysis
Overview.” National Renewable Energy Laboratory. NREL/PR‐560‐48360. May 2010.
Swider, D. J. 2007. ʺCompressed Air Energy Storage in an Electricity System with Significant
Wind Power Generation.ʺ IEEE Transactions on Energy Conversion 22(1): 95‐102.
Texas Energy Storage Alliance. 2010. Storage Participation in ERCOT. January 3, 2010.
Texas Legislature Online. Proposed Senate Bill 943 (Carona).
http://www.capitol.state.tx.us/BillLookup/History.aspx?LegSess=82R&Bill=SB943.
Tobie, B., C. Liuchen, L. David. 2002. “A low speed flywheel system for wind energy
conversion.” Canadian Conference on Electrical and Computer Engineering. Winnipeg,
Manitoba. May 2002. pp. 251–256.
Tzeng, J., R. Emerson, P. Moy. 2006. “Composite flywheels for energy storage.” Composites
Science and Technology 66: 2520‐2527.
UC Berkeley Law, Center for Law, Energy & the Environment, and UCLA Law, Emmett Center
on Climate Change and the Environment. 2010. The Power of Energy Storage: How to Increase
Deployment in California to Reduce Greenhouse Gases. July 2010.
U.S. Department of Energy. 2009. Annual Progress Report – Vehicle Technologies Program. January
2009. http://www1.eere.energy.gov/vehiclesandfuels/pdfs/program/2008_energy_storage.pdf
(accessed: March 2011).
U.S. Department of Energy. “Flywheel Energy Storage: an alternative to batteries for
uninterruptible power supply systems.” Federal Technology Alert. DOE/EE‐0286.
U.S. Department of Energy. 2011a. Quadrennial Technology Review Framing Document. March 14,
2011. http://www.energy.gov/qtr/documents/DOE‐QTR_Framing.pdf.
207
U.S. Department of Energy. 2011b. Energy Storage Program Planning Document. February 2011.
http://www.oe.energy.gov/DocumentsandMedia/OE_Energy_Storage_Program_Plan_Feburary
_2011v3.pdf.
U.S. Department of Energy, Electricity Advisory Committee. 2011c. Energy Storage Activities in
the United States Electricity Grid. May 2011.
http://www.oe.energy.gov/DocumentsandMedia/FINAL_DOE_Report‐Storage_Activities_5‐1‐
11.pdf.
U. S. Department of Energy, Office of Energy Efficiency & Renewable Energy. February 2011.
http://www1.eere.energy.gov/hydrogenandfuelcells/fuelcells/fc_types.html#comparison.
U.S. Treasury Department. 2011. Payments for Specified Energy Property in Lieu of Tax Credits
Under the American Recovery and Reinvestment Act of 2009. April 2011.
http://www.treasury.gov/initiatives/recovery/Documents/Guidance%203‐29‐
11%20revised%20(2)%20clean.doc.
Vit, B. M. Beatty Scott, B.J. DeBlonk, et al. 2004. “A review of technology developments in
flywheel attitude control and energy transmission systems.”2004 IEEE Aerospace Conference
Proceedings. Big Sky, United States. March 2004. pp. 2784–2800.
Wagner, R. 2005. ʺHigh‐Power Lead‐Acid Batteries for Different Applications.ʺ Journal of Power
Sources. June 2005.
Wagner, RC., R. David Boyle, D. Kent. 2002. “Commercialization of flywheel energy storage
technology on the international space station.” Proceedings of the Intersociety Energy Conversion
Engineering Conference. 37th Intersociety Energy Conversion Engineering Conference, IECEC.
Washington, DC. July 2002. pp. 146–150.
Western Power Trading Forum. 2011. “Comments of the Western Power Trading Forum on
Order Instituting Rulemaking Pursuant to AB 2514.”January 21, 2011.
http://docs.cpuc.ca.gov/EFILE/CM/129824.htm.
Woodbank Communications. Battery Comparison Chart.
http://www.mpoweruk.com/specifications/comparisons.pdf (accessed April 22, 2011).
Woody, Todd. 2010. “In California’s Mojave Desert, Solar‐Thermal Projects Take Off.” Yale
Environment 360. October 27, 2010.
Yang, Z., J. Zhang, M. Kintner‐Meyer, X. Lu, et al. 2010. ʺElectrochemical Energy Storage for
Green Grid.ʺ Chemical Reviews. September 2010.
208
209
Younicos. The Island Test Bay: Autonomous Supply of Complete Island Grid From Renewable Energy
Sources. February 2009.
Zunft, S., Jakiel, C., Koller, M., and Bullough, C. 2006. ʺAdiabatic Compressed Air Energy
Storage for the Grid Integration of Wind Power.ʺ 6th International Workshop on Large‐Scale
Integration of Wind Power and Transmission Networks for Offshore Windfarms. Netherlands.
