Energy Storage A Nontechnical Guide - Compress

Energy Storage
A Nontechnical Guide
By Richard Baxter
Disclaimer. The recommendations, advice, descriptions, and the methods in this
book are presented solely for educational purposes. The author and publisher assume
no liability whatsoever for any loss or damage that results from the use of any of the
material in this book. Use of the material in this book is solely at the risk of the user.
Copyright© 2007 by
PennWell Corporation
1421 South Sheridan Road
Tulsa, Oklahoma 74112-6600 USA
800.752.9764
+1.918.831.9421
[email protected]
www.pennwellbooks.com
www.pennwell.com
Director: Mary McGee

Managing Editor: Stephen Hill
Production / Operations Manager: Traci Huntsman
Production Editor: Tony Quinn
Senior Design Editor: Robin Remaley
Book Designer: Susan E. Ormston Thompson
Cover Designer: Sheila Brock
Library of Congress Cataloging-in-Publication Data
Azar, Jamal J., 1937-

Drilling engineering / J.J. Azar, G. Robello Samuel.
p. cm.
ISBN-13: 978-1-59370-072-0 (hardcover)
ISBN-10: 1-59370-072-5 (hardcover)
1. Oil well drilling. 2. Gas well drilling. I. Samuel, G. Robello. II. Title.
TN871.2.A92 2007
622'.3381--dc22
2006034984
All rights reserved. No part of this book may be reproduced, stored in a retrieval system,
or transcribed in any form or by any means, electronic or mechanical, including photo-
copying and recording, without the prior written permission of the publisher.
Printed in the United States of America
3 4 5 6 7 12 11 10 09 08

Foreword
Electricity is the most useful and flexible of all energy sources. To provide
this capability, the power industry in modern industrialized societies developed
power stations of various sizes and capabilities to provide a continuous, reliable,
and affordable supply of electricity as the demand varied on daily, weekly, and
seasonal cycles. Lately however, this centrally organized and controlled market
design has become unstable. This has caused investment for new large power
stations to become riskier as repaying their development costs can no longer be
guaranteed through assured power sales in a highly regulated market.
Solutions to this challenge follow one of two competing strategies. The first
is simply to continue extending the power transmission grid in order to open up
additional markets for these new generation facilities. The second is to focus on a
distributed supply strategy reliant upon smaller and distributed power generation
and energy storage resources to provide a more stable and secure electricity
supply. Each of these strategies implies a different direction for the future of
the power industry. The first strategy represents a continued centralization of
power production to offset increasing transmission infrastructure costs, whereas
the second strategy represents a focus on local production and management of
electricity to avoid excessive infrastructure build-out and grid management costs.
This second option is the most promising one as it enables significant progression
toward improvements of energy efficiency, leads to enhanced energy security,
and—above all—promotes wind and solar power for the electricity supply.
Richard Baxter shows in his book how diverse the possibilities of electricity
storage really are. These technologies have been widely overlooked by the power
industry for many years as the industry’s focus has been fixated on large-scale
supply strategies. Through this fundamental book for the energy industry of
tomorrow, Richard Baxter has broadened the industry’s horizon by showing
how it will be revolutionized by energy storage technologies—enabling greater
use of renewable energy, and promoting a more flexible, efficient, and stable self-
correcting energy infrastructure.
Dr. Hermann Scheer

General Chairman World Council for Renewable Energy
President EURSOLAR
Recipient of the Alternative Nobel Prize
Member of the German Bundestag
Contents
List of Figures ............................................................................................. xi
Acronyms and Abbreviations ..............................................xiii
Foreword ......................................................................................................... xv
1. Storage and the Electric Power Industry.......... 1

Misperceptions and Realities ............................................................... 3
Wholesale Power........................................................................................ 6
Facility utilization .................................................................................... 6
Cycling damage ...................................................................................... 8
Renewable energy ................................................................................10
Transmission and Distribution .........................................................12

Infrastructure underinvestment ..........................................................12
Transmission congestion .....................................................................15
Ancillary services ..................................................................................17
Retail ...............................................................................................................19
Power quality ..........................................................................................19
Cost of energy ........................................................................................21
References ....................................................................................................24
2. Storage in Other Energy Markets .............................25

Petroleum Market ...................................................................................27
The petroleum industry and the need for storage ........................27
Facility types and operations .............................................................29
Market uses and economics ..............................................................30
Coal Market ................................................................................................31

The coal industry and the need for storage ...................................32
E N E R G Y S T O RA G E : A N O N T E C H N I C A L G U I D E
vi
Natural Gas Market ................................................................................36

The natural gas industry and the need for storage ......................36
Lessons Learned ........................................................................................45

Value of storage ....................................................................................47
Technological change ..........................................................................49
Regulatory change ................................................................................51
References ....................................................................................................53
3. Electricity Storage Technologies ..............................55

System Components ..............................................................................56
Storage medium ....................................................................................56
Power conversion system (PCS)........................................................57
Balance of plant (BOP) ........................................................................59
Engineering, procurement, and construction .................................60
Pumped-Hydroelectric Storage (PHS) ..........................................60

Summary .................................................................................................60
Historical origins ...................................................................................61
Design and operations.........................................................................62
Cost issues .............................................................................................64
Installations ............................................................................................64
Prospects and challenges ...................................................................66
Major developers...................................................................................67
Compressed Air Energy Storage (CAES) .....................................68

Summary .................................................................................................68
Cost issues .............................................................................................74
Installations ............................................................................................74
Related Technologies............................................................................78
C O N TE N T S
vii
Flow Batteries ............................................................................................80

Summary .................................................................................................80
Cost issues .............................................................................................90
Installations—vanadium redox ..........................................................92
Installations—zinc bromine ...............................................................94
Installations—polysulfide bromide ...................................................95
Installations—cerium zinc ..................................................................96
Sodium Sulfur Battery ...........................................................................99

Summary .................................................................................................99
Historical origins ................................................................................ 100
Design and operations...................................................................... 101
Cost issues .......................................................................................... 103
Installations ......................................................................................... 104
Prospects and challenges ................................................................ 106
Major developers................................................................................ 108
Lead-Acid Battery ................................................................................. 108

Summary .............................................................................................. 108
Cost issues .......................................................................................... 113
Installations ......................................................................................... 114
Nickel Cadmium Battery .................................................................. 117

Summary .............................................................................................. 117
Cost issues .......................................................................................... 121
Installations ......................................................................................... 122
viii
Flywheels ................................................................................................... 125

Summary .............................................................................................. 125
Cost issues .......................................................................................... 129
Installations ......................................................................................... 130
Electrochemical Capacitors ............................................................. 135

Summary .............................................................................................. 135
Cost issues .......................................................................................... 139
Installations ......................................................................................... 140
Superconducting Magnetic Energy Storage ........................... 142

Summary .............................................................................................. 142
Cost issues .......................................................................................... 147
Installations ......................................................................................... 147
Thermal Energy Storage .................................................................... 151

Summary .............................................................................................. 151
Cost issues .......................................................................................... 155
Installations ......................................................................................... 156
References ................................................................................................. 160

C O N TE N T S
ix
4. Applications ................................................................................... 166

Key Design Issues .................................................................................. 169
Energy versus power ......................................................................... 170
Cycling issues ..................................................................................... 171
Ownership costs ................................................................................. 173
Wholesale Power................................................................................... 175

Commodity arbitrage ........................................................................ 177
Contingency reserves ........................................................................ 181
Blackstart capability .......................................................................... 184
Transmission and Distribution ...................................................... 186

Frequency regulation ........................................................................ 188
Voltage regulation .............................................................................. 193
Transmission and distribution asset deferral .............................. 196
Retail Market ........................................................................................... 200

Power quality ....................................................................................... 201
Energy management ......................................................................... 206
References ................................................................................................. 212
5. Renewable Energy and Storage ..................... 214

Resource Utilization Challenges ................................................... 217
Noncoincident peak........................................................................... 219
Nondispatchable ................................................................................ 219
System stability................................................................................... 220
Remote Power ........................................................................................ 222

Island Grid ................................................................................................ 225
Grid Connected ..................................................................................... 233
Dispatchable wind ............................................................................. 235
Capacity firming ................................................................................. 237
Baseload wind .................................................................................... 241
References ................................................................................................. 245

x
6. Our New Energy Future...................................................... 248

The Road to Here .................................................................................. 249
U.S. federal government ................................................................... 249
State energy programs ..................................................................... 251
International programs ..................................................................... 253
Industry programs .............................................................................. 254
User’s Goals for Storage Technologies ....................................... 257

Commercial and industrial .............................................................. 257
Utilities .................................................................................................. 258
Energy service companies ............................................................... 259
Regulators ............................................................................................ 260
A Road Map for the Future ............................................................. 261

Federal support ................................................................................... 262
State support ....................................................................................... 264
Industry support .................................................................................. 265
Hurdles to overcome ......................................................................... 269
References ................................................................................................. 271
Appendix—Vendor Web Sites ............................................ 272

Bibliography ............................................................................................ 274
Index ................................................................................................................ 288
1 STORAGE AND
THE ELECTRIC
POWER
INDUSTRY
The electric power industry has some immense challenges before

it—that, we can all agree, is glaringly obvious; the good news is that
energy storage technologies offer real solutions to some of the most
pressing of these issues. Many of the worst problems stem from issues
built into the system through the market structure. One of most vexing
is that the current power system is built around a central tenet: Electricity
must be produced when it is needed and used once it is produced. This
rule necessitates rigid procedures for operating the system—raising
inefficiency, lowering reliability, and reducing security. Although radical
solutions from pundits abound, most industry veterans understand the
sheer scale and interconnectivity of the system mean that change here
comes most readily through evolutionary and not revolutionary means.
Because energy storage technologies are usually enabling technologies
and not disruptive ones, their expanded use will enhance the value of
existing assets by providing more flexibility and options—supporting the
inherent infrastructure-centric nature of the market.
2
Although the industry may prefer slow, evolutionary change, it

should also be understood that change is endemic in the power industry.
In the short-term operation of the electric power market, what happens
on a daily basis can take weeks or months in the natural gas market. In
the long term, evolving regulatory, economic, and technological forces
have affected the market since its inception, and will continue to do
so in the future; while some of this change is intended, much is either
unexpected or even unwanted. First, it exposes weakness in the market;
episodes like the 2003 blackout in Canada and the northeastern United
States show that there is a real need for the existing infrastructure to be
coordinated in a far more effective manner. Second, change creates a
need for new market tools—existing assets are built to perform in certain
market conditions, and when these change, some of these assets are
simply not readily adaptable. Finally, change also produces opportunities;
as the market moves from a tightly controlled structure toward a more
open and flexible stance, many market participants can more easily take
advantage of innovative business models and technological advances
such as storage technologies.
This does not mean that energy storage technologies are a panacea or
a solution for every market problem; existing power market assets remain
the backbone of providing service to customers. However, the normal
operation of conventional technologies propagates the inefficiencies
inherent in the industry, which often produce the very situations where
the flexible capabilities of a storage asset are most needed. In fact, these
instances of system instability frequently produce the greatest cost to the
system—cost that can be reduced through greater adoption of energy
storage technologies. In the previous incarnation of the market, standard
regulatory-led practices lent themselves to capital-intensive solutions in
meeting these challenges as the redundant equipment could simply
be rolled into the rate base. Although functional, this was an inelegant
solution to a chronic problem; in an increasingly competitive market, this
inefficient solution is no longer acceptable.
S T O RA G E AND THE ELECRIC POWER INDUSTR Y
3
For these and other reasons, energy storage technologies provide

an option (often the least costly one) that the industry can call on to
solve some of its most difficult challenges. Energy storage technologies
break the linkage between electricity production and demand, allowing
the storage of power for later use. By maintaining a ready-reserve cache
of energy, storage technologies can help industry participants overcome
such challenges as:
• Improving low utilization of power facilities

• Relieving transmission congestion
• Improving the market potential of renewable energy
generation
• Preventing losses from unreliable power quality for end-use
consumers
By absorbing or providing power—even a small amount—at

precisely the right time and place, energy storage technologies can relieve
significant system stress and costs, effectively acting as a shock absorber
for the industry.
Misperceptions
and Realities
Storing energy is a seemingly simple and familiar concept, yet even
electric power industry insiders often misunderstand its real potential.
The technology is straightforward—storage technologies convert
electrical power into chemical, mechanical, or electrical potential energy
and retain the ability to reinject it into the power grid when called
on. Unfortunately, the stored electrical power is then viewed simply
as a static repository of energy, with any value ascribed strictly to its
4
commodity value. This is a simplistic, one-dimensional view of storage

technologies—the deliverability of that energy is another key aspect. The
ability to decouple the linkage between power production and power
demand allows the industry to use storage technologies as both a sink
and a source of energy—and both are valuable resources in a dynamic
environment such as the electric power market. As a sink, storage
facilities can not only absorb energy slowly from generators to optimize
their operations, but also absorb rapid surges, preventing imbalances
that can affect the power grid’s stability. As sources of energy, storage
technologies not only can be used to arbitrage energy between off- and
on-peak usage, but they can also be used to adjust the rate of energy
production and purchases. The ability to affect the supply/demand
balance of energy delivered to customers provides the capability to then
reduce ramping stress or otherwise optimize the use of generation and
transmission assets, or to prevent consumer equipment damage in the
retail market from power fluctuations.
Although treated as a new technology, the electric power industry

has actually maintained significant storage capability for quite some
time now. Therefore, the issue going forward is rather the rate and type
of further adoption—not introduction—of these technologies in the
market. From the wholesale to the retail market, firms have recognized
their flexibility and have invested significant amounts of capital into
these technologies for many years. In the wholesale power market, U.S.
utilities began in earnest to build a number of pumped-hydro facilities
in the 1970s and 1980s, resulting in 20 GW of capacity by the early 1990s
(nearly 3% of all U.S. summer capability at the time). Besides their use
for arbitraging off-peak power to peak demand, they have proved very
useful for providing much-needed stabilization for the transmission
system. In the retail market, commercial and industrial firms drive the $7
billion spent globally on uninterruptible power supply (UPS) equipment
to provide enhanced power quality and reliability as the demand for
better power quality grows.1 Even the U.S. power industry itself keeps
banks of batteries on hand at each of the estimated 100,000 (or more)
5
substations and every major power facility to run the control equipment
in the event of a power outage. Existing applications will continue to
grow with the market, and new applications will be enabled through
expanding capabilities of existing technologies and the continued
commercialization of newer ones. As one sign of a maturing storage
technology market, some storage technologies now target existing
markets already supported by existing battery technologies.
Therefore, energy storage technologies are not simply a solution in

search of a problem—they have real uses both now and in the future. In
some instances, they enable applications that generate real revenue, whereas
in other applications, they either reduce capital expenditures or prevent
losses from damage-process interruptions. As these are all valued roles now,
no paradigm shift in thinking, market structure, or technological basis is
needed for these technologies to expand their presence in the market. As
the market continues to change and new demands are put on the industry,
these existing and new technologies will enable a more flexible way of
operating and provide the real optionality managers need. For this reason,
the continued introduction of these technologies will not simply be vendor-
driven, but actually demanded by the market. In fact, three broad themes
at the core of most corporate and public policy goals will promote the
continued deployment and use of energy storage technologies:
1. Efficiency. Storage technologies can be used to improve the

operations and use of generation and transmission facilities,
enabling a more flexible and effective market. Even if their
capability is short-lived, so too are the times when they are
needed most, such as when the wholesale price of power spikes
during times of temporary transmission capacity shortage.
2. Reliability. Storage technologies can enhance the ability of
utilities to provide electricity service to customers without
any of the normal or abnormal fluctuations in the quality of
the power impacting their customers by improving the power
system’s stability.
6
3. Security. Storage technologies can provide a means to blunt

short-term disruptions on the power grid before they become
large-scale problems by giving system operators the fast-
response capability provided by a ready-reserve resource—
or they can harness additional resources to support restoration
of service once a blackout does occur.
Wholesale Power
The wholesale power market continues to prove far more challenging
than envisioned by the architects of its transformation, as competitive
power prices have brought both new opportunities and new setbacks.
Although recent entrants garner much of the market’s attention, owners
of the existing $500 billion in generating assets struggle to maintain their
profitability as the rules in the market change—changes that reveal
that previous operating strategies incurred far higher costs than first
imagined.2 These market changes also affect developers of renewable
energy projects who find that while many support the idea of developing
these domestic resources, integrating them into the market can be far
more difficult than first envisioned.
Facility utilization
The move to a competitive power market has stressed the ongoing
operations of many existing coal facilities; improving the utilization of
these units is one direct way to markedly improve their profitability.
Because coal-fired units provide half of all U.S. power generation,
improving their value ranks as a public policy concern. As it stands,
ramping down during off-peak hours (nights, weekends) to match
the current load leaves these units with appreciable unused capacity.
Although generally termed baseload, coal-fired units do not enjoy the
must-run status of nuclear units, leaving the fleet average of all coal
7
units around 70%, whereas nuclear units have been able to increase
their levels well above 90% (fig. 1–1). Increasing the utilization of these
facilities would lower average operating costs—not only increasing
overall revenue, but also the profitability of these facilities. The off-
peak production cost of power could also benefit (from higher plant
utilization), as some market studies point to increases in the heat rate
of individual units of anywhere from 5% to 20% when running at low
power.3 Anecdotal evidence even points to some older units whose total
operating costs reportedly increase upwards of 40% when operating at
half power. Beyond lower operating costs, per-kWh-hour emission levels
would also decline, as the power facility is able to operate more closely
to full power. Although this overall improvement may not be dramatic,
any improvement on this issue in an era of ever-tightening environmental
restrictions could prove beneficial.
Fig. 1–1. Coal units lag nuclear in utilization improvements (Data: U.S.
Department of Energy).
Large-scale energy storage facilities can help increase the utilization of

coal facilities—or other units with excess capacity—by acting as an energy
sink during energy-usage off-peak periods. Coal facilities with the most to
gain would be those with significant excess production capacity as they
8
should have the lowest marginal cost for additional power production.
Existing large-scale energy storage facilities (pumped-hydro, compressed
air energy storage [CAES]) already operate in this arbitrage role today, and
even offset the need for some additional peaking capacity during peak
demand. Producing additional power at night can even help daytime
air quality, especially during hot summer days prone to ozone warnings,
because the emission levels of a CAES facility (and especially those of a
pumped-hydro unit) are far lower than those of a peaking combustion
turbine unit.
Cycling damage
Matching the changing daily load can also have negative impacts
on mid-merit power facilities, and to a far greater degree than previously
thought. Recent market strategies have evolved so that older fossil-steam
(and even some combined-cycle) units in the mid-merit role are forced
to cycle with the load, resulting in many facilities cycling heavily every
day—with some even cycling on and off daily . What is becoming better
understood is that the cost previously ascribed (and used currently in
dispatching decisions) to these activities is significantly below their real
value; the true cost of this cycling is much higher—sometimes more
than 10 times the previously used value—making the current dispatch
decisions far from optimal. One of the leaders in the field of power plant
failure/performance analysis is APTECH Engineering Services, which has
spent many years evaluating these problems. Based on analysis of real
operating data from more than 200 power generating facilities, Steven
Lefton, vice president of the firm, notes that if the true cost of cycling was
used, an average utility power system could reduce its total production
costs by 5% (and increase profitability significantly). This heavy cycling
and rapid load following is taking its toll by damaging plant equipment
and requiring far larger maintenance costs. Although the older fossil units
have proved far more rugged than first thought, the resulting number
of warm or even cold starts is far more than first envisioned for these
units, especially as some are now older than many of their operators.
9
Even combined-cycle units—having improved significantly from their

first entrance in the market—still find the cycling troublesome. The
resulting effect on these units is that the wear and tear is accelerating,
resulting in higher operation and maintenance (O&M) costs, longer
and more frequent forced outages, higher heat rates, and shorter life
expectancies for critical plant components (fig. 1–2).
Fig. 1–2. Power generation cycling cost breakdown (Courtesy of Aptech

Engineering Services, Inc.)
Large-scale storage facilities can help these mid-merit power facilities

in two ways. First, they can provide a sink for energy at night sufficient to
keep these plants from shutting down, precluding many of the excessive
warm starts that create the costly and cumulative damage to the units.
Second, the storage facilities can provide faster-responding load following
services so that the demand on power generation facilities is not as
dramatic during times of rapidly changing demand, allowing these older
units (and combined-cycle facilities) to ramp along their optimal design
path rather than what the market demands. Both of these roles would
reduce the cyclical wear on the power facility equipment significantly,
greatly extending its life and lowering the maintenance replacement costs.
Extending their analysis to include this application of energy storage,
10
APTECH Engineering Services estimates that including storage into the mix
can improve the performance of a large utility system by anywhere from
10% to 28% from reduced cycling damage and dynamic heat rate effects.4
Renewable energy
Increasing the use of renewable resources such as wind is one of the
prime goals of U.S. energy policy makers. Not only does wind provide a
domestic source of power, it is an environmentally clean source of energy.
Wind power in particular has seen dramatic growth, with a total of more
than 6,000 MW of wind turbines installed by the end of 2003. Far more is
possible, with the American Wind Energy Association (AWEA) estimating
that the United States alone has three times the wind power potential
compared to its current electricity usage today. However, these new
resources face market penetration challenges as they become integrated
into the power market. Wind energy production potential is generally
noncoincident with peak demand periods; most (roughly two-thirds) of
the energy that can be produced from wind turbines is outside of the
weekly peak demand (and pricing) periods. In addition, wind resources
are variable—making the stable delivery of power from the wind turbines
hard to achieve. Most wind resources developed to date have been in
remote locations where the existing utility transmission capabilities are
generally limited.
Advancements in wind turbine technology have created many

new opportunities for even more development, but challenges
remain. Although modern wind turbines can now produce power
competitively with natural gas combined cycle units, there is a growing
realization that the wind energy potential in many areas will remain
stymied because of the above-mentioned challenges. Noncoincident
production means that much of the power is only sold at off-peak
11
prices. Unpredictable power means that not only is the capacity

payment heavily discounted, additional ancillary services are needed
to integrate the wind power into the power grid—again lowering
the value of the wind energy to the system operators. Much effort
(especially in Europe) has gone into extending weather forecasts into
wind-power forecasts, but weather forecasts have their limits, especially
in the time frame required by the power industry. Distant resources
mean additional costs for transmission or even the potential loss of
sale because the wind energy could be constrained off the power grid
during peak demand. Therefore, without addressing these attributes
of the wind resource, modern wind turbines may very well be able to
turn vast quantities of the nation’s wind resources into electrical power,
but that wind energy will continue to be penalized or even unusable
in the market.
Storage can improve the value of wind energy (and reduce the project
risk) by both increasing the value of the wind energy and reducing the
current discounting of its output value. At its heart, storage offers a means
to decouple the production of wind energy from demand, and to provide
that power in a dispatchable and stable manner to the market. These
capabilities can directly mitigate the negative aspects of wind resources
to improve the value of the wind energy and promote a greater market
penetration of wind power. For instance, on small or isolated power grids,
wind resources are plentiful, yet the variability of that potential resource
could mean additional generating units must be added to stabilize the
power grid, actually increasing the overall cost of power. A growing number
of wind and storage projects with short paybacks are proving the success
of this strategy today. With storage, these small power grids can rely on
wind power to a far greater degree, reducing or even eliminating the need
for the diesel generator and its expensive fuel needs. In the larger wholesale
power market, a number of wind and storage projects are currently being
evaluated. Although some will provide a capacity-firming capability to
support the wind turbine output, others plan to optimize the design
12
and use of supporting components such as the transmission facilities.

Storage can even help ensure that distant wind resources are able to get
their output to market if the transmission capability is severely limited,
threatening to strand the renewable energy far from where it is wanted.
Transmission and
Distribution
The transmission and distribution sector is arguably the area most in
need of attention in the power industry. Developed with an emphasis on
reliability rather than efficiency, the system was built out to ensure suf-
ficient capacity during peak demand periods—leaving much of the system
largely unused during off-peak times and producing an average system uti-
lization that rarely surpasses 60%. The $100 billion and $250 billion worth
of assets in the transmission and distribution markets, respectively, make
it apparent why it is necessary to improve their capability and cost effec-
tiveness.5 Although the industry is attempting to transform the transmis-
sion system into a real national grid with self-correcting market controls,
the distribution networks struggle to satisfy consumer demand for ever-
improving levels of power quality and reliability. Confounding the plans for
these changes are falling infrastructure investments, helping to cause grow-
ing congestion problems during peak periods as strategies to improve and
standardize the rules for operating the system remain slow in coming.
Infrastructure underinvestment
Underinvestment in the underlying infrastructure of the nation’s
power grid is a well-known and growing problem. Reasons for this lack
of sufficient investment are a bit circumspect as they include historical
construction patterns, regulatory risk, and poor financial performance.
13
It should be noted first that the system was built around a central
tenet that electricity can only be produced when it is needed, and it
must be used once it is produced. Operating the power grid as simply
one big just-in-time delivery system in this way, however, is extremely
wasteful. It requires significant capital to be tied up in a production and
delivery system that must always be capable of meeting the highest
expected demand, but the highest expected demand is only required
infrequently—for much of the year, the system remains significantly
underused. In addition, economic necessity and technical standards
have curbed development and installation of large upgrades or additions
to the system, producing the familiar stair-step expansion pattern for
transmission capacity. For example, as demand grew rapidly throughout
the 1950s and 1970s, significant transmission investment took place
to provide the wide safety margins common in those days. After the
energy shocks and recession of the early 1980s, however, lower demand
growth permitted existing assets to meet demand longer than previously
expected, allowing transmission expansion budgets to be pared back. As
increasing regulatory changes fueled the uncertainty as to what would be
the final structure of the industry, investment continued to taper off as
utilities looked for markets with higher returns and that were perceived
as less risky. Unfortunately for the transmission market, these regulatory
changes favored the generation sector throughout the 1990s, making that
market a far more attractive investment opportunity and thus reinforcing
the decline in transmission investment, which continues today.
As investments in the infrastructure continue to fall behind the

need for them, raising the efficiency of the existing assets becomes
even more important. Although the U.S. Department of Energy (DOE)
estimates that the high-voltage transmission network is comprised of
assets worth $100 billion, the Edison Electric Institute (EEI) estimates that
the industry spends only $3 billion annually on maintenance, and $2 to
14
$2.5 billion for capacity expansion (230 kV and above). In fact, over the
past 25 years, the group has noted that investment in the transmission
system has declined by $115 million per year, leading to the present state,
where annual investment (in nominal dollars) has reached half of what it
was in 1975, whereas actual kWh sales have doubled. Without additional
investment, badly needed upgrades will continue to go undone, further
stressing existing equipment already operating past its expected life span
and making it much more susceptible to a breakdown. Although the EEI
estimates that an additional $56 billion will need to be spent to catch up
with the legacy of underinvestment, the problem continues to grow as
the DOE expects demand to grow 22% from 2000 through 2010, whereas
the North American Electric Reliability Council (NERC) is only estimating
the expansion of the transmission power grid by 5%.6
Although many hold out for a regulatory-led movement to

increase the level of investment in the industry’s transmission
infrastructure, storage can help increase the usefulness of existing
assets—a far more effective means of investment for utilities in today’s
environment. By making existing equipment capable of meeting future
demand and postponing otherwise needed upgrades (especially
in lower-voltage systems where the investment problem is even
more acute), the increased use of these assets will help improve the
financial performance for their owners. In theory, the case for a storage
component in the power industry would therefore seem obvious. With
a storage component, the owners and market coordinators must build
out only what is necessary to carry a heavy, but more typical load—
resulting in a much higher utilization of their existing equipment, and
the system overall. Even a few installations of storage technologies could
provide badly needed supplemental capacity (bulk energy and system
stability) during peak demand, lowering capital costs and mitigating
supply disruptions. At the transmission level, storage technologies are
already a cheaper alternative for Wisconsin Public Utilities than existing
15
voltage stability equipment or capacity expansion alternatives. At the

distribution level, other utilities such as PacifiCorp are beginning to
use other storage technologies to postpone the need for upgrading
remote power lines, which is a constant challenge that is both costly
and lengthy.
Transmission congestion
One result of the lack of investment has been the increasing
level of congestion because of a deficiency of excess transmission
capacity during peak demand periods (fig. 1–3). Congestion is a
serious problem for the transmission market, affecting large areas
of the power grid when vital links such as the infamous Path-15 in
California become oversubscribed, contributing to system instability
and more volatile power prices. As this condition worsened through
the later the 1990s, independent system operator (ISO)–levied
congestion charges amounting to hundreds of millions of dollars
per year were passed on to customers. The total bill for congestion
charges is now measured in the billions, with more charges expected
in the foreseeable future. Besides direct monetary costs, congestion
also affects planning for power project development, such as where
to locate; while some target load pockets where prices may be high,
others such as wind developers are penalized because their location
is often predetermined, and the question becomes whether they
can get the power to market. In fact, some wind farm developers
do not build out their site to its fullest extent because of the lack of
transmission capacity during peak times. Building out vast excesses
of transmission capacity is difficult (beyond cost, simply permitting a
new high-voltage line today requires Zen-like patience), so other, more
realistic solutions must be found to raise the carrying capacity of the
existing network. Ongoing efforts to raise the carrying capacity of the
transmission network through more efficient operation, expanding
existing right-of-ways, and using flexible AC transmission systems
16
(FACTS) devices are being promoted, particularly in susceptible

areas of the power grid. By improving the carrying capacity at
these bottlenecks, the entire carrying capacity of the power grid
can be improved.
Fig. 1–3. Transmission load relief incidents on the rise (Data: NERC).
Congestion issues also extend into the distribution market. The

previously mentioned underinvestment grows as one moves down
in transmission size, leaving the system stressed during peak demand.
Because of its geometry, congestion at this level can produce serious
power quality and reliability issues for customers in highly localized areas
by preventing power from reaching end users; long-distance, low-voltage
distribution lines are especially prone to voltage problems for this reason.
Demand on these lines is extremely variable. Therefore, even if the average
utilization of the line is low, there can be times when demand exceeds the
transmission capacity of the power line. In fact, only a few instances of
maximum usage (or near) per month are enough to warrant an upgrade
17
in transmission capacity to alleviate the problem. When this occurs, it is

common for the line to be upgraded significantly to postpone the next
required upgrade, ensuring that the average utilization of the power line
will remain low.
Storage technologies can help alleviate both the lack of transmission

capacity and the stability problems that ensue during periods of high
demand. Sited past a bottleneck, they can provide a prepositioned source
of energy, which will allow the system to ride through a few short-term
peak demands during the month—postponing the need for an expensive
upgrade on the line until a more sustained level of demand warrants it.
With accompanying power electronics, they can also provide a means to
maintain voltage stability at these vulnerable locations—although the
act of providing real power in these stressed areas can also improve the
system’s voltage stability. Although there are alternative means to supply
short-term power (small peaking facility) or voltage stability (capacitor
banks or FACTS equipment), neither of these choices is as flexible as an
energy storage facility. Many storage technologies developed for this role
are also mobile, so they can be moved as the conditions and needs of
the power grid change.
Ancillary services
Ancillary services are essential to maintain the stability and reliability
of the power grid for the transmission of electric power; unfortunately,
they remain mostly misunderstood—even by many within the industry.
Frequency regulation, voltage control, and contingency reserves are
just a few of the components that must be monitored and constantly
regulated because changing load conditions will interact with the
geometry of the system to affect the reliability and deliverability of power
to customers. As power generators provide many of these services, the
move toward divestitures (and a focus on peak power sales) has left
providers with less incentive to provide these vital support functions
just as the need is growing. Once simply a concern of utility system
18
operators, the Federal Energy Regulatory Commission’s (FERC) desire

to create a more self-correcting electric power market led it to create
a competitive market for some of these services coordinated at the ISO
level. This unbundling of services is modeled after its previous successful
actions in the natural gas market, where substantial growth in flexible
contract sales occurred once the transmission charges were unbundled.
Although more work is needed in the power industry, this unbundling
has finally provided the beginnings of much-needed price visibility of
these ancillary services. Because the provision of these ancillary services
is such a central component to the working of the wholesale market, it
would be impossible to think that the wholesale market could operate
competitively and successfully without real reform of these functions.
Most well-regarded analyses of the U.S. ancillary services markets

originate from work done by Eric Hirst and Brendan Kirby. In a 1998
study, they estimated that the United States spent $12 billion on ancillary
services in 1996—fully 10% of direct energy costs. Other researchers have
found similar (or higher) costs. Most assumptions point to a market that
has grown significantly since then, as many of these values represented
embedded costs from an era prior to deregulation and not opportunity
costs in a competitive market. Specifically, many of these ancillary
services are linked to the spot price of power, or simply the scale of the
overall power market—both of which have grown significantly since the
mid-1990s. However, simply using Hirst and Kirby’s original $4.15 per
MWh metric for all ancillary services, today’s total ancillary service costs
amount to more than $15 billion annually; with market-derived prices for
these services, the total market value could easily be far larger.7
A handful of storage technologies is already providing ancillary

services to the power market, with others set to begin competing in
the near future. One core strength of these storage facilities is their
responsiveness. Because the role of these ancillary services is to
counteract rapid variations in power flows, responsiveness has real value
for these applications. For example, CAES facilities have the ability to
19
come to full power within 10 minutes (and pumped-hydro facilities can

respond within even less time), whereas natural gas turbines can require
25 minutes, and combined cycles 45 minutes in some instances. Coupled
with a large storage reservoir, energy storage facilities can provide load-
following or other longer-term energy supply ancillary services. Smaller
storage technologies can also support the stability needs of the power
grid. Placed at strategic locations, these units can directly reduce the
stress at weak points on the power grid by providing both rapid energy
balancing and voltage control. Some of these units can be moved as
the stability conditions of the power grid change. In the near future,
more storage technologies are targeting a role in this market as well. For
instance, a MW-scale, high-cycle capability storage system could provide
frequency regulation services by absorbing and discharging essentially
the same energy far more quickly than a power facility could respond.
Although not to the same scale as a power facility, the much-smaller
storage facility could, nonetheless, offer real value by providing support
to troubled areas of the power grid.
Retail
Energy use is of growing importance to many commercial and
industrial customers—but not by choice. According to the DOE,
commercial and industrial firms spent $133 billion for electricity service
in 2000; as the U.S. economy continues to grow and to increase the
penetration of electrical equipment, their electrical demand is expected
to be 20% higher by 2010. Although most firms will not be fully exposed
to volatile time-of-use rates over the next few years, some are already
having to create strategies not only to save money on their electric bills,
but, more importantly, to prevent problems with electric power quality
negatively impacting ongoing operations. For the future, most customers
understand that the end result of restructuring will be that they will
be far more responsible for looking after their own best interests than
in the past.
20
Power quality
Poor power quality is a growing concern to commercial and industrial
firms. This is because poor power quality is quickly becoming a hidden
drag that threatens to erode much of these firms’ hard-won operational
improvements. It is not a decline in the quality of delivered power,
however, that is the culprit here; it is, rather, the increasing expectations
customers have for their electricity service. Whereas industrial customers
fear even transient power fluctuations that can disrupt their high-speed
processes, commercial firms worry about momentary power sags as fault-
intolerant information technology spreads throughout the firm. Beyond
simply causing a loss of in-process work, larger power-quality events can
even damage this increasingly expensive equipment. Surprisingly, most of
these power quality disturbances are short-term, according to the Electric
Power Research Institute (EPRI), with 98% lasting less than 30 seconds and
90% lasting less than 2 seconds.8 What these firms want—either from
their service provider, or, increasingly, from their own energy management
strategy—is a means to essentially provide a capability for loss prevention
from these transient frequency variations or voltage surges and sags. The
sad fact is that the real impact of these disruptions will continue to grow,
however, as most firms have “wrung out” any savings in their operations,
and any interruptions in the resulting schedule are increasingly expensive.
Estimates vary, but the annual cost to the U.S. economy from poor
power quality currently ranges from $119 to $188 billion (EPRI) to $150
billion (DOE) from interruptions in operations, lost work time, and damage
to increasingly expensive equipment.9,10 Because of the magnitude of
these losses, which stymie economic growth in hard-hit areas—and their
ever-upward direction—it is no wonder that the DOE has raised this issue
as a central part of its national energy policy. To put these values in the
perspective of an individual firm, one survey from a UPS manufacturer
estimates that an average manufacturing facility loses $1 million per
hour during a power interruption. For some firms, such as electronics
manufacturers, these costs are even higher, with estimates from the DOE
21
putting even a five-minute unexpected shutdown at some chip-fabrication

facilities at more than $5 million from product loss and restarting the
process.11 Information technology poses even a larger problem, where
literally billions of dollars of activity can be affected within that first hour
after a shutdown at a bank or other financial-transaction firm.
Because of this impact on the bottom line, many commercial and

industrial firms already use energy storage technologies to protect
critical loads from poor power quality. Made up of power electronics
and an energy storage component, these UPS systems are designed to
protect electrical equipment from momentary but potentially damaging
power sags and outages. As the impact of poor power quality grows,
the global market for UPS systems was estimated to be $7 billion in
2002, with growth over the foreseeable future to average 7% per year.12
With sufficient power, some of these UPS systems are also able to act
as a ride-through or bridging-power source for essential equipment in
the event of a power failure. Generally, 15 to 20 seconds are needed
to either bring online a backup power generator or switch to a
different local circuit. Maintaining a backup generator on-site remains
a popular strategy to ensure power for critical loads, as evidenced by
the multibillion dollar market for these systems. However, many firms
requiring utility-comparable power quality (or better) from their on-site
power units have found their capability wanting—especially if the load
remains variable—because the smaller on-site generators will have a far
more difficult time in responding to any changing load than the utility’s
system power would have. Here too, newer storage technologies such as
flywheels show promise in providing frequency regulation to enhance
the quality of power when the facility is isolated from the power grid.
Cost of energy
All commercial and industrial firms would like to spend less money
on their energy costs; in fact, the desire of these firms to lower their
energy expenditures has been one of the core driving forces behind retail
22
restructuring. The importance of reducing energy costs has created new

strategies—such as the growing use of adjustable speed drives to replace
older, more energy-intensive fixed-speed motors. Now, however, these
firms’ original desire for simply lower utility bills has been replaced by
the specter of utility bills rising unpredictably based on activity outside
of their control in an increasingly complex retail power market. For many
firms, even a small rise in energy costs could wipe out their hard-won cost
reductions in operational efficiency, so many of the variable expenses have
been put (again) under review to look for savings. Many find that both the
commodity charge and the demand charge contribute to the variability
in the total amount a firm spends on electricity service. Searching for a
better means to manage their energy needs, many are now concentrating
on power-intensive activities that occur during peak periods, looking to
reduce the demand charge to lower their variable energy costs.
For some firms, the cost of energy is a central component of costs,

warranting the development of an involved energy management policy;
for others, the effort can quickly cost more than the potential savings. For
example, in energy-intensive industries such as the aluminum industry,
fully one-third of the cost of production is electricity. However, the cost
of electricity generally represents only 2% to 3% of the total cost of most
commercial and industrial firms, and only 1% for most office buildings.
Surprisingly, even for many of these firms, where the overall cost would
not seem to warrant any action, what is commonly found is that certain
components of their energy usage may be suitable for incorporating
energy storage technology into part of their overall load. For instance,
the DOE notes that 25% of commercial usage for electricity is for cooling
load, a highly cyclical operation. Although heavy energy use increases
the commodity energy charge, highly cyclical loads such as cooling or
batch processing can increase the demand charge component of the bill
dramatically. Currently, roughly 30% of all firms are subject to demand
charges; if much of the reason for this stems from a highly variable load
such as cooling, then many times it may make sense to devise a means
to eliminate that daily surge in demand. By reducing the volatility in their
23
energy usage, the firm’s demand charges will decline, and total costs will
be far more predictable—reducing significant amounts of uncertainty in
their potential earnings.
Energy storage technologies can be—and already are—used to lower

companies’ power bills. Obviously, firms with high-priced tariffs stand to
benefit from storing low-cost power at night and using it during peak
demand the next day. However, many others (even those where energy
costs make up a small portion of costs) can also reduce their cost of
electricity if one part of their load is highly variable—for instance the
previously mentioned cooling load. Here, thermal energy storage is already
used by many commercial firms to essentially create large blocks of ice at
night, which are then used to assist with the daily air-cooling load for air
conditioning. According to some vendors, these units can many times
reduce cooling peak power demand by upwards of 50%, producing a
30% overall reduction in the cost for cooling. If integrated into the design
of a new building, these systems can even result in a reduction of the
cooling infrastructure equipment size requirement, sometimes by 40% to
60%—providing additional benefits for the firm. Because of these results,
typical payback periods for these installations (new and retrofit) can be
realized in one to three years.13 Newer storage technologies are enabling
a wide variety of other useful applications as well. For instance, flywheels
are capable of capturing wasted energy in repetitive motion situations.
This role of acting as a dynamic sink and source for power fits well with
transportation sector applications, where repetitive starts and stops
produce very inefficient use of energy. Although container port lifting
cranes and light-rail/subway systems all exhibit these usage patterns, they
are currently under served because previous battery technologies could
not support these applications, leaving these firms with no solution
to their problems. If an energy storage solution is used to lower the
operator’s costs, even the local utility stands to save because the load
swings on the power lines feeding these installations are reduced.
24
References
1. World UPS markets: Alternative energy storage solutions. 2003.
San Jose, CA: Frost & Sullivan.
2. U.S. Department of Energy. 2003. Grid 2030—A national vision for
electricity’s second 100 years, 3. Washington, DC: U.S. Department
of Energy.
3. Grimsrud, P., S. Lefton, and P. Besuner. 2004. Energy storage systems
(ESS) provide significant added value by reducing the cycling costs of
conventional generation. Sunnyvale, CA: APTECH Engineering.
4. Ibid.
5. U.S. Department of Energy. 2003. Grid 2030—A national vision for
electricity’s second 100 years, 3. Washington, DC: U.S. Department
of Energy.
6. Edison Electric Institute. 2002. Energy infrastructure: Electricity
transmission lines. Washington, DC: Edison Electric Institute.
7. Hirst, E. and B. Kirby. 1998. Unbundling generation and transmission
services for competitive electricity markets (ORNL/CON-454). Oak
Ridge, TN: Oak Ridge National Laboratory, U.S. Department of Energy.
8. EPRI distribution system power quality monitoring project (DPQ study).
2003. Palo Alto, CA: Electric Power Research Institute; Knoxville, TN:
Electrotek Concepts.
9. Electric Power Research Institute. 2003. Electricity technology road
map: 2003 summary and synthesis, 14. Palo Alto, CA: Electric Power
Research Institute.
10. Swaminathan, S., and R. Sen. 1998. Review of power quality
applications of energy storage systems (SAND98–1513).
Albuquerque, NM: Sandia National Laboratories.
11. Gyuk, I. Electrical energy storage (presentation). Electricity Storage
Association, 2000 Summer Meeting. Seattle, WA.
13. MacCracken, M. 1993. Thermal energy storage myths.
ASHRAE Journal. 45 (9): 36–42.
2 STORAGE IN
OTHER ENERGY
MARKETS
Storage is a fundamental component of the petroleum, coal, and

natural gas markets. As these industries transitioned from highly regulated,
commodity delivery networks (creating logistical issues) to highly
competitive, deregulated markets (fostering hedging/speculative strategies),
storage has been there—first enabling the transition, and then becoming a
key provider of services. Those involved with public policy development in
the electric power industry should review these other markets’ experiences
to learn how storage can become another tool in promoting a better
working power market. In these other energy markets, the introduction
and use of storage has undergone a surprisingly similar learning curve. First
recognized for leveling variations in physical supply, a more integrated use
of storage allowed lower-cost transmission upgrades and system expansion
costs. Deregulation in these markets—with particular respect to natural
gas—leveraged the ability of storage facilities to provide ready-reserves
of the commodity to create greater customer choice and, surprisingly,
a greater demand for the use of storage facilities. As wholesale physical
trading begat even greater financial trading, storage assets served not
26
only as the basis for many of these contracts, but also as a tacit physical
backstop for the trading activity. In fact, it was the availability of storage
facilities in the natural gas industry that allowed the overall market to
weather the restructuring of that market from deregulation.
Because they represent a physical hedge, storage facilities are most

useful in wholesale commodity markets. By improving the service past
a bottleneck in a network through the use of a prepositioned cache, or
arbitraging between markets with uneven service (improving deliverability
of the resource to take advantage of the uneven local prices), physical
and price volatility can be more easily mitigated. Storage facilities also
have a role as the market matures. As a market clearing price and forward
price curve develop through greater trade, storage facilities act as the
basis or physical backstop for trading—first physical and then purely
financial trading. An interesting point is that although storage facilities
need wholesale markets to reach their full values, wholesale markets in
turn need storage facilities to operate properly and facilitate lower end-
use costs and additional customer choices.
As the electric power industry finds itself in the midst of its own
fundamental change, significant lessons can be learned from these other
energy markets in how storage can be usefully applied. However, because
electricity is not simply a commodity but a service where moment-to-
moment quality distinctions are important, there are real applications
for storage in all three power market segments: wholesale, transmission
and distribution, and retail. Lessons learned from the petroleum, coal, and
natural gas markets can provide significant guidance in how storage can act
as a substitute for transmission upgrades or to improve customer choice.
The electric power industry already uses storage to improve customer
service throughout the value/supply chain, although this practice receives
little recognition. Unlike natural gas storage, however, energy storage in the
power industry has not seen a similar level of public policy support because
of different market structures and competing technologies. However,
maturing storage technologies and changing market conditions in the
S T O RA G E I N O T H E R E N E R G Y M A R K E T S
27
power industry are now providing an opportunity for expansion of energy

storage for use in a variety of applications throughout the industry. Even if
the raw quantity of commodity storage in the power industry may never
equal that in other industries, the speed of the power market provides for
the opportunity to have as great or even a greater overall impact in service
quality. For that reason, the question may not be why there should be a
storage component to the electric power industry, but, rather, why not?
Petroleum Market
The petroleum industry is one of the largest industries on the planet.
Not only does it literally span the globe, it is also one of the most mature
and complex delivery markets of any kind. As the industry has grown
(from the U.S. perspective) from a national one to one where more
than half of all petroleum comes from overseas, the use of storage has
increased from improving the flow of petroleum in interstate pipelines
to becoming the basis for more than $600 billion in trading that is now
only loosely based on physical delivery.
The petroleum industry and the need

for storage
Wholesale petroleum market. Americans in the United States
spent more than $87 billion on petroleum products in 2002. To deliver
the 15 million plus barrels per day (bpd) that were consumed, the 6
million bpd of domestic production were supported by 9 million bpd of
imported oil.1 This growing reliance on imported oil has been an area of
concern for many who fear a supply disruption. Although efforts to find
additional domestic resources continue, significant effort is underway to
streamline the delivery networks to speed the transit time of product,
while providing redundancy to the system so localized disruptions do
not affect larger areas.
28
The need for storage. Petroleum stocks are used primarily to

keep crude oil and refined products moving efficiently from wellhead to
end users. As an immediate or ready-reserve source of supply, petroleum
stocks provide a physical backstop against both expected and unexpected
market disruptions. For instance, in production regions stocks allow for
production to be disconnected from pipeline supply; otherwise, well
output must be controlled to match demand, significantly affecting the
life (and total output) of the well. Along the pipeline, storage provides an
essential balancing role in case of a shortfall or in case regional demands
change. However, these disruptions while the petroleum is moving are
essentially logistical issues because of the speed of the crude oil. On
average, crude oil and products move through a pipeline at 3 to 8 miles
per hour, and crude oil tankers travel at 16 knots, requiring 12 to 18 days
for product to travel from Texas to New York, or 30 days for crude to
reach the United States from the Middle East.
Historically, the first storage of oil (stocks) simply consisted of large

wooden boxes or even simply large pits in the production regions.
However, because crude oil and petroleum products are volatile and
evaporate easily, storage techniques to contain these vapors began
to improve after the U.S. Civil War (1865), when riveted wrought iron
tanks began replacing the wooden variety. This style of storage facility
remained the primary way to maintain oil stocks for the next 50 years.
Advancements in welding technology in the 1920s led to another round
of improvements as steel storage tanks were introduced as more durable
storage facilities. Throughout all of this period, the sheer volume of oil
stocks continued to increase in step with production, with relatively high
levels required because of the poor transmission coordination capability
of these far-flung enterprises. As oil demand grew quickly following World
War II, stocks of petroleum products increased to maintain deliverability
in the ever-growing network of pipelines—and increasingly import
terminals. With the greater reliance on imported oil came more supply
instability stemming from political instability in foreign-production
regions, increasing the need for additional supplies to mitigate temporary
29
supply disruptions. After the early 1980s, however, stocks of petroleum

products began to decline as efficiency improvements reduced the need
for excessive amounts, and increasingly volatile prices increased the
carrying costs for commodities.
Regulatory impact on storage. Besides health and safety

regulations, the only regulatory development concerning storage has
been the establishment of agreements by International Energy Agency
(IEA) member countries to hold stocks of crude oil to mitigate supply
disruptions, which are often politically motivated. With the oil-sharing
rules of IEA, member countries share the burden of an oil supply shortage,
creating some resilience to the demand component of the global supply
network. These agreements require that each participating nation hold
stocks equal to 90 days of imports. Most member countries meet the
requirement with industry-owned stocks that can be commandeered in
an emergency; however, the United States, Japan, and a few other nations
also hold government-owned stockpiles. These 90 days of separately held
stocks were estimated to be worth around $100 billion in 2004. Crude oil
stocks owned by the U.S. government are held in the Strategic Petroleum
Reserve (SPR), originated in 1975 as a temporary replacement for imports
during significant supply disruptions.
Facility types and operations

Storage tanks fall within three levels that correspond to their position
or level within the supply chain:
1. Primary. Facilities are used for crude oil storage, refinery

operations, and petroleum products storage at wholesale
marketing facilities with more than 50,000 barrels of capacity.
Above-ground tanks are the most common type of primary
storage facility in the United States, with the average facility
holding in excess of 500,000 barrels.
30
2. Secondary. Facilities include wholesale marketing installations

with a capacity of less than 50,000 barrels and retail establish-
ments. These facilities balance the distribution demands of
tanker trucks delivering products to retail markets as the
demand fluctuates because of seasonal and economic factors.
3. Tertiary. Storage comprises end-use consumers’ home-heating
fuel tanks and vehicle fuel tanks.
Market uses and economics

An important role of oil stocks is as an indicator of spot price
movement. Normal fluctuations in crude oil stocks occur as the market
participants attempt to balance seasonal changes in the demand
for refined products and the projected supply of crude oil. These
discretionary stocks also effectively serve as a physical hedge against
price disruptions in the financial trading market. With 10% to 15% of the
world’s primary petroleum stocks discretionary, movements in this small
subset of oil stocks are followed closely.
In the short term, stocks act as a real-time indicator of the relative

availability of deliverability. An inventory increase or decrease that
appears small may actually result in a large price change because this
movement can actually represent an important shift in the usable
volume of inventories. Because of the depth and competitive nature
of the petroleum market, the forward price curve causes a feedback
impact on stock levels, and together they act as a self-correcting agent
for the market in general. For instance, when stocks are low, spot market
purchases increase, driving up prices, which begets more production
and finally an increase in stock levels. When stock levels reach points
higher than historical norms, spot purchases decline, begetting lower
production and encouraging limited additional stock fills. Besides the
spot trading market, stocks of petroleum products influence refiners’
day-to-day operational decisions in the short term. By balancing the
31
cost of storage with other issues, such as refinery margins and expected
prices for certain products, refiners attempt to maintain sufficient
stocks for operational needs while boosting profitability. Outside of
the refineries, storage of particular products is also used to bolster the
supply of a product during its demand season to assist with optimizing
refinery usage.
In the long term, stock levels are governed primarily by the carrying
cost of storage, which is made up of two parts: the actual cost to store
the petroleum and the opportunity cost of ownership. The physical
storage cost can vary significantly over time, depending on the type of
oil (crude or product) being stored, the current availability of storage, if
the storage capacity is owned or has to be rented, the price of the oil,
and so forth. The opportunity cost is then the implicit cost of working
capital tied up in owning the oil during storage. For instance, according
to estimates from the Department of Energy (DOE) based on average
prices in the first half of the 1990s, holding crude oil for a year would
cost a company approximately $1.50 per barrel if it had its own storage
and $4 per barrel if it had to rent storage tank space.2 In this way, the cost
of storage can be important to the overall strategy of the firm, even if it
is not widely evident.
Coal Market
Power generation derived from coal provides more than 50% of
all electric power in the United States, and coal is expected to remain
the single most important fuel to the electric power industry for
the foreseeable future. Moving this fuel from mines to major power
facilities, both of which number in the hundreds, across sometimes
thousands of miles, drove the need for a flexible storage solution
capable of maintaining continued operation through sometimes-
severe supply disruptions.
32
The coal industry and

the need for storage
Wholesale coal market. Slightly more than 1 billion tons of coal are
mined each year in the United States, with around 90% of that destined for
the power sector (the remainder is for steel making). Coal comes in a wide
variety of heat content types and contains varying amounts of impurities,
depending on the seam of origin. These impurities—sulfur, moisture, ash,
and others—greatly affect where the coal can be used, requiring many
power facilities throughout the United States to source coal from thousands
of miles away to obtain the coal appropriate for the power plant.
During the 1970s and 1980s, many utilities—responding to expected

shortfalls of natural gas supplies, problems with nuclear power facility
building programs, and high inflation—entered into long-term coal supply
contracts of sometimes 20 years, or more. The fears of insufficient natural
gas supplies for the power industry, in addition to industrial and consumer
use, restricted the construction of power plants that use petroleum or
natural gas as their primary fuels, and encouraged the building of many
new coal-fired facilities. The move to longer-term contracts was considered
a wise one at the time because there were no significant new coal
resources—or improved mining techniques—expected to increase the
supply of cheap coal. After most of these contracts were signed, however,
new underground mining techniques were developed and enhanced, along
with vast surface mining operations in the western United States (especially
in Wyoming’s Powder River Basin), which drove down the cost of coal
production dramatically. As the long-term contracts began to expire in the
1990s, many power generators were able to move to shorter and far more
flexible contract terms, and a greater reliance on spot purchases. Many
coal customers have, therefore, been free to experiment with using a wider
variety of coals, such as the low-sulfur coal from Wyoming’s Powder River
basin, which now produces 33% of all U.S. coal, up by more than one-half
of its output from only 10 years ago. With a significantly more flexible and
33
shorter-term delivery focus, many power facilities now require a far more
detailed logistical strategy for handling and storing coal.
The need for storage. In the 1950s and 1960s, many of the new
power facilities were powered by petroleum and natural gas because the
high price of delivered coal—governed by high mining and transportation
costs—drove away new plant orders. To provide more competitive prices,
the coal-mining industry worked with the railroads to create longer
unit-trains of 100 cars or more that could provide lower transportation
rates—resulting in lower delivered costs. Each of the 100 cars carried
100 tons of coal (125 tons in newer cars), so a unit train could move
10,000 tons of coal or more at once. However, to take advantage of these
special rates, the train needed to be loaded and unloaded in four hours
or less (far more quickly than the sometimes day-long events common
before) to meet more demanding scheduling requirements. These larger
deliveries, therefore, resulted in the need for larger, more flexible storage
facilities at the mine mouth and at the power stations. In tandem with
this transportation evolution, new coal-fired power plants continued to
increase in size as developers attempted to reduce the cost of producing
power—also requiring significantly larger storage facilities to hold and
handle the fuel for these behemoths. Some power facilities contained
individual boilers rated upwards of 600 MW each—up by nearly an order
of magnitude from the previous generation of power facilities.
Regulatory impact on storage. Whereas the coal mining industry

operates under an extensive regulatory regime, no regulatory changes
have targeted—or are expected to target in the near future—coal storage
activities directly.

Coal storage is primarily used at both the mine mouth and at the
power facility. These storage facilities come in a variety of shapes and sizes,
ranging from simply open storage piles to enclosed bins and silos, which
34
is determined by the need to balance maintaining the maximum amount

of coal in one place with the inherent problems of storing that much coal
in one place—namely dust control, oxidation, or added moisture from the
environment that can cause spontaneous combustion or degradation of
the quality of the coal. These piles can become quite large; for example, a
pile 70 feet high with a 200-foot base can hold upwards of 18,000 tons of
coal, and more massive facilities of even 100,000 tons do exist.
Although many times derided as simply a “big pile of rock,” coal

storage facilities are governed by a surprisingly complex set of operational
rules. For coal that is to be stored for an extended period of time, layers of
coal are “sealed” by interlaying layers of very fine coal to reduce the chance
for air and moisture to enter the pile, which otherwise could catch fire
or significantly degrade in quality. For example, the most common style
of coal storage is simply an open conical pile with the collection point
underneath the center of the pile, which can result in a large amount of
coal in dead-storage between the edge and collection hole. Coal in the
dead-storage area does not turn over as rapidly as the coal in the middle,
meaning that it has a greater chance to decline in heating value and, thus,
usability. To combat this eventuality, the storage facility can periodically
move the coal to the collection point with a bulldozer, fill the dead-storage
areas with earth-fill (significantly lowering the facility’s capacity), or create
many more collection points throughout the pile.

As evidenced by the sheer physical size of the stocks involved, the
historic development coal storage facilities has been more of a physical
hedge against supply disruptions than an economic hedge against vola-
tile power markets. In the producing regions, mine mouth coal stockpiles
were needed to balance the slower production rates and the ability to
fill the unit trains in a timely manner. Along the transportation path,
some coal is held at a trans-shipment point for converting the method
of travel (barge to rail, etc.) or to produce a blended coal product from
35
one or more varieties. However, the vast majority of coal storage is held
at power facilities, where on-site storage prevents supply disruptions
from interfering with generation schedules.
Unfortunately, holding a massive amount of coal on-site entails

carrying charges—expenses modern power facilities would like to avoid
as much as possible. As deliverability of coal becomes more secure, coal
supplies at power facilities have slowly declined for many years, with the
average coal-fired plant in 2004 having only 40 to 45 days of supply on
hand, falling from an average of 60 days on hand during much of the
1990s. Even this is lower than levels common in earlier times of more
strident miner unrest and poor transportation reliability, where many
power facilities targeted a 90-day (or more) stock level—a period that
some say just happened to equal the average length of a United Mine
Workers’ strike in West Virginia and Kentucky coal mines.
Besides the desire to reduce carrying costs, a more integrated power

market has made stockpile management a far more important issue in
overall power plant strategies over the last few years—requiring a closer
monitoring of stock levels, the greater use of blending coals, and a real-time
monitoring of coal quality. One issue in particular has been the focus of this
effort—trimming losses because of weathering of coal to provide sufficient
availability of high heat-content coal during peak demand periods. This
issue becomes more problematic as the seasonal high and low stock levels
are taken into account. Seasonal stock levels can fluctuate by one-third
during the year; however, some plants can experience far greater variability.
This problem became acute during the tumultuous consolidation of the
rail industry in the mid-1990s, when a number of power facilities were
required to reduce their power output because of low stock levels when
deliveries to the plants were delayed significantly—resulting in significant
lost sales. As the integration of the power market continues, the operation
of coal stockpiles will be increasingly affected by volatile prices in power
markets, driving a continued focus on optimizing the stock level as carrying
charges vie for availability.
36
Natural Gas Market

Natural gas has become the most economically important energy
resource for the U.S. power industry. Not in sheer volume—coal still
reigns by far—but for pricing issues, most power plants on the margin are
fueled by natural gas, making it the fuel that sets the price for electricity
much of the time. Demand has also grown significantly, with natural gas
use growing 250% in the power industry in the past 25 years—80% of
the total natural gas demand increase during that period.3 This demand
growth has only been possible through the expansion of the natural
gas storage market; expected continued growth of natural gas power
generation will fuel the need for even more storage.
The natural gas industry and the

need for storage
Wholesale natural gas market. Natural gas is a widely used fuel
that provides a quarter of U.S. energy needs. In 2002, according to the
DOE, 22.5 Tcf of natural gas were consumed domestically, representing
more than a $90-billion retail market. Still dominated by space
conditioning demands in both the residential and commercial markets
(35% of demand), industrial users hoping to reduce emissions and
power generators looking to build combustion turbine–based power
facilities have also been drawn by the fuel’s promise. The natural gas
power generation market has seen the greatest demand growth over
the last decade and is expected to continue growing rapidly, with 90%
or more of new power plant construction natural gas-based. Currently,
power generators use 22% of the natural gas in the United States
(up from 17% in 1990), with this ratio expected to expand to 29% by
2020.4 Space conditioning will likely continue to be a major demand
market for natural gas. However, many industry pundits expect overall
37
industrial use to remain stable as industries that use natural gas as fuel
stocks, (or other price-sensitive industries) curtail their domestic use of
natural gas in response to the elevated wholesale price, which is now,
and is expected to remain, far above historical norms.
North America is largely self-sufficient for natural gas use; only 15%
of natural gas used in the United States is imported, with roughly 90% of
this coming from Canada through pipeline and the remainder entering
the market through the four liquefied natural gas (LNG) facilities—
through which a growing portion of the new demand is expected to be
met (along with new LNG import facilities). As much more imported
LNG becomes essential, associated storage at the terminals will also
become important. Most natural gas used in the United States currently
is produced along the Gulf Coast and in the western states. Although
some of it is co-located with oil deposits, the majority is retrieved from
natural gas–only deposits in either salt or permeable rock layers. Because
the largest demand markets are in the upper Midwest and along the
eastern U.S. seaboard, long-distance interstate pipelines are required to
deliver the natural gas to market.
With more natural gas being delivered to a wider variety of end

users, a marketing and trading component grew up alongside the
physical delivery of natural gas. Besides promoting lower costs, reducing
volatility, and providing additional choices for consumers, this trading
also supported the development of a forward price curve for the market.
Initially, the trading was essentially physically based; however, the need
and desire for hedging/speculation promoted the development of
financial trading that dominates the market today. Because of the
integrated nature of the trading, each unit of delivered natural gas on
average is actually traded 3 times or more. The financial trading market
is far more active, with the number of contracts being traded actually
equaling roughly 10 times the amount of natural gas delivered. However,
almost 98% of these contracts are settled prior to delivery.
38
The need for storage. Natural gas storage is the primary means of
managing fluctuations in supply and demand and is an essential component
of a low-cost, efficient, and reliable interstate transmission and distribution
network. Natural gas storage is used for primarily two purposes: to meet
seasonal demands for natural gas and to meet short-term peaks in demand
that can range from a few hours to a few days. Because there is a distinct
seasonal variation to natural gas demand stemming from the heavy space
heating demand, storage’s impact is significant, making these facilities an
integral part of the natural gas industry. For instance, the DOE estimates
that more than 15% of all natural gas delivered resides in a storage facility
at one time, and on some peak days, stored natural gas delivery accounts
for more than 30% of all daily natural gas demand.5
Seasonal variation in demand poses a persistent challenge to the

market. How does one economically serve a market a few thousand miles
away when most of the demand only happens during a few months of the
year? Early on, the natural gas industry decided to solve its annual cyclical
demand problem by injecting natural gas into underground reservoirs
near the demand regions during the summer and releasing it during the
winter to meet the peak demands. The first underground natural gas
storage field in the United States began operation in 1916 near Buffalo,
New York. Providing additional supply during peak demand from natural
gas produced in the summer was very attractive to distributors and utilities
in the Northeast. At that time, the steel industry was not capable of
producing pipelines of the size required to meet the winter peak demand
loads of these regions. The only other option was to lay additional, smaller-
diameter lines, which would have been uneconomical because they would
have been mostly unused during the summer.
Therefore, underground natural gas storage facilities in the

downstream market areas were the only feasible option to allow the
industry to continue expanding far from the producing regions. As
they grew to become an integral part of the supply network, the long-
haul pipelines could remain at nearly full utilization throughout the
39
year—which kept the industry profitable—and continued to draw in

investment for additional expansion. The DOE, in fact, estimates that
natural gas storage is responsible for avoiding 50% of the required
transmission system upgrades to the systems. 6 Over the following
decades, underground storage capacity continued to expand and was
used by pipelines and local distribution companies (LDCs) alike to
optimize long-haul pipeline capacity by balancing the flow of natural gas
with the injection or removal of natural gas as needed. As the physical
natural gas demand increased, so too did the need for more natural gas
storage facilities.
Regulatory impact on storage. Regulatory changes have

significantly impacted natural gas storage. Before deregulation of the
natural gas industry 20 years ago, expansion of the natural gas storage
market was used primarily to defer pipeline capacity additions or
upgrades; natural gas was simply shipped to and stored in underground
reservoirs during the summer for use during the winter months. As part
of the restructuring of this industry, natural gas storage was separated
from the bundle of services previously rolled into the transmission cost
of natural gas, and end users were forced to choose how, or whether,
they would use the service. The results have been impressive because as
the market changed, so too did storage’s role. Storage originally had little
capability to respond dynamically to the market. However, in response
to the need to support shorter-term storage demands, far more flexible
storage facilities have dominated new development.
The origin of today’s natural gas industry traces its roots to the
Natural Gas Policy Act of 1978, which reversed an earlier decision
to establish price controls over interstate natural gas prices, giving
the Federal Energy Regulatory Commission (FERC) jurisdiction over
interstate pipelines and, in subsequent FERC orders, their storage
facilities. In the early 1980s, FERC attempted to instill competition
in the wholesale market through FERC Orders 436 and 500, which
allowed consumers to contract with the pipeline to reserve capacity
40
for their own use and allowed natural gas producers to sell directly to
consumers. By the late 1980s, the federal government forced pipeline
companies to break apart their total delivery charge of natural gas into
separate services and made plans for a more far-reaching change. In
1992, FERC moved to change the roles of the existing players in the
market and created new opportunities and responsibilities for a host of
others through FERC Order 636. This later order brought together the
open access concepts found in FERC Orders 434 and 500 and outlined
the unbundling of services provided by the pipelines by identifying
and separating those services—one of those being storage. In Order
636, FERC concluded that natural gas pipeline companies needed to
separate their natural gas sales operations from their transportation
services, open their trading and transportation information systems to
third parties without discrimination, and adhere to new industry-wide
standards of behavior aimed at leveling the playing field. With these
changes in place, the DOE estimated that 80% to 90% of all working
natural gas in underground storage would become available to end-use
customers. This had become a primary goal of the federal government
in its attempt to foster a secondary market for storage capacity, and,
hopefully, additional services to the end-use customer.
The availability of storage to the end user was a key goal of the
government because of the optionality storage brought to the market.
FERC Order 636 also changed the role of existing participants in the
underground storage market and opened the door for a host of new
players. Interstate pipelines were allowed to retain ownership and
use their storage facilities for load balancing and related services,
but since then, these facilities have been required to operate as
service providers to any shipper or end-use company and to allow
the customer to contract for its own natural gas service. Intrastate
pipelines and producers continued to use storage to increase peak
demand deliverability and to improve the system reliability because
they were regulated by state, not federal, regulations. In both cases,
41
these pipeline firms continued to invest in storage to both enhance the

efficiency of their transmission network, and to extend their offerings
of higher-margined services. As planned, consumers such as LDCs and
large commercial and industrial (C&I) firms began to use storage to
supplement their winter season. Through using storage, they could
effectively hedge against decreases in natural gas supplies, arbitraging
the difference between winter and summer prices with storage fees.
Marketers also used storage to support arbitrage and hedging of natural
gas contracts; although possessing greater risk, this strategy held out
the greatest promise for profit and was, therefore, better suited for
independent players than the more regulated actors in the market.
Besides providing the opportunity for additional storage use,

FERC Order 636 also changed the economics of natural gas delivery
by providing additional incentives for customers to take advantage of
storage. This extremely important change raised the cost of pipeline
transportation for consumers and resellers with a low load factor
(but wide variations in natural gas use) by mandating a straight-fixed-
variable (SFV) rate design. SFV essentially shifted all transmission
fixed costs to a monthly demand charge for reserving the pipeline
capacity. In response, some shippers increased their storage capacity
to offset reductions in pipeline capacity reservations. Another
impetus for shippers to use storage facilities under FERC Order
636 was that pipeline companies were given significant latitude in
penalizing shippers with out-of-balance accounts (monthly true-ups).
Shippers with fluctuating demand were again the hardest hit, further
promoting the use of short-term storage to maintain balance (load
management) and thus avoid penalty.
Besides federal action, state-level unbundling of natural gas pipeline

services has also affected the evolution of natural gas storage usage. As
state public utility commissions (PUCs) unbundled retail natural gas
42
services, LDCs and other large end-use consumers became more peak
sensitive and required more balancing of services from storage facilities
for their trading activity across the various shippers.

Storing natural gas underground can be accomplished in more than
75% of the United States, with different geological formations supporting
different types of natural gas storage facilities:
• Depleted reservoirs. Depleted natural gas or oil fields can

become storage facilities by simply refilling the permeable
rock structure that held previous hydrocarbon deposits, taking
advantage of existing wells, gathering systems, and pipeline
connections and saving significant start-up costs. Because of
their geological stability, there are few leaks during operation.
• Aquifers. Some aquifers (mostly in the midwestern United
States) have been converted to storage reservoirs. With an
impermeable layer above, these facilities create a bubble of
natural gas within the water-bearing layer. Unfortunately, there
is a significantly greater risk of leakage because the original
structure did not hold natural gas under pressure. Also,
significant amounts of water can be included in the natural gas
during withdrawal, requiring additional de-watering equipment.
• Salt caverns. The most recent type of natural gas storage
facility is an open cavern in a salt dome created by hollowing
out the volume with a water jet. Because of the reliance on
salt domes, these facilities are limited primarily to areas with
suitable salt-bearing geology, found primarily along the U.S.
Gulf Coast. A few salt caverns have also been constructed in
salt formations in northeastern, midwestern, and southwestern
states. Cavern construction is more costly than depleted fields,
but the facility has a far greater operational flexibility and
capability and essentially no leakage.
43
The different physical properties of varied geological formations

(porosity, permeability, and retention capability) determine how that
natural gas will be stored, and at what rate and to what extent the
natural gas can be retrieved—leading directly to operational uses for the
natural gas in storage:
• Base gas. The volume of natural gas that becomes

permanent inventory to maintain adequate pressure and
deliverability rates throughout the withdrawal season. Some
of this natural gas is permanently unrecoverable, trapped by
physical forces in the pores of the facility walls.
• Working gas. The volume of natural gas that is intended for
cyclical injection and withdrawal for sale to the marketplace.
Therefore, this is the real measure of interest to the market.
• Deliverability. A measure of the amount of natural gas
that can be withdrawn from a storage facility daily. The
deliverability of a given storage facility is variable and
depends on factors such as the pressure within the reservoir,
compression capability of the reservoir, and so forth. In general,
a facility’s deliverability rate varies directly with the total
amount of natural gas in the reservoir; it is at its highest when
the reservoir is full and declines steadily as working natural gas
is withdrawn.

Storage facilities located in different geographical areas lend themselves
to different market roles. Like oil storage facilities, natural gas storage facilities
in production regions allow for the field’s output to be disconnected
from production; otherwise, the output of the wells must be massaged to
match demand. This lowers the life of the well and requires fewer, but more
productive wells—improving the economics of the field. Along the pipelines
44
and in the distribution markets, storage facilities provide much-needed

physical balancing services. This enables pipeline companies to optimize
pipeline pressure and overall performance by maintaining as constant and
high a flow rate (pipeline utilization) as possible—allowing pipeline firms to
maintain service with lower capital investments.
Marketing diverse facility types is also tied to their varied capital and
operating costs. Because the majority of natural gas storage is used for
seasonal heating demands, depleted natural gas reservoirs and aquifers
make up the largest component of underground natural gas storage,
comprising 95% of the total working gas. In 2000, it cost on average
$0.48/MMBtu per season to store natural gas in a depleted field.7 These
facilities cycle once per year, injecting natural gas from May to October
(214 days) and withdrawing it from November to April (151 days).
Because it takes approximately 180 days to fill and approximately 120 days
to withdraw, the operators and customers of these facilities are generally
price insensitive as they have very little operational flexibility. With the
capability of only one cycle per season, the profitability of the facility
becomes more of an issue of seasonal price changes than operational
activity. Because of the low deliverability of these primary storage
facilities, the amount of working gas in storage has become a leading
indicator of short-term natural gas prices, especially during the later half
of a heating season when stock levels—and thus deliverability—make
the market more susceptible to volatility in a cold snap.
Unfortunately, the increasing demand from natural gas–fired power

facilities (mostly during the summer) is putting stress on this historical
pattern and affecting the seasonal demand injection/withdrawal
schedule. As more natural gas–fired plants are brought online to
supply power for cooling loads, these plants compete for natural gas
supplies in the middle of the injection season. The need for more
flexibility—higher deliverability—has, therefore, put a premium on
salt-dome storage facilities. These facilities only require approximately
20 days to fill and approximately 10 days to fully withdraw the natural
gas in storage (depending on size). Although they represented only
45
4% of total working gas capacity in 2002, they provided 17% of total

deliverability.8 With their faster deliverability, the salt-dome facilities
provide a host of load management services:
• Peak supplies for sudden weather shifts

• Load balancing for shippers (to avoid paying imbalance fees)
• Emergency supply backup in case of supply disruption (hedge
for end users against price changes)
• A tool for marketers to speculate on price changes
Although salt domes have high capital costs, their greater flexibility
gives them a much lower operational cost—making them profitable over
a wider range of market conditions, especially multiple cycling throughout
the season. For instance, salt-dome storage costs $1.08/MMBtu for one
cycle per year, but this cost drops to $0.28/MMBtu when the facility is
cycled five times.9 Therefore, this ability to cycle its working gas several
times in a season can easily offset the additional cost of high-deliverability
storage—if the operators correctly anticipate most of the market price
swings. With this flexibility from high-deliverability storage, the value
potential of the natural gas in salt caverns changes from simple long-
term arbitrage to market timing—something of far more value.
Lessons Learned
Groups involved with setting public policy and developing energy
storage facilities in the power industry can learn much from how other
energy markets have integrated storage into their value chain. These storage
assets have proved to be a key component for providing reliable service,
low prices, and flexibility in each of these other energy markets. Although
the power industry is obviously quite different in many ways, many of the
economic forces are similar (especially in the natural gas industry), leading
to operational insights and leverageable market strategies.
46
In these other energy markets, the introduction and use of storage has
undergone a surprisingly similar learning curve. Although initially used to
level variations in supply, a more integrated use of these facilities has resulted
in fewer required transmission upgrades and lower system expansion costs.
Deregulation in these markets—again with particular respect to natural
gas—enabled the storage facilities to provide ready reserves and create
greater customer choice, and, surprisingly, it created a greater demand for
the use of storage facilities. As wholesale physical trading begat even more
financial trading, storage assets served not only as the basis for many of these
contracts, but also as a tacit physical backstop for the trading activity.
Although the petroleum and coal storage industries provide valuable

lessons, the natural gas industry has especially valuable and relevant lessons
to teach the electric power industry in the use of storage in the wholesale
and transmission markets. There, the use of storage facilities first was
recognized as a means to lower network expansion costs, and they have
been incorporated into a nationally integrated production, transmission,
and end-use service network. This did not happen overnight, but rather
through a combination of regulatory prodding and innovative use of new
technology. This storage capability has proved to be a crucial part of the
natural gas industry—without it, the industry would not currently be able
to deliver the capacity or provide the flexible services that were required
for market expansion. Nor would the natural gas industry have been able
to emerge so successfully from restructuring without a storage component
to dampen a wholesale market, provide optionality, and raise the level of
confidence in the market. As the electricity market looks to complete its
own restructuring (or simply start again), it can look to this history in the
natural gas industry for guidance.
The most valuable lessons learned for the electric power industry are
focused along three avenues:
1. Value of the storage asset

2. Impact of technological advances
3. Impact of regulatory reform
47
Through reviewing the experience of these markets’ use of storage

assets as they underwent change, groups looking to expand the presence
of energy storage assets within the electric power industry can gain some
significant understandings in what works, and what does not.
Value of storage
The value of a storage facility is linked to its usefulness, which is
related to the
• Speed of the market

• Scale and location of the facility
• Deliverability of the facility
These criteria dictate much of how a storage facility is able take

advantage of changing demand patterns and market activity to provide
enhanced services to the customer. As a market matures and integrates
storage deeper into its operations, the uses, as well as the values derived
from those uses, will change. The maturity of the industry’s use of storage
is thus implied by the leveragability of these assets.
The speed of a market describes the rate at which the price and
availability of a commodity change. Slower, physical-delivery commodity
markets can largely use the flow of the commodity only to replenish
physical demand or rebalance a shortfall. However, as activity in the
market increases, the ability to leverage the stored quantities of the
commodity increases and raises a storage facility’s inherent value. For
instance, a highly valued storage facility could take the form of a highly
deliverable ready reserve to rebalance a supply-demand imbalance—one
where it would provide a value through action or as a reserve unit. This
speed-to-value index is taken to the extreme example in the electric
power industry, as there is effectively no commodity flow to replenish
a shortfall; demand cycles that happen daily here can take weeks or
months in the natural gas market.
48
The storage facility’s location and scale are other aspects that
affect how the unit will be used, with increased use translating into
higher value. One of the initial values ascribed to a new storage facility
is its ability to preclude the outlay of additional capital for building or
refurbishing transmission facilities that would be used to supply peak
demand periods—because the peak demand occurs infrequently, the
entire system is then vastly underused. In the natural gas industry, storage
has been rightly credited by the DOE with avoiding 50% of the required
transmission system upgrades and allowing the average utilization of
the system to remain more than 90%.10 This contrasts with past electric
utility strategies that dictated generation and transmission capacities
to be so overbuilt that the system’s current utilization rate is effectively
capped at 60%. Positioning a cache of supply past a capacity-constrained
portion of the transmission system also creates additional value to the
commodity if it can be quickly delivered. This strategy of capturing
locational scarcity rent was a central theme in many merchant power
facilities that were built during the 1990s construction boom looking for
transmission cul-de-sacs or load pockets.
Finally, the last issue driving the value of a storage facility is the
deliverability of that unit. For some applications, a slow withdrawal
capability means the facility is essentially useless, leading to a poor value
attributed to that asset for that particular application. With greater
deliverability, the range of applications grows, making the unit more
flexible and responsive to market needs. For instance, experience with
salt-cavern storage facilities has also shown an increased number of
shorter-duration injection and withdrawal activities throughout the year,
expanding the possibilities for new roles as a swing-supply—especially
for the growing demand from power facilities. As high-deliverability
storage has become integrated into short-term supply strategies, the
market has shown its increasing comfort level in storage’s deliverability
by holding lower storage inventories. Figure 2–1 shows that as storage
throughout the 1990s grew to provide first 10% and then 15% of all
demand—increasingly from a growing number of high-deliverability
49
salt-dome facilities—the amount of natural gas in storage actually fell

as a portion of total use, signifying the stabilizing component storage
provided to the market.
Fig. 2–1. High deliverability enhances the value of storage

(Data: US Department of Energy).
Technological change
Technological change can be driven both from inside and outside
a particular market. When the natural gas industry was new, the
inability of the steel industry to produce pipes large enough to meet
summer demand levels drove the need to store large quantities of
natural gas underground near the demand regions. This quandary led
to a much higher use and efficiency of the entire industry. Subsequently,
ever-larger coal facilities in the 1950s and a declining competitiveness
of coal in general because of high transportation costs led to the unit
train delivery of coal—requiring much larger and better managed coal
stockpiles. Because of the improved reliability of the entire supply chain,
50
coal stockpiles have been able to decline for more than 20 years, greatly
lowering the carrying cost of the coal and thus improving the operating
margins of the facilities.
New technologies can change the existing economics in the

market, which in turn make the introduction of more innovative
technologies easier. The introduction of underground natural gas
salt-dome storage facilities is a good example of how new storage
technology with a lower operating cost can alter the use of storage
within a market. One of the central benefits of salt-dome storage
facilities is the much higher deliverability over other types of natural
gas storage—allowing these facilities to provide multiple injection
and withdrawal cycles per year. Although these salt-dome storage
facilities are more capital intensive, their ability to have upwards of 10
cycles per year allows for the average injection and withdrawal cost
to be lower than the depleted natural gas field storage facilities. For
instance, according to Simmons & Company International in 2000,
it cost on average $0.48/MMBtu per season to store natural gas in
a depleted field. That same cost was $1.08/MMBtu in a salt-dome
facility for 1 cycle per year, but this dropped to $0.28/MMBtu when
the facility was able to cycle 5 times per year.11 This advantage allowed
salt-dome facilities to spring from essentially 0% of the market in 1990
to more than 15% of all natural gas in storage by 2002, and nearly
20% of deliverability capability for all storage facilities.
The electric power industry is arguably the energy market most

affected by technological change, with this change coming through
either revolutionary (disruptive) or evolutionary means. However, as
the production and delivery of power is an infrastructure-oriented
market, the electric power industry is hardwired toward accepting
evolutionary, rather than disruptive technology. Because electricity
storage technologies are more geared toward improving the efficiency
and operational capability of other components in the system rather
51
than “establishing a new paradigm” in how the industry operates, these

energy storage technologies should find a receptive audience, providing
the technology is mature and cost-effective.
Regulatory change
Regulatory change can have a profound effect on the use of storage
facilities. Because of the importance of energy markets to the national
economy, regulators have long looked for ways to leverage any assets
within a particular market to create greater customer choice and lower
costs. For instance, in the petroleum industry, strategic petroleum reserves
have been created (in many countries) to act as a buffer for major supply
disruptions. Acting together, the groups controlling these reserves can have
a large impact on the market, which makes their mere presence a means
to lower supply uncertainty and, hence, prices. Regulatory changes are at
the heart of the wholesale power and transmission market opportunities
facing energy storage in the electric power industry today.
Regulatory action within the natural gas industry provides a good

example of how policy makers can successfully leverage storage assets.
Consumer demand for contract relief and for more choices in natural
gas delivery options led the federal government to desire a more self-
regulated and independent storage market. By forcing the unbundling
of the delivery charges (FERC Order 636), the transportation, storage,
handling, and delivery services were then priced competitively; natural
gas storage facilities were able to operate as third-party service providers
in a deregulated wholesale market, serving a much larger number of
end users. After delivery charges were unbundled, demand for more
reactive storage services expanded. It is widely believed that the natural
gas industry would not have been able to emerge so successfully from
its restructuring without a storage component to dampen a potentially
volatile wholesale market, provide optionality for sellers and buyers alike,
and raise the level of confidence in the market.
52
With the passage of FERC Orders 888, 889, and 2000, the electric
power industry finds itself in a position similar to that of natural gas
after FERC Order 636 was passed. Unfortunately, storage assets simply
do not exist in the electric power industry to the same degree they did
in the natural gas industry prior to deregulation. Because of the inability
to economically transmit electric power more than a few hundred
miles (normally), the electric power industry developed into a far more
vertically oriented industry compared to the horizontally oriented natural
gas industry—leading to the early recognition of the value of storage
facilities by a wider number of market participants in the natural gas
versus the electric power industry. For this reason, a single natural gas
storage facility was able to serve a number of customers easily once the
delivery charge was unbundled.
As the electric power industry moves toward a similar horizontal

focus, large-scale energy storage facility development may mimic
the experience of natural gas storage facilities. For example, current
pumped-hydro facilities provide significant power grid stabilization
services to the utilities that own them, but if they were slated to operate
in a merchant role through an ISO, their stabilizing services could be
provided to a far wider audience. For the power industry to mimic the
natural gas industry, ancillary services must be further broken out—in
a fuller and more systematic way (i.e., replicated nationally)—from the
delivery charges. With the market determining the cost of (and need
for) these essential services, this action would significantly support the
development of a self-regulating market. As in the case of early natural
gas storage, federal involvement will be key; not to give electricity
storage facilities an advantage, but simply to level the playing field to
create a market for their services.
53
References
1. Energy Information Agency. 2002. Annual energy review 2002,
159–211. Washington, DC: U.S. Department of Energy.
2. Energy Information Agency. 1997. Causes and effects of lower
inventories. Petroleum 1996: Issues & trends. Washington, DC: U.S.
Department of Energy.
3. Energy Information Agency. 2004. Natural gas. Annual energy review
2002, 213–232. Washington, DC: U.S. Department of Energy.
4. Energy Information Agency. 2004. Annual energy outlook 2004.
Washington, DC: U.S. Department of Energy.
5. U.S. Department of Energy. Gas storage white paper (draft).
Washington, DC: 6100 Fossil Energy Working Group.
6. Ibid.
7. Dietert, J. A., and D. A. Pursell. 2000. Underground natural gas storage.
Houston, TX: Simmons & Company International.
8. Energy Information Agency. Annual energy review 2002, 213–232.
9. Dietert and Pursell. Underground natural gas storage.
10. U.S. Department of Energy. Gas storage white paper.
11. Dietert and Pursell. Underground natural gas storage.
3 ELECTRICITY
STORAGE
TECHNOLOGIES
As the name implies, energy storage technologies absorb electrical

power during periods of excess capacity so that it can be released later
when it has more value. When released, the energy can either be delivered
in large amounts for commodity use, or in a measured, controlled manner
to optimize the operation or enhance the reliability of the power delivery
system. The storage unit can also be run by constant charge/discharge
cycling to dampen unwanted fluctuations in the system’s power level.
Alternating current electricity itself is not stored directly, but rather, it
is converted and stored by mechanical, chemical, or electrical potential
energy methods. Each of these storage methods has its own particular
operational range and capabilities, and the use of a particular method
usually predisposes the storage unit into a set of applications for which
the facility’s technology is best suited.
56
System Components
Each energy storage unit facility is made up of three distinct
components: the storage medium (or subsystem), the power conversion
system, and the balance of the plant. The size and cost of each of these
components vary, based on the setting and the application for which
it is used, leading to a range in costs even within one technology.
These costs will decline as the core storage technology matures,
improvements are made in the production processes, and economies
of scale such as production levels continue to rise for many of these
recently commercialized technologies. In the near term for technologies
in pilot-phase installations, initial engineering and installation costs
can also be significant because of the lack of experience—a cost
component that also declines with wider commercialization.
Storage medium
The heart of every energy storage facility is the energy reservoir or
storage medium, which can take the form of mechanical, chemical, or
electrical potential energy. The differences in how energy is stored in
the various storage media help to define each technology’s capabilities,
leading to one or more technologies being better suited for certain
applications than the others. For instance, large-scale storage technologies
that can arbitrage energy between long time periods rely on air or water
because they both are low-cost and both can be stored a long time
without loss. Storage media that require greater energy expenditures for
support systems such as necessary environmental controls for the storage
medium would not be suitable for such roles. However, if their storage
medium was able to cycle through a charge/discharge cycle quickly and
repeatedly without damage to the unit, they would be a good choice for
other applications.
E L E C T R I C I T Y S T O RA G E TE C H N O L O G I E S
57
Storage medium costs vary widely because the different media vary
significantly from air to chemical electrolytes. An important aspect to
their costs is the energy density of the medium; high-density storage
media allow for smaller supporting equipment, whereas lower-density
material requires a large storage facility, with its associated expenses.
Structurally, the costs for each storage medium can generally be broken
down into two components—the initial capital cost of the medium
itself (including replacement material), and the costs associated with
maintaining the storage medium in a charged state until a discharge is
requested. For most systems, the cost of the storage medium is one-half the
cost of the entire facility, and it is expected to remain in that range for the
foreseeable future. The reason for this is that as the level of technological
progress continues to improve, the cost of the overall unit will decline
(other subsystems are improving as well) for the same application, or
more capable units are developed to address larger applications.
Power conversion system

All storage technologies, except for some mechanical energy storage
systems (i.e., pumped-hydro storage or compressed air energy storage
[CAES]), need dedicated power electronics for alternating current–direct
current (AC–DC) and DC to AC conversion (in pumped-hydro and
CAES plants, these are part of the motor/power generator train). A
power conversion system (PCS) acts as the electrical interface between
the customer or utility power grid and the storage subsystem, and
consists of such components as a converter, DC and AC switch gear, and
programmable high-speed controllers.
The PCS of an energy storage subsystem has two major roles. First, the
PCS is used to convert the AC electrical energy of the power grid into DC
for storage. When the storage system is being charged, the converter acts
as a rectifier (changing AC to DC); during discharge, the process reverses
and the converter operates as an inverter, changing DC to AC. Second,
the PCS conditions the power during conversion so that no damage is
58
done to the storage facility or the customer’s electrical system. A related

system of power electronics monitors and controls the condition of the
storage subsystem (this is of a greater concern when the storage medium
is a chemical-based battery that can be damaged through misuse). The
actual PCS for any one particular storage technology or installation is
heavily dependent on the balance of capabilities desired (reliability,
responsiveness, etc.), which, in turn, affects the cost of the unit.
The U.S. Department of Energy’s (DOE) storage program has identified

the customization of the PCS for each storage technology as one of the
most important energy storage system integration issues. Because each
energy storage technology operates differently during charging, standing
by (floating), or discharging, the life span of the storage facility will be
greatly reduced if the PCS does not take that into account. In particular,
the design and cost of a PCS greatly depends on the primary activity
of the storage facility, and thus most PCS components are generally
compared on a dollars-per-MWh (commodity arbitrage) or dollars-per-
MVA (power quality) basis, depending on the need for real or reactive
power (which the PCS provides).
In terms of capital expenses, PCS costs for standard-scale installations

are typically in the range of $100 per kW to $200 per kW for those
focused on energy arbitrage (MWh) to $200 per kW to $400 per kW for
those focused on power quality (MVA). Because customization is still the
rule rather than the exception, the PCS often represents anywhere from
33% to nearly 50% of the entire cost of the storage installation (depending
on the application and scale of the facility), but continued effort keeps
reducing the costs of these units. With advances toward modularization
and mass production of these units, unit costs are expected to decline
further, with those systems focused on energy management estimated to
see declines of 10% from current levels. Power-quality PCS systems could
see price declines in the coming years of up to 25%.
59
Balance of plant
The balance of plant (BOP) encompasses the facility (and control
systems) to house the equipment, the environmental controls, and
the electrical connectors between the PCS and the power grid.
This aspect of an energy storage installation can vary tremendously
based on the requirements of not only the technology, but also the
application. For example, the BOP typically includes transformers,
electrical interconnections, surge protection devices, a support rack
for the storage medium, the facility shelter (or component of total),
and environmental control systems. Because storage media can
operate throughout a wide range of environmental conditions, the
space conditioning is sometimes the target of cost cutting. Although
undersizing of the electrical components is not often an option,
the supporting, particularly the environmental, controls in existing
(normally lead-acid batteries) energy storage installations are sometimes
lacking, especially if the storage unit is operating in a reserve or backup
power mode. Unfortunately, savings here will many times shorten the
operating life and hence increase replacement costs, driving up the
overall cost for the use of the unit in question.
For these reasons, the BOP is by far the most variable cost component
of an energy storage facility. In general, these costs tend to rise when the
installation is more customized and are less when modular storage units
are used. Normally, however, the BOP costs represent 10% to 25% of the
total cost for a typical storage facility. For example, lead-acid batteries
should be maintained within a certain operating range to prolong their
operating life, necessitating space conditioning for the sometimes-
significant floor space required to house the units. Alternatively, flywheels
are far more compact and can operate in a much wider temperature
range—mostly obviating the need for space conditioning of the unit.
60
Engineering, procurement,
and construction
Similar to the BOP costs, the engineering, procurement, and
construction (EPC) costs to build an energy storage facility are highly
dependent on the complexity of the equipment and the amount of
preparation required for the site. Overall, these engineering and construction
costs can add an additional 10% to 15% (or more as the uniqueness level
rises) to the total cost of the facility. Some reduction in these costs will
be expected over time as engineering service companies gain experience
installing energy storage technologies, and the continued modularization
of the PCS and storage subsystems allows for a plug-and-play installation.
Pumped-Hydroelectric
Storage (PHS)
Summary
Pumped-hydroelectric (hydro) storage (PHS) is the oldest and
largest of all of the commercially available energy storage technologies.
This technology also retains the largest installed base (by capacity) of
any storage technology with more than 20 GW in the United States,
where it currently represents roughly 2.5% of total summer generating
capability. Conventional PHS facilities (fig. 3–1) consist of two large
reservoirs, one located at a low level and the other situated at a higher
elevation. During off-peak hours, water is pumped from the lower to
the upper reservoir where it is stored. During peak hours, the water is
released back to the lower reservoir, passing through hydraulic turbines
to generate electrical power. These facilities generally operate on a
daily schedule, with some facilities also operating as a conventional
hydropower facility for irrigation or other public uses. Older designs
61
had round-trip efficiencies of more than 60%, but repowering of some

plants recently has pushed this higher than 75%. Although it is a
proven and valued technology, difficulties in siting new facilities and
total system costs that can reach upward of $2,000 per kW curtail most
future prospects for development, especially in developed countries.
Fig. 3–1. Pumped-hydro storage (Courtesy of Montgomery Watson

Harza (MWH)).
Historical origins
PHS units first appeared in Italy and Switzerland in the 1890s,
and arrived in North America in 1929, when the 32-MW Rocky River
pumped-hydro storage facility in Connecticut entered service. Moving
into the 1950s, PHS facilities were becoming more widespread as the
designs for them began incorporating a single reversible pump-turbine—
previous designs incorporated separate pump impellers and turbine
runners. Besides being more capital intensive, these previous designs
limited the unit’s reaction time to reverse direction as the channel not
being used had to be drained of water. The peak construction period
for these facilities in North America stretched from the 1960s through
the 1980s, as utilities found these units to be a valuable way to manage
system loads throughout their service territories. Pumped-hydro
62
facilities were, in fact, the only commercially available energy storage

technology for utility-scale applications until the 1970s. Interest in
pumped storage waned in the 1990s because of the technology’s high
capital costs and difficulty in locating new operating sites because of
environmental concerns.
Design and operations

A typical PHS facility consists of two large reservoirs, one located
anywhere from 30 m to 650 m above the base level, with 300 m of
hydraulic head (the difference between the two reservoirs) generally
considered the preferred height for new development. These facilities
can either use freshwater or seawater as the working medium. As PHS
facilities share a large portion of their equipment with the much larger
conventional hydroelectric industry, this has allowed them to take
advantage of advancements in variable-speed motors and improved
impellers to increase their capability and efficiency.
During off-peak hours, water is pumped from the lower to the

upper reservoir, where it is stored. To generate electricity during peak
demand hours, the water is released back down to the lower reservoir,
passing through hydraulic turbines and generating electrical power.
Energy in PHS facilities is stored as potential (mechanical) energy; water
is the working fluid, and gravity provides the driving force. Whereas the
generating capacity (kW) is a function of the rate of water outflow and
the hydraulic head, the energy (kWh) that can be stored in the facility
is a function of the reservoir volume and hydraulic head. Therefore,
attaining the greatest hydraulic head for the facility is normally the most
critical design issue, because the ability to expand the upper reservoir (to
expand to the facility’s storage capacity) is usually more expensive and
difficult. This aspect allows variations in the design for a facility with the
same storage capability; a larger head requires a smaller volume of water
to provide the same amount of energy storage as a larger and lower
reservoir. Increasing the hydraulic head also requires smaller-diameter
63
turbines and the accompanying capital equipment. Assuming that a

wide, straight pipe links the two reservoirs (to reduce frictional losses)
and that evaporation is minimal, the main losses in PHS storage operation
stem from the inefficiencies of the pump and turbine.
PHS facilities possess a fast reaction to go along with their large size,
providing a unique capability to the system operators, which allows
them to cover a wide range of applications within the wholesale power
market—including providing reliable system-wide frequency regulation
and contingency reserves in addition to arbitraging commodity power
for sale. Although the water held in the upper reservoir could be held
there indefinitely, PHS facilities generally operate on a daily schedule—
although it is not uncommon for a facility to operate upward of eight
times per day. Compared to generation units, these facilities are able to
quickly respond to changing market conditions, as short as 10 minutes or
less from complete shutdown (or full reversal of operation) to full power;
maintained on standby, these facilities can even reach full power within
10 to 30 seconds. As these facilities can also reverse the direction of flow
quickly, they provide an important balancing role for system stability as
well, switching from absorbing a few hundred MWs of excess system
power to producing an equal amount quickly in case of an emergency.
Over a longer-term horizon, this strategy is also used to balance or
optimize thermal facilities elsewhere on the power grid. These attributes
help PHS facilities maintain an availability rating that can surpass 98%,
and a forced outage rate below 1.5%.
Whereas older designs had round-trip efficiencies of less than

60%, improved designs coupled with repowering with modern control
equipment have recently pushed this higher than 75%. Besides losses
incurred as the water travels through the motor/turbines, the efficiency
of these facilities is also greatly affected by the effective head loss because
of friction and turbulence in the pipeline connecting the upper and lower
reservoirs; shorter distances and larger pipes incur less energy loss. The cycle
time requirements for these facilities can be expressed as 1:1 for charge/
64
discharge at full power, but operational requirements of the local power

grid normally dictate the use of the unit as a flexible system optimizer
instead of a rapid commodity energy sink/source. The effective cycle life
of these systems is normally viewed in decades of operating rather than
a set number of cycles, as uses such as providing frequency regulation
would not require full charges/discharges; component replacement and/or
upgrades are more along the lines of other power facilities.
Cost issues
PHS facilities are capital intensive, as their costs include not only
significant civil construction, but also reversible pump-generating
units and reservoir(s) that require frequent drawdowns. In total, these
requirements push the average capital costs to $1,500 per kW to $2,000
per kW for development costs. Because most pumped storage facilities
are developed with many hundreds of MWs, the total cost for such a
facility will normally reach into the hundreds of millions of dollars in
developed countries such as the United States or in Europe, where
environmental mitigation requirements exist. However, with much of
the capital equipment already in place, existing sites provide extremely
cost-effective upgrade opportunities of core components such as the
impellers. To improve the economics of this technology, developers and
engineers are looking at creative ways to reduce costs when constructing
new projects. Some of these ideas have included using exhausted mines
as the lower reservoir (including sealing the cavern walls) or adding a
PHS unit to an existing hydroelectric facility.
Installations
Typical PHS turbines range in size from 30 MW to 350 MW, with
facility capabilities ranging from 300 MW to 1,800 MW. Currently more
than 90 GW of capacity exist in more than 240 PHS storage facilities
around the world—roughly 3% of global generating capacity. This is an
65
increase from 1990, when only 74 GW of installed generating capacity

existed. The current 20 GW of capacity in the United States (up from
15 GW in 1990) represent 2.5% of the nation’s summertime generating
capability. Other developed countries have also invested heavily in these
facilities, with Japan maintaining roughly 10% of its total generating
capacity in PHS storage facilities, and European Union (EU) countries
maintaining more than 32 GW of PHS storage capacity.
Example—Rocky Mountain, GA.1 The Rocky Mountain facility—

developed in 1991—is a three-unit, 848-MW pumped storage hydroelectric
plant located in Floyd County near Rome, Georgia. Oglethorpe Power’s
Rocky Mountain pumped-storage facility was originally intended to run in
a peaking mode, but in its first calendar year of service ran every day except
one to take advantage of strategic power marketing deals. The facility won
the 1998 Powerplant Award from Power Magazine for demonstrating the
value of large-scale energy storage to a regional power market.
Example—Dinorwig, Wales, UK.2 The Dinorwig plant in Wales,

UK, is one of the most well-known pumped storage plants in the world.
It was constructed between 1976 and 1982 in Europe’s largest man-made
cavern under the hills of North Wales. Each of the station’s six generating
units acts as both pump and turbine, delivering 317 MW of power,
sustaining 1,800 MW for a total of five hours. Working volume for the
facility is 6 million cubic meters, with a head of approximately 600 m. If
held as spinning reserve, the entire plant can reach maximum output in
less than 16 seconds.
Example—Okinawa, Japan.3 The world’s first seawater PHS plant

was built in Kunigami Village, Okinawa Prefecture, Japan, in 1999 with
a rated capacity of 31.4 MW. Taking advantage of Japan’s abundance of
coastline, these facilities can be located in rural areas near power facilities
to lower power transmission losses. This PHS storage facility was built by
the Electric Power Development Company for the Agency for Natural
Resources and Energy of the Ministry of International Trade as a pilot
facility. The facility’s upper reservoir is located 500 m from the ocean
66
on a 150-m cliff. The facility possesses a 160-m working head, and the
powerhouse is located halfway between the reservoir and the ocean, 150
m below the surface. Because the working fluid is saltwater, significant
corrosion protection was included in the design, such as simplifying
the design of the interior and the inclusion of cathodic protections to
prevent crevice corrosion.
Prospects and challenges

Prospects for this technology are limited in most developed counties
because of high development costs, long lead times, and reservoir
design limitations, such as environmental concerns and large land needs
requiring placement far from load centers and existing transmission
facilities. International prospects for development are better but also are
affected by these issues, plus the requirement for a sufficiently developed
power grid with a sufficient difference in on- and off-peak prices and
demand levels to make development economically viable. Currently, new
PHS facilities are continuing to be developed in countries with rapidly
expanding power sectors, as in China, where 3 GW of capacity are being
added to its existing total of 2.4 GW.
According to Peter Donalek, system studies manager at MWH

Global (a developer of PHS facilities), one of the more important recent
developments for PHS technology is the introduction of adjustable speed
machines for the pump/turbine in both new and retrofit projects. This
advancement provides a number of benefits over previous, fixed-speed
pump/turbines. Operational benefits include both partial speed pumping
(60%–100%) and the ability of the facility to provide frequency regulation
in both pumping and generating modes (previously this was just available
during generation). This flexibility allows the unit to absorb power in a
more cost-effective (and for the power grid, useful) means, and increases
not only the pump-turbine life, but also produces less stress on supporting
equipment such as the seals and bearings. Overall, using adjustable speed
67
machines can improve the efficiency of a pumped storage facility by 3%

or more because the unit can be optimized for peak efficiency in both
pumping and generating modes.
One proposed variant of PHS technology design (primarily from

Japan) uses an underground cavern instead of a free body of water as the
lower reservoir. Because these facilities can be isolated from existing water
resources, there is significantly more freedom in the choice of plant sites,
and a reduced impact on the environment. These facilities are designed
to move the same water up and down in a directly vertical direction,
giving a greater energy-per-unit volume than a natural system—which
must pump the water up at an angle and is limited by geography to
a maximum vertical distance. This also means, however, that the head
must be significantly higher to make up for the reduced volume of
working fluid.
Although many U.S. utilities have expressed a wistful desire for more
of these facilities, the current focus for development of this technology
in countries like the United States is to upgrade existing PHS facilities.
Examples like Ameren UE’s upgrade of the Taum Sauk facility (on the
Black River in Missouri) show that utilities can add more than 100 MW
of capacity for $250 per kW or less. That these facilities continue to be
viewed as premium resources can be attested to by the prices garnered
during auctions. Current upgrades for these facilities normally include
uprating existing facilities with advanced pumps/turbines, impellers,
control systems, and variable-speed drives to increase capacity by 15%
to 20% and operating efficiency by 5% to 10%, and to improve the
unit’s charging rates.4
Major developers
Although no firm is actively developing large PHS facilities in the
United States, firms such as MWH Global and American Hydro are
actively upgrading existing facilities, such as the Taum Sauk plant
68
previously mentioned (American Hydro was retained by Ameren UE to

do the upgrade between 1996 and 1999; original construction was in
1963). Replacing the impellers and other work increased turbine capacity
(from 350 MW to 450 MW), round-trip efficiency (43% to 70%), and
station flexibility at a cost of only $25 million—effectively representing a
new 100-MW power plant for only $250 per kW.
Compressed Air
Energy Storage (CAES)
Summary
Compressed air energy storage (CAES) systems use off-peak power
to pressurize air into an underground reservoir, which is then released
during peak daytime hours to power a gas turbine for power production.
In a gas turbine, roughly 66% of the power produced is required to
compress the air for combustion and high-temperature expansion. The
strategy behind this technology is thus to substitute lower-cost energy
from an off-peak baseload facility for the more expensive natural gas fuel
used to power a separate compressor to precompress and store the air
in an underground chamber. The air is later fed directly into the expander
combustion train (without a compressor) to produce electricity. This
allows a CAES facility (fig. 3–2)—without the parasitic compressor
load—to produce essentially three times the electricity as a gas turbine
from the same amount of natural gas. To improve the efficiency of the
unit, the exhaust gas is passed through a recuperator to preheat the air
coming from the high-pressure storage cavern. CAES and PHS facilities
are the only storage technologies in commercial operation able to
provide large-scale storage deliverability (more than 100 MW) for use in
the wholesale power market.
69
Fig. 3–2. Compressed air energy storage (CAES) (Courtesy of Ridge

Energy Storage)
Historical origins
CAES technology development can trace its roots to the early
1960s during the early evaluation of gas turbine technology for power
production. The technology gained additional support during the 1970s
to provide load-following and peaking-power support because of its
promising fuel efficiency and response capabilities compared to early
natural gas turbines. Because of its load requirements during off-peak
hours, the technology was also considered a means to provide additional
off-peak demand for the growing number of nuclear power facilities
to improve their overall utilization rates. The first unit (which is still in
operation) was developed in Germany in 1978, followed by another unit
in the United States, completed in 1991. A number of other potential
follow-up projects around the world were investigated, but none came to
fruition. However, the continued development of gas turbine technology,
especially when expanded into the combined-cycle concept, crowded
out much of the subsequent interest until recently.
70

A CAES facility consists of five major above-ground and under-
ground components:
1. Power train motor/generator that employs dual clutches

to provide for the alternate engagement of the compressor
and turbine
2. Air compressors, which require low pressure (LP) and high
pressure (HP) casings in two or more stages (intercoolers and
after-coolers) to achieve efficient compression and cool the
air before discharge into the cavern (which also reduces the
moisture content of the air)
3. Expansion or expander turbine train (high- and
low-pressure turbines)
4. Controls for operating the combustion turbine,
compressors, and equipment to handle the switchover from
charging to generation
5. Balance-of-plant auxiliary equipment consisting of fuel supply,
mechanical, electrical, piping, and cooling systems
A recuperator (part of the power train) air-to-air heat exchanger

is used to boost the efficiency of the system by preheating the cavern
air before it enters the first expansion stage. The recuperator heats the
inlet air with exhaust from the turbine to allow the first expansion to
meet the needed inlet temperature and pressure to the combustors.
The facility also contains small inlet air heaters to keep the expander
and recuperator warm for a fast response to full load. The secondary
compression in the cavern as the pressure increases adds temperature,
which is offset by cooling in the wellhead control valve on release of the
air. The reservoir used by the system can be a salt cavern, an abandoned
hard-rock mine, depleted gas fields, or an aquifer—although this last
71
setup would only allow for a low-pressure release of air (550 psi vs.
1,200+ psi for salt cavern or mine) and a subsequent lowering of energy
storage capacity of the unit. Man-made underground storage caverns
can also be used, having either a rubber or steel layer along with a
concrete lining to contain the air.
CAES systems use off-peak power to pressurize air into an

underground reservoir, which is then released during peak daytime
hours to power an air expander/gas turbine for power production.
Normally, roughly 66% of the power produced in a gas turbine is
required to compress the air for combustion and high-temperature
expansion. Instead of using high-cost natural gas for the turbine to
compress the air when the gas turbine is operating, the strategy
behind this technology is to use off-peak power from low-cost
sources (generally baseload coal and nuclear power facilities, but also
including wind power) to run a separate compressor to precompress
and then store the air in an underground chamber. This saves the
increasingly expensive natural gas fuel to only be used when the
compressed air is released from the underground reservoir and fed
directly into the expander combustion train—without the parasitic
compressor load—to produce electricity. Using the air compressed
with off-peak baseload power, a CAES facility can produce roughly
three times the electricity of a normal gas turbine from the same
amount of natural gas, giving the facility a heat rate of 4,250 Btu
per kWh HHV or less during operation. Other operational benefits
accrue to the unit because of the high-pressure air reservoir and the
off-peak power. These units have a far greater response time and
operating range, allowing them to operate with no ambient de-rating
and with shorter ramp rates at times—for instance on a hot summer
day—than a gas turbine. Because of the lower fuel consumption
during power generation of these units, the direct emissions per
MWh are substantially lower than the equivalent capacity from gas
turbines installed in a peak or mid-range combined cycle plant.
72
Because of their large scale and fast reaction capability, CAES

facilities are among the most flexible sources and sinks for bulk energy
in the wholesale power market, making the units able to undertake
frequent startups and shutdowns, swinging quickly from generation
to compression or simultaneous compression and generation for load
optimization. As a source, a CAES facility can provide low-cost energy
on the scale of PHS units with an operational range better than a gas
turbine. Because the air is prepressurized, there is no derating of the
facility because of increasing altitude or summer ambient temperatures
(because of less-dense air). This is a major advantage for a peaking facility
targeted to operate during the peak electricity periods (summer cooling
electrical loads) because for a typical combustion turbine, a 15% derate
penalty can occur for a 40°F rise in ambient temperature. A combined
cycle facility under the same conditions would suffer a 10% derating.
Conversely, CAES facilities do not suffer appreciably from a rising heat rate
under partial load as the competing combined cycles and gas turbines
do. For example, dropping from a 100% load to a 50% load is estimated
to increase a combined cycle’s heat rate up to 25% and a combustion
turbine’s heat rate by up to 30%. With their responsiveness maintained,
CAES facilities are able to ramp three times as fast as gas combined cycle
facilities—which is especially useful on a hot day. However, the real value
of CAES is as a mid-merit facility—able to provide this capability at low
cost over a much larger period of the year instead of only the 1,000 hours
of a peaking gas turbine.
As a sink for power, the decoupled compressors of the CAES

facility in the current concepts proposed can play an important role in
load management as well, either during off-peak periods (to improve
utilization of nearby baseload power facilities by lowering cost and
reducing wear and tear) or during normal operation. Although most
interest lies in the ability of a CAES facility to absorb large amounts
of power during off-peak hours, the ability to reduce or shut off the
compression power is another means to rapidly return power to the
power grid. It can, therefore, provide a means to respond to very large
73
and rapid shifts in the balance of power output for a region—with

additional power output range available from the generator as well. The
fixed-speed compression load of a CAES facility is controlled through
the use of variable inlet guide vanes and pressure control, which allow
the motors to vary their compression load quickly anywhere from 60%
to 100% of full load. With a 100-MW compression load, this permits
the ability to provide 40 MW of regulation or variation in overall
load. (Additional reduction can be achieved through the generating
unit—as is the practice with other power plants today.) By controlling
the position of these guide vanes, the load absorption on the motor
can be controlled, besides controlling the starting and stopping of the
unit. In fact, because the compressors and expander turbine train are
decoupled, each component can operate at any time.
Although it may seem counterintuitive to operate the compressors

while generating power, this gives the facility the flexibility needed to
provide a number of ancillary services for the system, such as frequency
regulation and load following, voltage control, and contingency
reserves. The value of providing these capabilities has always been
embedded within the value of existing power generation facilities, but
until recently—with the development of an actual market for these
capabilities—there has been no price discovery of what the real costs
to provide these services are, and how much the market values them.
As these markets develop, it is apparent that they are indeed important
and valuable, and the faster response capability of a CAES facility gives
these units an edge over other providers for these services, increasing
the overall value of the facility. Because of this wide range of operations,
CAES facilities are also able to benefit wind power. If associated directly
with the wind farm, the entire output of wind energy can be structured
for the market. Even if a CAES unit is simply located in the same region
(because of these facilities’ flexible siting arrangements), the excess off-
peak power from the wind farms can be put to good use while the local
baseload units are allowed to operate more efficiently. As wind energy
penetration continues to climb, balancing this variable resource will
become more important.
74
The average round-trip efficiency for existing CAES facilities ranges

between 75% and 80%; those in development should achieve upward of
85% efficiency. The charge/discharge ratio for each facility can vary greatly
because the charging rate depends on the size of the compressor and
the size and pressure of the reservoir versus the output of the generators.
Finally, the cycle life of these units has proved to be far greater than
similar gas turbines, and the existing facilities are expected to continue
operating for some time to come.5
Cost issues
System capital costs for next generation CAES facilities are estimated
to be approximately $450 per kW according to the EPRI-DOE Handbook
of Energy Storage (from the Electric Power Research Institute [EPRI]–
DOE), with much of the core components benefiting from the overall
advancement in turbine technology. Roughly half of this cost is attributable
to the power plant components, with the BOP and the cost of the storage
reservoir accounting for the remainder of the cost. Construction costs
are greatly reduced if an existing salt dome is used rather than having
to construct a storage facility. Aquifer facilities are expected to cost less;
hard-rock facilities more. Comparatively, modern CAES facilities are
expected to have capital costs roughly equal to, or possibly higher than,
that of a combined cycle facility on a dollar per kW basis, but—more
importantly—will have a much lower cost of energy (produced power)
based on favorable off-peak compression power costs.
Installations
Compressed air energy storage technology has been used for nearly
30 years to support electricity generation. Two plants (discussed in the
following sections) currently operate daily, with another two under
development in the United States alone. Since the Huntorf facility
75
began operation, other proposed facilities have undergone some level

of development:
• 500-MW facility in Texas

• 1,050-MW facility in the former Soviet Union
• 300-MW (3 x 100-MW) facility in Israel
• 100-MW facility in Luxembourg
• 25-MW facility in Sesta, Italy
• 35-MW facility in Japan
Additional interest was developed for CAES facilities in South Africa,

Korea, and Morocco.
Example—Huntorf, Germany.6 The world’s first commercial CAES

facility was the E.ON (German utility) 290-MW Huntorf facility built in
Huntorf, Germany, in 1978 to provide originally a blackstart capability for
some of the nuclear units near the North Sea. The facility (built by BBC
Brown Boveri [now ALSTOM Power]) has successfully run since startup
and has averaged 90% availability and 99% reliability. The underground
salt cavern reservoir had a volume of 10 million cubic feet, holding air
at a pressure up to 1,000 psi. The charging efficiency of the electrically
driven compressors for the cavern has averaged 83%, with the cavern
normally charged over an eight-hour period to generate 290 MW for up
to four hours. With experience, the flexibility of the plant has allowed
for additional market uses to be adopted since startup of the facility,
increasing its value to both owner and the system operator.
Example—McIntosh, AL.7 The second CAES facility is the Alabama

Electric Company’s 110-MW McIntosh facility (power and compressor
train from Dresser Rand) in McIntosh, Alabama, developed in 1991
with design assistance from EPRI. The $65-million facility has found its
most success operating in the fast load changes during the morning and
afternoon peaks. During normal operation, the facility has been started
76
several times a day, ramping quickly to a high load, and after 30 minutes
unloaded and returned to hot-standby. A 50-MW compressor is used to
fill the 19 million cubic foot underground salt cavern and can provide
sufficient air for 26 hours of operation at 100 MW (2,600 MWh). During
operation, the air pressure in the chamber draws down from 1,100 psi
to 650 psi. The unit is designed to come online within 14 minutes but
can start in less than 10 minutes during an emergency (and synchronize
within 5 minutes); the facility can run down to 10 MW. It incorporates
several improvements over the Huntorf facility, including a recuperator
(air-to-air heat exchanger), reducing fuel usage by approximately 25%.

CAES is the only other technology in commercial operation besides
PHS able to provide large-scale storage deliverability (more than 100 MW)
for use in the wholesale power market. As with pumped-hydro storage,
CAES plants have a proven track record, cycling daily and operating
efficiently during partial load operations. Most areas in the United States
and Europe can support CAES facilities because they possess acceptable
geology for the underground storage reservoir. Some areas such as Japan
have no such salt layer (strata or domes), prompting them to develop
hard-rock mines for the air cavity, like the proposed Norton, Ohio facility.
Planned facilities are targeting an availability rate in excess of 95%, with
increases in the recuperator efficiency (85% vs. 75%), faster charging
times, and a faster airflow from the cavern. These enhancements have
the potential to lower the heat rate to near 4,000 Btu per kWh.
Hindering the further deployment of this technology is the perceived

unconventional nature of the technology, although CAES relies heavily
on commercially available turbine technology. Unlike other power fa-
cilities, CAES facilities have significant up-front site development costs
associated with prefeasibility tests and underground excavation. For both
aquifer and salt-dome reservoir development, preliminary detailed stud-
ies of underground conditions are essential and can be very expensive.
77
However, with the expanded use of underground storage facilities for

natural gas storage, these costs and technical requirements are well un-
derstood. Development of the currently proposed plants could provide
additional operational experience, increasing the confidence in the tech-
nology and validation of its economic potential.
Major developers
There are two projects under development for CAES facilities, as
described in the following sections.
Example—Iowa Stored Energy Project. 8 The Iowa Stored

Energy Project (ISEP) is an integrated power facility, consisting of an
84-MW wind farm connected to a 200-MW CAES unit (Dresser-Rand is
providing the power and compressor trains). The $200-million project is
set for start-up during the summer of 2006. The compressed air for the
CAES plant will be stored in an underground aquifer located 1,200 feet
below ground, and it will be able to store the compressed air at 500 psig.
Because of the advantageous geology of the area, both air and natural
gas will be stored in different strata—providing arbitrage opportunities
for gas storage, as well as relieving stress on the local gas pipeline
system during peak demand periods. The ISEP will be operated as an
intermediate load power plant, producing a mix of wind and power from
the CAES facility 12 to 16 hours per day, 5 days per week. During off-peak
periods, the ISEP will compress and then store air into an underground
aquifer using power from the wind farm (and supplemented by off-peak
system power). Overall, the facility is expected to maintain a utilization
rate of 50%, with wind power responsible for 33% of the total output.
Combining the competitive economics of wind turbines and the CAES
facility, the ISEP project is expected to be cost-competitive with other
intermediate power plants in the region. Gradually, with the addition of
more wind turbine generators and compressors, the facility could evolve
into a baseload facility in both scale and capability.
78
Example—Norton Energy Storage.9 The CAES Development

Company is developing the Norton Energy Storage 2,700-MW CAES
facility in Norton, Ohio that will use an abandoned limestone mine as
the reservoir. CAES Development Company, a Haddington Ventures,
LLC subsidiary, plans to develop the facility in stages, with the first
300-MW phase expected to cost approximately $200 million. Additional
construction will continue until the facility is developed to its full capacity
of 2,700 MW. ALSTOM Power and MAN TURBO are potential suppliers
for the power and compressor trains. Compressed air will be stored in
an abandoned limestone mine that reaches 2,200 feet deep and will be
capable of storing 338 million cubic feet at 1,500 psi. There is substantial
local support for the CAES project because alternative prospects for the
abandoned mine included developing it to store toxic waste. With a full
charge to the cavern, the full 2,700-MW facility would be able to operate
for an entire 16-hour period, or allow one unit to operate continuously
for 18 days (43,200 MWh).
Related technologies
Other variations of CAES technology are also being developed,
with one in particular termed thermal and compressed air storage
(TACAS). This is essentially a stand-alone and smaller version of CAES
technology. TACAS stores compressed air in conventional high-pres-
sure cylinders, so it can be more easily located on-site at a commercial
or industrial facility. Unlike CAES, which uses natural gas to heat the
compressed air, TACAS systems store thermal energy in a steel thermal
storage unit (TSU) (fig. 3–3).
79
Compressed
Air Cyllinders Short-Term Energy Storage
Flywheel AC
Motor/
Generator DC
To DC
Load
Thermal Turbine/ DC
Storage Alternator AC
Compressor
Long-Term Energy Storage
Fig. 3–3. Thermal and compressed air storage (Courtesy of Active Power, Inc.).
TACAS was designed to replace existing battery cabinets in a

standard uninterruptible power supply (UPS) system. When low-cost
utility power is available, the TACAS system fills its air tanks with its
compressor and uses electric resistance heaters to bring the TSU to full
operating temperature. After that, TACAS maintains itself in a passive
standby mode. A low-cost flywheel is used to maintain power quality
and to provide a few seconds of bridging power during operation of
the unit. During discharge events, pressurized air is quickly heated in
the TSU in route to an expansion turbine. There, the turbine efficiently
extracts energy from the high-pressure, high-temperature air to
produce power. As this unit is designed to operate indoors, the output
temperature is lower than room temperature. When electric utility
power returns, the TACAS recharges itself.
TACAS is being developed by Active Power under the trade name

of CleanSource XR, in power ranges between 40 and 85 kW. The unit is
designed to provide extended run times of 15 minutes or more, with the
run-time length determined principally by the volume of air stored on-
site. Installed cost and footprint of TACAS are estimated to be comparable
80
to that of flooded batteries for comparable energy storage. Operating

expenses (maintenance and replacement costs), however, would be
significantly less than conventional UPS batteries, leading to a lower life-
cycle cost. As Joe Pinkerton, CEO of Active Power, notes in a personal
correspondence, “We believe that the CleanSource XR is the first true
minute-for-minute alternative to batteries for the UPS industry. It is ideal
for applications where batteries would fail prematurely because of ambient
temperatures or heavy cycling loads.”
Flow Batteries
Summary
Flow batteries store and release energy through a reversible
electrochemical reaction between two electrolytes. There are four types of
flow batteries currently in production or in very late stage of development:
vanadium redox (fig. 3–4), zinc bromine (fig. 3–5), polysulfide bromide
(fig. 3–6), and cerium zinc (fig. 3–7). Flow batteries are typically made
from three subsystems (cell stacks, electrolyte tank system, and control
system) plus the PCS system. The power and energy ratings of the flow
battery are independent of each other. Polysulfide bromide systems are
designed at a system level, requiring specific arrays of cell stacks for the
particular power rating desired and specific storage tank size for the
energy rating desired. Zinc bromine and cerium zinc manufacturers have
settled on modular units with fixed power (cell stack dependent) and
energy (storage tank dependent) capacities, respectively. Because of this
modular design, increasing the capacity of the systems requires discrete
units of power and energy capacity. Vanadium redox manufacturers
have designed a system using both strategies. During operation, the two
electrolytes flow from the separate holding tanks to the cell stack for
the reaction, with ions transferred between the two electrolytes across
81
a membrane; after the reaction, the spent electrolytes are returned to

the holding tanks. During recharging, this process is reversed. As this
technology is highly flexible—both in its physical makeup and its activity
potential—it is able to support a wide variety of applications in markets
including transmission, retail, and renewable energy.
Fig. 3–4. Vanadium redox flow battery (Courtesy of Sumitomo Electric

Industries, Ltd. (SEI)).
Fig. 3–5. Zinc bromine flow battery (Courtesy of ZBB Energy Corp.).
82
Fig. 3–6. Polysulfide bromide flow battery (Ardour Capital Investments).
Fig. 3–7. Cerium zinc flow battery (Courtesy of Plurion systems, Inc.).
Historical origins
Vanadium redox. Early work by NASA on iron-chromium (Fe-Cr)
flow batteries in the 1970s was the inspiration for research into vanadium
redox technology by Professor Maria Skyllas-Kazacos and her team at the
83
University of New South Wales, in Sydney, Australia, beginning in 1984.

Her work focused on vanadium electrolyte stability at high concentration
and related processes and technologies. The intellectual property rights
surrounding this research and the resulting technology were sold to
Pinnacle VRB in 1998. Another line of follow-up work to the NASA
research was begun at the Electrotechnical Laboratory (ETL) in Japan
in the 1980s. Sumitomo Electric Industries (SEI) then acquired this ETL
technology and, in collaboration with the Kansai Electric Power Company,
continued to develop the cell stack design and further integrate complete
vanadium redox systems (with some additional licensed technology from
Pinnacle VRB). Currently SEI is the sole manufacturer of cell stacks, but
VRB Power Systems is actively pursuing its own manufacturing capability.
VRB Power Systems has recently reached an agreement with SEI to
obtain licensing rights to all related intellectual property surrounding
vanadium redox technology. Besides the facilities developed by SEI and
VRB Power Systems, other development efforts include integration
with photovoltaic (PV) arrays (Thai Gypsum Products, later Cellenium
Company), an emergency backup system for submarines (Australian
Department of Defense), and load leveling (Mitsubishi Chemicals, along
with Kashima-Kita Electric Power Corporation).10
Zinc bromine. The zinc bromine flow battery technology represents

some of the earliest research into any flow battery technology, but any
commercial application was difficult due to early poor cycling properties
and chemistry challenges. In the 1980s, two veins of research and
development started with the hope of using the technology for stationary
uses and electric vehicle development. The first area of research began
at Gould, Inc., and was later expanded by Energy Research Corporation
(which became Evercel). The second area of research centered around
Exxon, which developed key components of the technology and licensed
it to a number of manufacturers in the United States, Austria, Japan, and
Australia. By the mid-1990s, the electric vehicle research focus had waned.
Research then focused on stationary power for load-leveling purposes,
with a number of stationary pilot projects in the low multi-MW range,
84
and trailer-mounted mobile systems rated at both 200 kW per 400 kWh
and 1.8 MW per 1.8 MWh. The first commercialization attempt came
through PowerCell Corporation, which in 1998 produced its Powerblock
unit that was rated at 100 kW per 100 kWh, but which was shut down
by 2002. This event left the zinc bromine flow battery field to ZBB Energy
Corporation, a firm that traces its history from Johnson Controls, which
licensed the technology from Exxon. Premium Power Corporation,
building off the technology developed by PowerCell, has recently begun
manufacturing units for sale.11
Polysulfide bromide. The original work on polysulfide bromide

chemistry was developed by Ralph Zito (United States), an independent
inventor who assigned his rights to the technology to Regenesys in the
early 1990s. Besides its involvement in and expansion of the body of
knowledge on the polysulfide bromide chemistry since the early 1990s,
Regenesys Technology had contracted with a number of engineering
firms to develop other components of the system. One of these firms,
Electrosynthesis, was acquired by Regenesys to assist in the development
of the cell stack. After the late 1990s, additional work with Innogy (prior
to its acquisition by RWE AG) was undertaken on the first installation-
site, Didcot Power Station (100 MWh), before the site selection was
changed to Little Barford Power Station facility in the United Kingdom.
Besides this power sector installation, the UK’s Ministry of Defense
had been provided a system to determine the feasibility of including
the flow battery technology in an all-electric ship demonstration. The
demonstration project was designed to be capable of delivering 1 MW of
electricity for periods of 10 minutes under at-sea conditions. VRB Power
Systems has recently purchased the technological base of Regenesys
Technologies to continue the development of this technology type.12
Cerium zinc. Research into cerium zinc flow battery technology

began in the early 1990s at Plurion Systems, with the first practical
application of this technology developed during 2000 and 2001. Pilot
phase demonstration projects are being planned.
85

Flow batteries comprise three subsystems plus the PCS systems.
Since power (MW) and energy (MWh) ratings of flow batteries are
independent of each other in the unit’s design, increasing the power
from the facility necessitates additional cell stacks, whereas expanding
the energy capacity of the system is accomplished by expanding
the amount of electrolyte stored within the unit. The power (MW)
available from each cell is related to the cell voltage (vanadium redox:
approximately 1.2 V; polysulfide bromide: approximately 1.5 V; zinc
bromine: approximately 1.8 V; and cerium zinc: approximately 2.4 V)
and the current density across the membrane. For all flow batteries,
each cell is composed of two half cells separated by a microporous
membrane (vanadium redox—proton exchange; polysulfide bromide—
cation selective); each positive and negative half cell is itself made up of
an electrode and a passage for the electrolyte to flow through. Banks
of these cells can then be linked together to create a bipolar modular
cell stack where the electrodes are shared between the adjacent cells,
with the cathode of the first cell becoming the anode of the next cell,
and so forth. Linked in series, sufficient cells in these bipolar modules
can then form the desired voltage for the cell stack.
The energy storage capacity (MWh) of the facility is determined

by the size of the electrolyte holding tank and the concentration of
electrolyte used. Normally, rubber-lined or plastic tanks are used to store
the electrolyte, with larger systems having flexible geometry to allow for
various sizes and shapes. Accompanying the electrolyte tanks is a system
of pipes, pumps, and control valves to move the electrolyte from the
storage tank to the cell stack and back. Managing all of this is the control
system, which manages the activity of the cell and integrates the cell
stack electrical connections to the PCS.
86
Vanadium redox. Energy in vanadium redox flow batteries is stored

chemically in two ionic species of vanadium suspended in an aqueous
sulfuric acid solution with approximately the same acidity level as that of
a lead-acid car battery. During operation (discharge), the two electrolytes
flow from the separate holding tanks to the cell stack for the reaction,
with ions transferred between the two electrolytes across the proton
exchange membrane. The concentration of each ionic form of the
vanadium electrolyte changes when the flow battery is discharged, when
the potential chemical energy is converted to electrical energy. After
the reaction, the spent electrolytes are returned to the holding tanks,
but because of the changed chemical nature of the electrolytes, self-
separation occurs. During recharging, this process is reversed. Vanadium
redox flow batteries operate at normal temperatures, and there is no
discharge of the electrolyte solutions from the facility during operation.
The electrolytes also have an indefinite life, so they can be used in follow-
on installations after removal of the facility.
Zinc bromine. Energy in a zinc bromine flow battery is stored

chemically in an aqueous solution of zinc and bromine ions that only
differ in their concentration of elemental bromine. During operation
(discharge), the electrolytes flow from the separate holding tanks to
the cell stack for the reaction. The metallic zinc dissolves into the
electrolyte, and zinc ions and bromine ions are allowed to migrate
across the microporous polyolefin membrane to the opposite
electrolyte, equalizing the charge and converting potential chemical
energy to electrical energy. Unlike other flow batteries, the electrodes of
the zinc bromine battery cell serve as substrates for the actual chemical
reactions. This requires that circulation be maintained to free up surface
area for further reaction, and that the battery be fully and regularly
discharged to prevent degradation of performance (stripping). During
recharging, this process is reversed; zinc is deposited on the negative
electrode, and bromine is formed at the positive electrode and remixes
with the electrolyte. After the reaction, the spent electrolytes are
returned to the holding tanks, but because of the changed chemical
87
nature of the electrolyte, self-separation occurs. Zinc bromine flow

batteries can operate in a wide temperature range of 20°C to 50°C.
Heat from the operation of the battery can be removed by the use
of a small chiller. No electrolyte is discharged from the facility during
operation. The electrolytes have an indefinite life, so they can be left in
a fully discharged state for extended periods of time with no problem.
Polysulfide bromide. Energy in polysulfide bromide flow batteries

is stored chemically in sodium bromide and sodium polysulfide
electrolytes. During operation (discharge), the two electrolytes flow from
the separate holding tanks to separate halves of the cell for the reaction,
separated by the cation-selective membrane—sodium bromide on the
positive side and sodium polysulfide on the negative. After the reaction,
the spent electrolytes are returned to the holding tanks, but because of
the changed chemical nature of the electrolytes, self-separation occurs.
During recharging, this process is reversed. Because of the difference in
the electrolytes, any failure of a membrane would cause undesirable
mixing (precipitating sulfur), so safeguards and monitoring systems are
in place to detect a leak. Polysulfide bromide flow batteries operate
ideally between 20°C and 40°C, but they are able to tolerate a wider
range, with the heat generated during operation removed through the
use of a plate cooler. Although there is no discharge of the electrolyte
solutions from the facility during operation, trace quantities of bromine
and hydrogen gases are produced. The electrolytes present minimal
hazards in handling, and from the reaction produce modest amounts
of sodium sulfate crystals. Regular maintenance requirements for the
facility include a biweekly removal of these sodium sulfate crystal
by-products. Other maintenance tasks must be conducted quarterly,
including replacement of spent absorbent, and replenishment and
maintenance of the electrolytic solution.
Cerium zinc. Energy in cerium zinc flow batteries is stored

chemically in zinc and cerium ions suspended in methane sulfonic acid
(MSA); the negative half-cell uses a zinc-MSA electrolyte, and the positive
88
half cell uses a cerium-MSA electrolyte. During operation (discharge),

the two electrolytes flow from the separate holding tanks to the cell
stack for the reaction, with the metallic zinc dissolving into solution, and
the cerium ion moving down one in charge as the potential chemical
energy is converted to electrical energy. After the reaction, the spent
electrolytes are moved to another set of holding tanks to wait recharging
of the system. During recharging, this process is reversed. The zinc ions
in solution form metallic zinc on the electrode, and the cerium ions
move up one in charge. Cerium zinc flow batteries have no discharge
of the electrolyte solutions from the facility during operation. The
electrolytes are environmentally benign and have an indefinite life. As
flow batteries are highly flexible—both in physical makeup and activity
potential—these are able to support a wide variety of applications in
diverse markets, including transmission, retail, and renewable energy.
Units can respond (from standby mode or from charging to discharging)
within 20 milliseconds or less, and can ramp up from a full shutdown
to full operation within 10 minutes or less. Flow batteries also can
sustain elevated output several minutes—some upwards of three times
normal—giving the facility a dual-use capability of long-term energy
and high response rate with instantaneous discharge capability. In the
transmission market, this technology has already proved useful as a
multifunctional facility, providing arbitrage, capacity asset deferment,
and frequency and voltage regulation. In the retail market, the dual-
output capability both acts as an uninterruptible power supply (UPS) to
protect against voltage sags, and as a peak-shaving facility in support of
energy management strategies. Finally, flow batteries can help integrate
renewable energy—both wind and solar—into useful roles in both
power grid and remote power applications. Before RWE had stopped
development of the technology, Innogy had plans for 12-MW to
100-MW polysulfide bromide flow battery facilities to support wind farms
in Denmark in integrating these larger-scale wind energy installations into
the power grid.
89
These wide-ranging applications are supported by the operating

characteristics of the different flow battery technologies:
Vanadium redox. Vanadium redox flow batteries have an efficiency

approaching 85% (with an estimated 75% AC-to-AC round-trip efficiency,
including PCS losses), partially because of the chemical efficiency of the
reaction and the small waiting loss while in the electrolyte holding tanks.
As the same reversible chemical reaction is responsible for charging and
discharging the facility, its charge/discharge ratio capability is 1:1 (although
that may not be the norm for each technology). With a cycle life of at
least 10,000 charges and discharges, vanadium redox flow batteries have
an estimated lifespan of 7 to 15 years, depending on the application and
wear of the cell membranes, pumps, and auxiliary components. Although
some gradual degradation of the proton exchange membrane occurs
(leading to the need for its replacement at the end of its cycle life), no
other degradation of capability is expected over the life of the unit.
Zinc bromine. Zinc bromine flow battery cells have an efficiency

approaching 80% (with an estimated 70% to 75% AC-to-AC round-
trip efficiency, including PCS losses), partially because of the chemical
efficiency of the reaction and the small waiting loss while in the electrolyte
holding tanks. As the same reversible chemical reaction is responsible
for charging and discharging the facility, its charge/discharge ratio is 1:1,
although a slower rate is often used to increase efficiency. With a cycle
life of at least 2,000 charges and discharges, energy storage facilities based
on this technology have an estimated service life of 10 years, depending
on the application and wear of the cell membranes, pumps, and auxiliary
components. Although some gradual degradation of the microporous
membrane occurs (leading to the need for its replacement at the end
of its cycle life), no other degradation of capability is expected over the
life of the unit, leading to it being capable of fully discharging without
damage to the electrolytes or electrodes.
90
Polysulfide bromide. Polysulfide bromide flow battery cells have

an efficiency approaching 75% (with an estimated 65% AC-to-AC round-
trip efficiency, including PCS losses), partially because of the chemical
efficiency of the reaction and the small waiting loss while in the electrolyte
holding tanks. As the same reversible chemical reaction is responsible
for charging and discharging the facility, its charge/discharge ratio is 1:1.
The life of these facilities is estimated to be 15 years—depending on the
application and the wear of the cell membranes, pumps, and auxiliary
components. Although some gradual degradation of the cation selective
membrane occurs (leading to the need for its replacement at the end of
its cycle life), no other degradation of capability is expected over the life
of the unit at the facility, leading to it being capable of fully discharging
without damage to the electrolytes or electrodes.
Cerium zinc. Cerium zinc flow batteries have a DC–DC efficiency

of 70%, partially because of the chemical efficiency of the reaction
and the small waiting loss while in the electrolyte holding tanks. As
the same reversible chemical reaction is responsible for charging and
discharging the facility, its charge/discharge ratio is 1:1. With a high
cycle life, energy storage facilities based on this technology have an
estimated life of 15 years, depending on the application and the wear
of the cell membranes, pumps, and auxiliary components. Although
some gradual degradation of the membrane occurs (leading to the need
for its replacement at the end of its cycle life), no other degradation of
capability is expected over the life of the unit, leading to it being capable
of fully discharging without damage to the electrolytes or electrodes.
Cost issues
System costs vary depending on the application desired. According
to the EPRI-DOE Handbook of Energy Storage, total system costs for a
typical multifunctional application of a vanadium redox amount to
$1,828 per kW, with 80% of this attributable to the storage module.
Also according to the EPRI-DOE Handbook of Energy Storage, total
91
system costs for a typical multifunctional application of a polysulfide

bromide flow battery are $1,094 per kW, with 80% of this attributable
to the storage module. According to this same source, total system
costs for a typical multifunctional application of a zinc bromine flow
battery amount to $639 per kW, with 60% of this attributable to the
storage module. According to Plurion Systems, total system costs for
a typical multifunctional application of a cerium zinc flow battery
range from $750 per kW to $1,000 per kW, with nearly 50% of this
attributable to the storage module. As flow batteries are a recently
commercialized product, additional cost reductions are envisioned,
which will be primarily driven by modularization and advancements in
manufacturing of key components, in addition to scale benefits from
larger production runs, rather than any new product breakthrough in
material science. It is estimated that the storage modules themselves
can be reduced in cost by upward of 25%.13,14
Since the power and energy components of flow batteries are

decoupled, this affects the cost directly as the units are optimized for
one capability over the other. Power-related cost drivers deal with the
cost of the cell stack component, such as the electrodes and membranes,
and are a large part of the cost of the battery module of the system. Any
replacement of these cell stacks is included in the final total ownership
of these systems. Incremental energy cost drivers deal with increasing
the electrolyte volume and its accompanying piping and storage system.
Energy related cost drivers deal with the electrolyte system, including the
cost of the electrolyte, the storage tank and piping required, and any
control systems required. Often described in terms of dollars per kWh,
the electrolyte for the vanadium redox cost is estimated to range from
$30 per kWh to $50 per kWh, with total energy storage costs of $300
per kWh to $1,000 per kWh (depending on system design). The costs
of the electrolytes for the zinc bromine flow battery are estimated to
range from $10 per kWh to $20 per kWh, with total energy storage costs
of $400 per kWh (depending on system design). The electrolyte costs
for the polysulfide bromide flow battery are estimated to range from
92
$10 per kWh to $20 per kWh, with total energy storage costs of $160
per kWh to $185 per kWh (depending on system design). Finally, the
electrolyte costs for the cerium zinc flow battery are estimated to range
from $50 per kWh to $70 per kWh.15
The design of the different flow batteries also affects the installation
cost. Polysulfide bromide energy storage systems are constructed on-site,
leading to somewhat higher EPC (turnkey) costs than purely modular
energy storage systems. Since zinc bromine and cerium zinc are modular
and not constructed on-site, they have a lower EPC (turnkey) cost than
technologies requiring extensive construction on-site. Vanadium redox
energy storage systems are designed to be delivered in modular and
constructed on-site designs. Another driver for installation costs is the
power density of each individual cell. This metric (watts per square meter
[W/m2]) helps to determine the number of cells required for a desired
power output, and, hence, the size of the facility.
Installations—vanadium redox
A number of vanadium flow batteries have been operating in a number
of countries, including Japan, the United States, South Africa, and Italy.
Example—Totorri Sanyo Electric Company, Ltd.16 Voltage sags

caused by lightning strikes on the local power grid can trip the sensitive
equipment in manufacturing facilities for liquid crystal displays—shutting
down the processing line and potentially destroying expensive material
in the process. SEI was asked by the Tottorri Sanyo Electric Company, Ltd.
in Osaka, Japan to develop a UPS solution capable of both protecting
the facility and providing flexibility to the facility in managing its energy
usage. The company installed a vanadium redox–based flow battery
system in April 2001 that provides 3 MW for 1.5 seconds in the UPS role
to combat short-term voltage sags. The facility also operates in a peak-
shaving role to lower energy costs for the facility by providing 1.5 MW
for 1 hour. As opposed to local practices of using distributed generation
93
units as a measure to counter voltage sags, the VRB has no emission—

an important benefit as the local government has been pushing for
nearby companies to explain and reduce their effects from operations
on the local environment. The energy storage facility has operated well,
compensating for 22 voltage sags in the first eight months after its
installation. This first commercial unit of the VRB is still operating.
Example—University of Stellenbosch, Cape Town, South

Africa.17 A number of utilities worldwide have investigated the use of
VRBs. In this example, Eskom, a South African utility, contracted with
VRB Power Systems to develop a 250-kW, 520-kWh vanadium bromide
battery for installation at the University of Stellenbosch near Cape Town
in September 2001. The storage facility installation was a 12-month pilot
project and test bed to investigate the technology’s UPS capabilities
under a number of operational conditions, including power quality, ride-
through, and emergency power backup. Besides Eskom and VRB Power
Systems, Highveld Vanadium and Steel Corporation was involved in the
development of the electrolyte.
Example—PacifiCorp, Moab, UT.18 In the United States, utilities

such as PacifiCorp (a subsidiary of Scottish Power) are looking to use
VRBs to postpone the need for upgrading remote power lines. Several
potential solutions were examined, including conventional planning
options such as line reinforcement, additional power lines, substation
upgrades, and increased reactive compensation; however, all were found
either impractical (due to environmental restrictions) or very expensive.
The use of energy storage as an alternative proved to be the most
economically attractive solution, and resulted in the installation of a
250-kW, 2,000-kWh (eight-hour) vanadium redox flow battery energy
storage system (VRB-ESS) from VRB Power Systems in late 2003 on a
distribution feeder near Moab, Utah. Continuous full power daily cycling
operations began in March 2004. To date, feeder voltage deviations have
improved by 2%, and feeder power factor improvements have reduced
line losses by 40 kW, more than offsetting the parasitic losses of the
94
VRB-ESS. The unit provides a variety of functions, including peak-shaving,

load-following, dynamic frequency control, voltage support, islanded
operations, and premium power for industrial and commercial customers
on the power line.
Installations—zinc bromine
Although no commercial zinc bromine flow battery facility is
currently operating, there have been a number of pilot phase projects
both in the United States and Australia.
Example—Melbourne, Victoria, Australia.19 United Energy Ltd

contracted with ZBB Energy to install a 400-kWh zinc bromine flow
battery system in a substation to supply a shopping area in suburban
Nunawading, Victoria, Australia in November 2001. The facility operates
as a test bed for a variety of operations under various seasonal conditions.
The system also acts in peak-shaving and system stability roles to defer
capacity upgrades on the local distribution system. Financing for the
project was provided by United Energy Ltd and the late Energy Research
Development Corporation (ERDC). The modular arrangement of the
system allows the battery to be isolated from the power grid, which
enables ZBB to demonstrate the facility to other utilities.
Example—Detroit Edison, Detroit, MI. 20 One utility with

significant experience with energy storage technology, Detroit Edison,
installed a 200-kW, 400-kWh zinc bromine flow battery from ZBB Energy
for a two-year set of trials in August 2001. Funding for the program was
provided through the Energy Storage Systems program of the U.S. DOE’s
Office of Power Technologies. The zinc bromine flow batteries were
chosen over lead-acid batteries for their high energy density (two to
three times that of lead-acid batteries), a deeper and far higher cycle life,
a faster reaction time, and a rapid recharge capability. Two sets of trials
over two years were conducted at Detroit Edison substations in both
95
Akron and Lum, Michigan. Each installation tested different applications.

The Akron installation tested peak shaving for deferral of substation
upgrading to alleviate system stress, and the Lum installation provided
voltage support for frequent voltage sags. The system was designed to
support areas with only seasonal daily peak loads; therefore, it was truck
mounted—allowing for quick redeployment of the system to other
locations. This capability is important, as areas of voltage instability
can quickly grow in areas previously not suffering from these problems
because of load growth and usage change by consumers.
Example—Broken Hill, New South Wales. 21 In June 2002,

Australian Inland (AI) Energy contracted with ZBB Energy to install a
500-kW zinc bromine flow battery system at a refurbished 20-kW solar
generating station for a utility substation in White Cliffs, western New
South Wales, Australia. The unit was designed and installed to show the
commercial benefits of load management. To support the installation of
the unit, the Australian Greenhouse Office (AGO) provided 50% of the
total project value of $530,000AU; the balance of funding was provided
equally by AI and ZBB Energy.
Installations—polysulfide bromide
Two demonstration facilities were contemplated for Regenesys
Technology’s polysulfide bromide flow battery; one in the United
Kingdom, and one in the United States.
Example—Little Barford Power Station (UK). 22 The first

commercial scale polysulfide bromide flow battery demonstration
plant from Regenesys was to be located at RWE’s Little Barford Power
Station in Cambridgeshire, UK. Site preparation work started in August
2000, and the 24,000-cell facility was expected to be in operation by
early 2004. The 12-MW, 100-MWh facility was built alongside the
Little Barford 680-MW combined cycle power station. It was designed
96
for blackstart capability (40-MWh reserved) and to demonstrate

the frequency regulation, voltage control, and arbitrage interaction
capability for the surrounding power grid.
Example—Tennessee Valley Authority, Columbus, MS.23 The

second demonstration facility for the polysulfide bromide flow battery
technology was at a Tennessee Valley Authority (TVA) substation in
Columbus, Mississippi, near the Columbus U.S. Air Force Base. The TVA
substation was nearing capacity and would have required $5 million in
upgrades to solve the capacity problem using conventional technologies
and strategies. The 12-MW, 100-MWh flow battery facility from Regenesys
Technologies was originally slated to be online in the first half of 2005.
The entire plant would have required two acres of land, containing a
building 175 feet long, 65 feet wide, and approximately four stories tall, in
addition to the two large electrolyte solution storage tanks. The storage
facility’s primary role was to provide higher reliability for the Columbus
Air Force Base. In addition to supporting the Air Force base, the facility
would have served the western edge of the TVA’s service territory by
providing voltage support, frequency regulation, spinning reserves, and
arbitrage opportunities.
Installations—cerium zinc
There are no current commercial or public pilot phase projects of
the cerium zinc flow battery technology at this time. Plurion Systems
is currently finishing the commercialization of the technology and is
readying several pilot projects for installation at demonstration-sites in
the near future.
97

Because of flexibility and their ease of scalability, many utilities are
very interested in flow batteries for both supporting transmission and
providing enhanced power quality to customers who want premium
power for their operations and a means to improve their load profile
and, thus, lower their costs. Besides PacifiCorp’s installation, both the
Municipal Utility Power Company for Boulder City, Nevada, and REMU—
a Dutch utility—commissioned studies on flow batteries to study their
potential applications for roles such as customer UPS, deferring power
grid additions, frequency regulation, and contingency reserve. As these
systems have an extremely low marginal cost for additional energy
storage, multiple hour capabilities are gaining increasingly competitive
cost comparisons versus alternative storage technologies. As Gary Colello,
CEO of Premium Power Corporation notes in a personal correspondence,
“Environmentally friendly Zinc-Flow energy storage is adaptable to a
variety of target markets and applications including telecommunications,
electric utilities, datacenters, and renewables such as wind and solar. In
addition to meeting a full spectrum of scalable AC and DC output power
requirements, Zinc-Flow energy storage is fully repeatable for thousands
of cycles and provides 10 times the energy storage when compared to
lead-acid batteries. Premium Power solutions are designed for turnkey
outdoor or indoor placement and installation can be accomplished
in under 4 hours. With only minimal required maintenance, Zinc-Flow
provides reliable operation for up to 30 years.” VRB Power Systems has
recently purchased the technological base of Regenesys Technologies
to continue the development of the polysulfide bromide flow battery
technology. The polysulfide bromide flow battery in particular is of
interest to those wanting larger-scale units because of its potential to
scale easily into the multi-MW range.
98
Although continued interest exists, failures of companies such as

PowerCell (zinc bromine) and Regenesys Technologies (polysulfide
bromide) continue to cast doubt over all flow battery technology.
Additional facility market penetration will help both in the manufacturing
of lower-cost cell stacks and on-site facility construction knowledge.
Other possible advancements such as improvement of the concentration
level of the electrolyte would reduce storage requirements (volume for
a given energy level), reducing costs for the auxiliary systems. Each flow
battery technology also many times has its own hurdle. Vanadium redox
flow batteries have the lowest power density and require the largest
number of cells to attain the same rated system power output. Plurion
Systems (the only manufacturer of cerium zinc flow batteries) has yet
to produce a commercial product. Greater commercial success is also
required of all of these technologies to establish a broader track record
for these technologies. As multifunctional facilities, flow batteries must
many times face single-purpose competitors that perform a particular
market role more cheaply. Storage technologies that provide multiple
benefits to multiple beneficiaries rarely are the least expensive for any
one role. What is not more inexpensive, however, is the current (but
often overlooked) requirement to buy all the requisite competing
technologies to match the flexibility of these systems. That message
must be conveyed to potential customers more forcefully for these types
of systems to be successful.
Major developers
Vanadium redox. Two major developers are currently producing
commercial scale vanadium-based flow battery systems: Sumitomo
Electric Industries, Ltd (SEI) and VRB Power Systems. Cellenium Company,
Ltd. has also undertaken some development work toward providing
vanadium flow cell batteries but has not entered into commercial-scale
production as of this writing.
99
Zinc bromine. There are two active developers of the zinc bromine
flow batteries: ZBB Energy Corporation and Premium Power Corporation.
Premium Power purchased the technology of PowerCell and has recently
begun to produce units for sale.
Polysulfide bromide. VRB Power Systems has recently purchased

the technological base of Regenesys to continue the development of
the polysulfide bromide flow battery technology. Previously, Regenesys
Technologies, Ltd., a subsidiary of RWE AG (formerly Innogy, formerly
National Power) was the only developer of this technology.
Cerium zinc. The only developer of the cerium zinc flow battery
technology is Plurion Systems, Inc.
Sodium Sulfur Battery

Summary
The sodium sulfur battery’s (NAS) battery cell is a cylindrical
electrochemical cell with a molten-sodium negative electrode in
the center and a molten-sulfur positive electrode on the outside,
separated from the negative electrode by the β-alumina solid electrolyte
(fig. 3–8). As a cell is discharged, sodium at the negative electrode
discharges electrons, and sodium ions pass through the β-alumina
electrolyte to the positive electrode, where they react with sulfur to form
sodium polysulfide. When a cell is charged, this reaction is reversed; the
sodium polysulfide at the positive electrode decomposes, and sodium
ions return to the positive electrode. Average round-trip efficiency is
89% for the storage module. Prospects for this technology are promising
in the retail market for energy management and power quality. One
significant benefit of this technology is its ability to provide output in
either a long-term or pulse mode. To date, the technology has been well
100
received in Japan, and international evaluation is underway, with one unit

operating in the United States, and additional market development work
in progress in Europe and Southeast Asia.
Fig. 3–8. Sodium sulfur (NAS) battery (Courtesy of NGK Insulators, Ltd.).
Historical origins
The Ford Motor Company conducted early development of the
NAS battery in the 1960s for electric vehicles, closely followed by ABB
in Europe and the New Energy and Industrial Technology Development
Organization in Japan. General Electric (GE) undertook other early
research, this time into stationary applications, during the late 1970s.
Starting in the early 1980s, the Tokyo Electric Power Company (TEPCO)
began a joint program with NGK Insulators to develop the technology
for load leveling at substations and customer sites (initial target, 2-MW,
16-MWh units). With demand for electricity increasing rapidly and a
poor system load factor, TEPCO needed to develop a means to entice
its customers to help improve the operation of the system—and save
101
themselves money at the same time. This need led NGK to establish
a research program with ABB in the late 1980s, which resulted in a
licensing agreement for ABB NAS technology in 1998. NGK’s NAS battery
development program has progressed through three significant stages:
1. β-alumina tube improvements

2. NAS cell and module design
3. Optimization of energy density, performance, and life parameters
By the early 1990s, NGK and TEPCO began installing the first
demonstration programs at customer sites. By 2002, the technology
became commercial in Japan, and the first demonstration unit was
introduced into the United States at an American Electric Power
(AEP) facility.24

The NAS battery cell is a cylindrical electrochemical cell with a molten-
sodium negative electrode in the center and a molten-sulfur positive
electrode on the outside, separated from the negative electrode by the
β-alumina solid electrolyte. Only the positive sodium ions are allowed to
migrate through the electrolyte to react with the sulfur to form sodium
polysulfides. The battery must be maintained at an operating temperature
of 320°C to 340°C to promote this reaction. As a cell is discharged, sodium at
the negative electrode discharges electrons, and sodium ions pass through
the β-alumina electrolyte to the positive electrode, where they react with
sulfur to form sodium polysulfide. When a cell is charged, this reaction
is reversed; the sodium polysulfide at the positive electrode decomposes,
and sodium ions return to the negative electrode. Each NAS cell operates
at approximately 2 V, and cells are configured in series/parallel arrays to
provide the specified voltage and energy storage capability (energy density
three times that of a lead-acid battery). Each cell has no self-discharge or
memory effect common among chemical batteries. All of the individual
102
cells are then housed in a thermally insulated enclosure to form a typical

NAS battery module of 50 kW, 360 kWh or 50 kW, 430 kWh. Through the
cell arrangement, these modules can then be optimized for power quality
(power) or peak shaving (energy). Because of the internal heaters, the
battery module is insensitive to ambient temperatures and can therefore
be easily placed outdoors.
NAS battery systems can provide power either in a single, continuous

discharge or in a larger, but shorter, pulse of power. This flexibility allows
the unit to perform a variety of market applications. Utilities including
TEPCO and AEP have evaluated the system for use at utility substations
to defer transmission upgrades. NAS batteries have also been evaluated
and deployed for use in energy management (peak shaving) and power
quality in the retail energy market. In the peak-shaving role, the unit is
called on to provide long-duration discharges for groups with large loads,
such as office buildings, factories, shopping centers, and schools. Because
of the multi-MW potential size of these units, the incumbent utility
also benefits from an improved load factor among its larger customers.
The short-term pulse capability of the system also enables the unit to
counteract power quality disturbances. This pulse-power capability is
rated at five times the unit’s continuous rating. The ability to inject power
for short periods is useful to mitigate power disturbances such as voltage
sags or momentary outages. For example, the ability to provide power
for 30 seconds is useful because it is a sufficient amount to address more
than 98% of all power quality events and to support a transition to a
backup generator. This pulse-power capability is also available even if the
unit is currently in the middle of a long-term discharge for peak shaving.
With a growing body of operational knowledge with the NAS batteries,
additional market roles are continuing to be evaluated, including using
the unit to stabilize the variable energy output of wind turbines to
improve the integration of wind power into the power grid.
103
These wide-ranging applications are supported by NAS battery

systems because of their strong operating characteristics. NAS batteries
have a high round-trip efficiency and maintain an 89% efficiency rate
during normal operation. With minimal cell degradation, the cycle life
is also far higher than that of other chemical batteries—at 100% depth
of discharge (DOD), the NAS battery has a cycle life of approximately
2,500 cycles. As with other chemical batteries, shallower discharges
provide for a longer cycle life. At 90% DOD, the unit has a 4,500 cycle
life; at 65% DOD, the unit has a 6,500 cycle life; and at a 20% DOD, the
unit has a 40,000 cycle life. Through an expected 15-year operational
life, the degree of degradation for the NAS battery cell is highly related
to the corrosion of the sulfur electrode. Whereas both power quality
and peak-shaving modules charge at 50 kW, each variant discharges at a
different rate, thereby giving a different charge/discharge ratio depending
on design and market role. The peak-shaving NAS battery module can
discharge at up to 100 kW for 15 minutes, in addition to discharging at
lower power levels for longer time periods. Conversely, the power quality
NAS battery module can discharge at up to 250 kW for 30 seconds or
longer, in addition to discharging at lower power levels for longer periods
of time.25
Cost issues
According to the DOE-EPRI Handbook of Energy Storage, total system
costs for a typical multifunctional NAS battery are $810 per kW, with
60% of this attributable to the battery module. As this is a recently
commercialized product, additional cost reductions are envisioned,
which will primarily be driven by manufacturing advancements and
scale benefits from larger production runs, rather than any new product
breakthrough in material science. It is estimated that the battery modules
themselves can be reduced in cost by upward of 33%.26
104
Installations
By early 2004, more than 50 MW of NAS battery systems were
operating at a number of locations, primarily in Japan, but also in
countries like the United States, where evaluation studies are underway.
In total, more than 88 of these systems were installed between 1992
and 2004, totaling more than 66 MW of capacity. Existing units range
from 2,000-kW, 14,400-kWh units for large industrial factories and
semiconductor manufacturing factories to 100-kW, 720-kWh units for
hospitals and amusement facilities for emergency power or peak-shaving
strategies. The largest installation to date is an 8-MW, 64-MWh facility
installed by TEPCO at a water treatment facility in Morigasaki, Japan,
in April 2004.
Example—Akiruno, Japan.27 A Fujitsu semiconductor factory

in Akiruno, Japan, suffered periodic voltage sags sufficient to cause
major processing disruptions (and damage to plant equipment) from
lighting strikes. In fact, during the year prior to finding a solution to the
problem, five major lightning strikes occurred. Through working with
the local utility, TEPCO, the customer chose to install an NAS battery
system from NGK Insulators that began operation July 15, 2002. The
NAS battery system is connected with the power grid of the customer’s
substation at 6.6 kV and has operated nearly every day in peak-shaving
mode, while providing enhanced power quality to the mission-critical
components of the semiconductor factory. For peak shaving, the unit
is capable of providing 1 MW for 7.2 hours, and has been able to be
retasked to alternative operation schedules according to the factory’s
changing demand pattern. For power quality protection, the NAS
battery provides up to 3 MVA for 13.5 seconds, with the transition
time between peak shaving to the pulse power output capable within
20 milliseconds. Since installation, four voltage sags caused by lightning
strikes have caused the NAS battery system to provide power quality for
the semiconductor facility. Although two of these power quality events
were multiple lightning strike events that occurred many seconds apart,
105
the NAS battery system was able to provide sufficient protection during
all events that no mission-critical component was damaged or tripped
because of the voltage sag.
Example—boat racing facility, Japan.28 In another example, an

NAS battery system was chosen to provide higher reliability and lower
cost power for lighting at a public gambling facility in Japan. At the 9,000-
person facility, boat racing is a very popular pastime, and night races, held
100 times per year, are especially popular. The facility’s owner contacted
Kyushu Electric Power Company to find ways to both improve the
reliability of the lighting system and lower the cost of service. The large
lighting system needed to support the races was a significant portion of
the facility’s 4,400-kW peak load and total energy usage of 5,200 MWh.
A 2-MW, 14.4-MWh NAS battery system was chosen, which was owned
and maintained by Kyushu Electric and leased to the owners of the boat
racing facility. Other options such as distributed generation (DG) or
other battery systems were not chosen because the NAS battery facility
required less operation and maintenance, was quieter as it has no moving
parts, and, with no fuel used, had much less environmental impact. After
installation, the peak load of the facility was reduced more than 40% to
2,600 kW. This reduction was a significant factor in the $1,170,000 annual
cost of electric service being reduced 28% to $842,000.
Example—Gahanna, Ohio.29 AEP installed the first NAS battery

system in North America at an office complex in Gahanna, Ohio, in
September 2002. The system is rated for 100 kW, 720 kWh in the peak-
shaving mode (7 hours), with a 30-second 500-kW pulse capability for
power quality applications (ABB provided the system’s power electron-
ics). The DOE and EPRI provided assessment during the two-year pilot
project phase of the facility’s peak shaving and power quality operational
performance to evaluate its economic costs and benefits for a variety of
customer and utility applications. During the evaluation phase, the unit
operated in a peak-shaving mode five days per week, shifting a total of 15
MWh per month. The unit was also subjected to 200 simulated power
quality events, besides experiencing 90 actual power quality events.
106

Prospects for this technology show promise both in existing
markets and new ones. The NAS battery compares especially
f avorably with existing lead-acid batteries . A s the system
encompasses additional capabilities that lead-acid batteries cannot
accommodate, these facilities can take on additional market roles in
harsher environments. Specifically, NAS batteries have:
• Three times the energy density of lead-acid batteries (more

efficient use of space)
• Longer life span because of high cycle ability
• Ease of maintenance (periodic inspection and cleaning) for the
sealed battery modules, as there are no moving parts such as
pumps or heat exchangers to the system
Because of these characteristics and the MW scalability of these

units, they possess strong environmental benefits compared to
on-site generation of peak shaving, as the NAS battery has no emissions
during operation. This aspect is especially important in congested urban
environments where there are strict emission rules.
To date, the technology has been well received in Japan, and

international evaluation is underway with one unit operating in the
United States and additional market development work underway
in Europe and Southeast Asia. As Ted Takayama, Manager of Business
Planning for NGK’s NAS Battery Division, notes, “Utility-scale energy
storage with the NAS battery is a commercial reality in Japan and ready
for introduction to the global market. Through extensive testing and
precommercial demonstration projects with the Tokyo Electric Power
107
Company, high reliability and safety have been confirmed. NGK’s

commercial scale manufacturing facilities are now in operation with
plans for expansion to meet the growing market.”
Barriers to low-cost and large-scale manufacturing have been largely

eliminated as the material of the batteries is inexpensive and abundant.
In addition, with a nod toward the life cycle responsibility of modern
manufacturing in Europe and Japan, dismantling the batteries has been
made easier, as more than 99% of the battery material can be recycled.
Challenges for this battery system include those similar to other energy
management strategies, which rely on arbitrage—the users’ energy cost
must be high enough to warrant the installation of the unit. However, the
pulse power capability of the system for power quality solutions provides
an additional market use of the unit, helping to make it more cost-effective
to customers in lower energy cost areas that need more than simply lower
energy costs to warrant the installation of the unit.
Going forward, this technology will continue to focus on retail en-

ergy management and power quality applications, but as transmission
and distribution upgrades become increasingly difficult everywhere, de-
ferring utility transmission capacity upgrades will become a larger target
for development. To that end, in 2004 the Japanese government began
providing research and development funding for advanced load-leveling
equipment, such as an efficient heat storage facility and the NAS battery
system, to pursue an emerging national load-leveling policy.
108
Major developers
TEPCO and NGK Insulators of Japan are the only developers of NAS
batteries today (commercial production began in the spring of 2002),
with a current manufacturing capacity of 65 MW, with plans for a 200-
MW capacity in a few years.
Lead-Acid Battery
Summary
Research and development of lead-acid (LA) battery technology
has been ongoing for more than 140 years. The two predominant types
of LA batteries are flooded (or vented) and valve regulated (or sealed)
(fig. 3–9). Flooded lead-acid batteries are used in three applications:
starting and ignition, deep cycle, and industrial uses, while VRLA are used
in applications such as industrial tools and backup power. The electrodes
in LA batteries are used both for part of the chemical reaction and for
storing the results of the chemical reactions on their surfaces. Therefore,
both the energy storage capacity and power rating are based on the size
and geometry of the electrodes. Because of their low cost and reliability,
LA batteries remain a favorite for a wide variety of market applications
in the transmission, retail, and renewable energy markets. Although the
possibility exists to use them in a variety of applications, environmental
and operational effects curtail the list of applications truly available to
them. However, LA batteries will remain an important energy storage
technology in a number of existing market applications for the foreseeable
future; they will always be the low-cost option for less-taxing applications
in the UPS, telecommunication, and remote/off-grid renewable markets.
However, prospects for this technology in the expanding role of energy
storage technologies are limited.
Next Page
109
Fig. 3–9. Lead-acid (LA) battery (Courtesy of GNB Industrial Power, a

division of Exide Technologies).
Historical origins
The first practical LA battery was developed by Gaston Plante, a
French scientist, in 1859. By the 1870s, this technology was integrated
into early power-delivery systems to provide load-leveling capability to
meet peak demands. Although rapidly expanding power stations and
transmission systems made the need for prepositioned energy sources
less necessary, the early 20th century saw increased use of LA batteries
to power early electric automobiles. As the internal combustion engine
quickly took over the market for engines, LA batteries found a niche
in the market as the starter for electric motors by the 1920s. Also by
the 1920s, LA batteries had become widespread in the utility market
for standby power systems in power plants and substations. Research
into less expensive and more efficient manufacturing of a more reliable
flooded LA battery created new applications, such as the energy
storage component of a UPS system. By the 1970s, the valve-regulated
Previous Page
110
lead-acid (VRLA) battery was developed, with a lower manufacturing

cost and easier maintenance, but a shorter life. Since then, both flooded
and VRLA batteries have found uses in the stationary market. By 2001, the
total world LA battery market amounted to $15 billion (wholesale), with
the stationary market accounting for approximately 8% of this value.30

Two types dominate LA battery design: flooded (or vented), and
valve-regulated lead-acid battery—VRLA (or sealed). Flooded LA
batteries are used in three areas:
1. Starting and ignition, where they provide a short burst of
strong power to start an engine
2. Deep cycle, where they provide a low, but steady level of
power for a longer period of time than a starting battery
3. Industrial, where they provide low, steady power over a much
longer period of time than a typical deep cycle battery
A flooded LA battery consists of a positive electrode comprised

of lead dioxide and a negative electrode comprised of lead; to improve
the performance of the battery, pure lead is not used, but rather an
alloy, such as lead-antimony. The electrodes are then immersed in
liquid electrolyte, a dilute solution of sulfuric acid and water (35%
acid/65% water).
VRLAs are used in a number of applications, including industrial

tools and backup power. Unlike flooded LA batteries, these batteries are
sealed to prevent water loss or release of gas during charging. Although
designed to be a longer-life replacement for flooded LA batteries in
applications such as UPS installations, VRLA batteries are actually far
more susceptible to variations in temperature and over/undercharging
that affect the life of the battery—giving them a lower operating life
than flooded LA batteries. Each VRLA cell is constructed from positive
and negative electrodes that are separated by a porous material, although
111
the electrolyte is immobilized into an absorbent separator or gel. VRLA

batteries also contain a resealable vent to allow off-gassing if the battery is
overcharged; this limits the rate at which VRLA batteries can be charged.
For both designs, the cells in a battery are connected in parallel series
to obtain the required current and voltage for the application. As each
battery has slightly different operating characteristics, monitoring systems
are needed to both balance the current flow and signal if a battery
is failing or fails (because of an internal fault) to prevent neighboring
batteries from damage.
The electrodes in LA batteries are used for part of the chemical

reaction and for storing the results of the chemical reactions on their
surfaces. Therefore, both the energy storage capacity and the power rating
are based on the size and geometry of the electrodes. A higher power rating
requires a larger surface area for each electrode, often leading to more and
thinner plates in a battery. However, the energy storage capability is based
on the mass of the plate, leading to fewer and thicker plates (in a given
sized battery). Typically, LA batteries have a cell voltage of 2 V, but as the
battery discharges, the voltage drops slowly. During discharge, the positive
electrode (lead dioxide) frees lead ions into the electrolyte solution, while
lead sulfate forms on the negative electrode (lead), as the sulfuric acid turns
into water. Reversing the electrical charge through the system recharges
the battery. When the cell is being recharged, the chemical reactions are
reversed, restoring the battery to its original condition. Because of off-
gassing, flooded LA batteries require far more monitoring and attention
(adding water) than VRLA batteries. In addition, the emission of acid
fumes from flooded LA batteries can cause corrosion of surrounding metal
supports unless vented to the outside.31
Because of their reliability and low cost, LA batteries remain a

favorite for a wide variety of market applications in the transmission,
retail, and renewable energy markets. These batteries can respond within
milliseconds and provide full power instantaneously. In the transmission
market, LA batteries have already proved useful as a multifunctional
facility, providing capacity asset deferment along with frequency and
112
voltage regulation. However, these demonstration projects showed

that larger MW-scale facilities can have operational difficulty, although
smaller-scale units have proved useful in niche locations. In the retail
market, the ability to produce double the normal output levels for short
periods allows these units to both provide commercial and industrial
facilities power quality protection against voltage sags and be a ride-
through resource. Some groups are also investigating combining LA
batteries with shorter-duration storage technologies to provide a longer
emergency power capability. Another important use of LA batteries
is to provide standby power for substations, power plants, and the
telecommunication industry. However, peak shaving for commercial
and industrial facilities would shorten the battery life because of the
higher cyclic requirements of that application, so integrating them into
an energy management strategy is generally avoided. Finally, LA batteries
are used extensively in remote renewable energy installations—especially
solar—to increase the value of the power by providing dispatchability.
Although the possibility exists to use LA batteries in a variety of ap-

plications, environmental and operational effects curtail the applications
truly available to them. The average DC–DC round-trip efficiency of an LA
battery is 75% to 85% during normal operation. Generally, an LA battery has
a useful life of approximately five years under normal operating conditions,
which corresponds to a cycle limit of 250–1,000 charge/discharge cycles,
with VRLA batteries having a much lower cycle life than vented LA batter-
ies. The cycle life of an LA battery can be significantly degraded because of
temperature, depth of discharge (DOD), fast charge/discharge cycling, and
other factors. Temperature may be one of the most important aspects af-
fecting the cycle life of an LA battery. The primary operating temperature of
an LA battery is roughly 80°F, but operating the battery 40 or more degrees
above this point can cut the life of the battery by 50%. (Many UPS systems
have minimal space-conditioning, leading to early failure.) The DOD also
affects the cycle life of an LA battery—with the deeper average discharge
corresponding to a shorter life. Normally, LA batteries for UPS and other
energy storage systems are designed for steady, deep cycle discharges of 50%
113
to 80%; above this, the effect becomes more pronounced. Finally, LA batter-
ies take far longer to charge than discharge, having an effective charge-to-
discharge ratio of 5:1 or more to prevent damage to the cell. Faster charging,
although possible, will lower the life of the battery; if charged at more than
2.4 V, gassing (electrolysis) occurs (this also occurs during overcharging),
producing oxygen and hydrogen and ultimately damaging the cells.32
Cost issues
Generally, LA batteries are the cheapest energy storage technology
choice for most retail and some transmission applications. According to
the EPRI-DOE Handbook of Energy Storage, a multi-MW, multifunctional
LA battery system has a system cost of $580 per kW, with a little more than
50% from the battery module. As these are mature technologies, there is
no real expectation for significant cost reductions over the near term from
technology or manufacturing advancements.33
However, cost comparisons should incorporate the real-world

operating requirements and conditions that often shorten the expected
life of LA battery units—driving up their actual total ownership cost
(in particular, high cycle, deep discharge conditions), even to the point
where they become more expensive than alternative energy storage
technologies. Design issues of matching capabilities to applications
are also important because the power and energy components of this
technology are linked as each cell has a specific aspect of each. Therefore,
the scale (and hence cost) of an LA battery system relies to a great
extent on the required number of battery strings. Each string requires
an accompanying monitoring system to prevent failure of one battery
from damaging other batteries within the string. Smaller groups of LA
batteries for UPS systems can be modular, but larger, multifunctional
facilities require some construction on-site for the housing, monitoring,
and space conditioning system. Although the space conditioning system
is looked on for cost savings, failure to provide an environment that does
not stress the batteries will greatly reduce their useful life.
114
Installations
Besides numerous UPS installations, LA batteries were the heart of
many of the early large-scale multifunctional energy storage facilities,
such as the 1986 BEWAG 8.5-MW, 8.5-MWh battery facility located in
Berlin, Germany.
Example—Vernon, CA.34 An example of a large LA battery UPS

system is the 3-MW, 4.5-MWh VRLA UPS system that was installed at an
LA battery recycling plant in Vernon, California. This facility was designed
to provide peak demand reduction and uninterruptible power. For a
number of years after its installation in 1996, the system greatly reduced
the power quality events that had previously resulted in increased lead
emissions and costly noncompliance fines.
Example—Chino, CA. 35 Southern California Edison (now a

subsidiary of Edison International) installed a 10-MW, 40-MWh vented LA
battery system during 1988 at a substation in Chino, California. The system
was installed as a pilot facility to evaluate battery systems. Its primary role
was to evaluate load-leveling operations of a battery energy storage facility.
Secondarily, the facility evaluated load following, transmission asset deferral,
economic dispatch, frequency regulation, voltage control, and blackstart
operation capability. The facility operated successfully through 1997.
Example—Sabana Llana, Puerto Rico.36 An early example was

the Puerto Rico Electric Power Authority’s 20-MW, 14-MWh (40 minute)
LA battery storage plant at its Sabana Llana substation, which was
installed in 1994. This $21-million facility successfully provided spinning
reserves, regulation, and voltage control services for the island during
normal operation. Because of the small size of the island’s power grid,
these capabilities were always in short supply. These capabilities proved
extremely useful during the fall of 1998 in the aftermath of Hurricane
George, when unreliable power supplies and transmission reliability
caused a number of load-shedding events and increased spinning reserve
requirements—requirements that the battery facility reduced during this
115
time of scant resources. Although the plant operated successfully for a

number of years, deteriorating batteries caused the premature closure of
the facility in 1999.
Example—Metlakatla, AK.37 In 1997, Metlakatla Power & Light in

Metlakatla, Alaska—a small village on an island in the southeast portion
of Alaska—installed a 1.2-MW, 1.4-MWh valve-regulated LA battery
system from GNB Industrial Power/Exide. The facility was designed to
have a 20-year life span. This Metlakatla facility is a well-known example of
how a battery energy storage facility can help integrate renewable energy
resources into a power grid, and how a storage facility can become an
integral part of an island grid’s balancing system. Here, the full cost of
the $1.5-million energy storage facility was recovered within three years
because the storage unit improved overall power quality and reduced
the operating costs of the existing generating assets. There were both
direct financial and nonfinancial benefits to using the system. The system
matched generation needs very well, almost totally replacing a peaking
diesel unit that was costing $400,000 per year in fuel costs alone. In
addition, the energy storage facility greatly reduced the environmental
risks to the community, as diesel emissions and noise were greatly reduced
(or eliminated).

LA batteries will remain an important energy storage technology in
a number of existing market applications for the foreseeable future; they
will always be the low-cost option for less taxing applications in the UPS,
telecommunication, and remote and off-grid renewable markets. However,
prospects for this technology’s usefulness in the expanding role of energy
storage technologies are limited because the requirements of many of
these applications would significantly curtail the life of the unit. Although
mature, further development will continue (primarily through using
different alloys with lead for the electrodes) to reduce maintenance and
extend life through reducing corrosion, plate buckling, and off-gassing.
116
Unfortunately, a number of challenges inherent in the technology

will continue to restrain further deployment of LA batteries for
stationary applications in the electric utility market. Recent advances
in the basic technology continue to fall short of the breakthrough
desired. This lack of technological innovations restrains significant
additional research and growth for this technology, and encourages
research into competing energy storage technologies. For these reasons,
the roles currently assigned to LA batteries will be the only practical
ones available; large-scale, multi-MW systems are impractical because
of short battery life and the extensive monitoring and environmental
controls required. Although the markets dominated by LA batteries
will continue to grow, a number of other energy storage technologies
will continue to specifically target these established LA battery markets
for growth. As environmental concerns about lead continue to grow,
standards and regulations concerning the manufacturing, handling, and
disposal of these batteries will continue to increase.38
Major developers
There are a number of manufacturers of LA batteries for a variety of
applications. Some of the larger manufacturers include C&D Technologies,
GNB Industrial Power/Exide, East Penn Manufacturing Company, Gill
Batteries (a division of Teledyne Continental Motors), Optima Batteries
(a Johnson Controls, Inc, company), Trojan Battery Company, and Crown
Battery Manufacturing Company.
117
Nickel-Cadmium
Battery
Summary
Swedish scientist Waldmar Jungner developed the first nickel-
cadmium (NiCd) battery (the pocket-plate style) in 1899. Three designs
dominate nickel-cadmium battery design: pocket plated, sintered
plated, and sealed. A typical design can be seen in figure 3–10. Pocket-
plated NiCd batteries are used extensively in markets such as industrial
and standby power, where ruggedness and durability are important.
Sintered NiCd batteries dominate in markets such as for starting
aircraft and diesel engines, where high energy per weight and volume
are important. Finally, sealed NiCd batteries are used commonly in
commercial electronic products, where lightweight, portable, and
rechargeable power is important. The reliability of NiCd batteries
makes them a favorite for a wide variety of market applications in the
transmission, retail, and renewable energy markets. For many reserve
power applications, NiCd batteries are one of the least expensive energy
storage technology choices for most applications in the retail and some
transmission applications. Nickel-cadmium batteries will remain an
important energy storage technology in a number of existing market
applications for the foreseeable future. Although slightly more expensive
than LA batteries, their operating capabilities and excellent reliability
span a wider operations envelope and longer cycle life, allowing for
lower ownership costs. Unfortunately, a number of challenges continue
to restrain significant further deployment of NiCd batteries for stationary
applications in the electric utility market.
118
Fig. 3–10. Nickel cadmium (NiCd) battery (Courtesy of Saft).
Historical origins
Waldmar Jungner’s pocket-plate NiCd battery had few applications
because of the high cost and difficulty in manufacture. For this reason,
much development activity for nickel-based batteries did not occur until
after Thomas Edison designed the nickel-iron battery. In 1932, Shlecht and
Ackermann invented the sintered plate design, which was thinner with
a higher surface area, allowing much higher load currents and improved
longevity. By 1946, Neumann developed the sealed nickel cadmium cells,
with early applications in the military market. Development toward
improving manufacturing and design continued, focusing on improving
capabilities and lowering the cost of the unit. Pocket-plate designs still
dominate for ruggedness and durability in market applications, whereas
sintered-plated designs dominate where higher energy per weight and
volume are important. Nickel-cadmium batteries are found throughout
the rail industry (40% of all NiCd batteries are destined for this market),
military uses, space applications, and standby power in the industrial and
119
electric power industry. By the 1970s, development of the technology

had continued so that NiCd batteries became widely used in the portable
power market for consumer electronics.39

The three dominant NiCd battery designs are pocket plated, sintered
plated, and sealed. Pocket-plated NiCd batteries are used extensively in
markets such as industrial and standby power, where ruggedness and
durability are important. Sintered-plated NiCd batteries dominate in
markets such as aircraft and diesel engine starters, where higher energy
per weight and volume are important. Sealed NiCd batteries are used
commonly in commercial electronics products, where lightweight,
portable, and rechargeable power is important.
A NiCd battery cell consists of a positive electrode comprised of

nickel hydroxide and a negative electrode comprised of metallic cadmium
commonly divided by a nylon separator. The electrolyte is aqueous
potassium hydroxide and is not significantly changed during the reaction,
with the cell voltage of the typical NiCd battery being 1.2 V. During
discharge, nickel oxyhydroxide combines with water to produce nickel
hydroxide and a hydroxide ion, and cadmium hydroxide is produced at
the negative electrode. Reversing the electrical charge through the system
recharges the battery. When the cell is being recharged, the chemical
reactions are reversed, restoring the battery to its original condition. During
charging, some oxygen is produced at the positive electrode and some
hydrogen at the negative electrode, requiring some venting and occasional
water addition, but far less than that required of an LA battery.40
Because of the reliability of NiCd batteries, they remain a favorite

for a wide variety of market applications in the transmission, retail, and
renewable energy markets. These batteries can respond within milliseconds
and provide full power instantaneously. In the transmission market,
NiCd batteries have recently been installed in the largest multifunctional
battery energy storage facility—the 40-MW Golden Valley Electricity
120
Association facility—to provide much-needed frequency and voltage

regulation on that utility’s system. As with LA batteries, however, only
experience will prove if this technology is capable of wider deployment
in MW-scale facilities for this market.
In the retail market, NiCd batteries have proved popular with

commercial and industrial facilities that need power quality protection
against voltage sags and a ride-through resource to on-site generation for
firms looking at total ownership costs instead of simple up-front costs.
These batteries are also finding a growing market as a replacement for
LA batteries to provide standby power in harsh environments such as
power plants. Although possessing a better cycle life than LA batteries,
peak shaving for commercial and industrial facilities would shorten
the NiCd battery life because of the higher cyclic requirements of that
application, so integrating them into an energy management strategy
is generally avoided. Nickel cadmium batteries are also finding many
new applications for remote renewable energy installations—especially
solar—where reliability and tolerance to heat is required (something
common where solar energy intensity is strong).
These wide-ranging applications are supported by nickel cadmium

batteries’ useful operating capabilities. Unfortunately, the average DC–DC
round-trip efficiency of a NiCd battery is 60% to 70% during normal
operation—somewhat lower than LA batteries. However, the expected
life of these batteries is much higher—rated at 10 to 15 years, depending
on the application—and depends on the design. Pocket-plate batteries
have a useful life of 1,000 charge/discharge cycles, and sintered batteries
are capable of sustaining 3,500 cycles. As with all batteries, the cycle
life of a NiCd battery can be degraded from a number of issues, such
as depth of discharge (DOD), temperature, and fast charge/discharge
cycling. Unlike LA batteries, NiCd batteries have a far greater cycle life
expectancy at small DOD—at only 10% DOD, life cycle expectations
reach 50,000 cycles. Nickel cadmium batteries are also more tolerant of
higher temperature operations, able to withstand occasional operations
121
at 120°F, making them far more tolerant for nonconditioned spaces for
standby power. As with LA batteries, NiCd batteries can be charged at
varying rates with accompanying effects on the battery and the ability to
transfer the full energy storage capacity of the battery. Nickel cadmium
batteries require a somewhat greater amount of float charge as they lose
between 2% to 5% of their charge per month at room temperature (LA
batteries lose 1%). Most of the discharge occurs just after charging, and
self-discharge increases with temperature.41
Cost issues
For many reserve power applications, nickel cadmium batteries are
one of the least expensive energy storage technology choices for most
applications in the retail and some transmission applications. According
to the EPRI-DOE Handbook of Energy Storage, a multi-MW, multifunctional
nickel cadmium battery system has a system cost of $600 per kW, with a little
over 60% from the battery module. As with other mature energy storage
technologies, there is no real expectation for significant cost reductions
over the near term from technology or manufacturing advancements.42
Although NiCd batteries have a slightly higher initial capital cost

compared to LA batteries, their superior operational properties actually
provide lower cost of ownership. Possessing a higher cycle life and a
lower maintenance requirement, these batteries have proved to be a
viable substitute for LA batteries for many applications. Similar battery
monitoring is necessary, but it can be done remotely, and maintenance
requirements are significantly lower, requiring only annual inspection and
care even for actively cycled batteries. As nickel cadmium batteries are
also more tolerant to higher temperatures, environmental management
and control common for LA or other battery systems will not add any
relative additional costs to the installation, and may even reduce or
eliminate them in certain circumstances. Finally, cadmium is a toxic
material, requiring close control throughout production, use, and
recycling of the material.
122
Installations
Nickel-cadmium batteries have been used in a number of applications
in the electric power industry as a replacement for LA batteries.
Example—Golden Valley Electricity Association, AK. 43 The

Golden Valley Electricity Association (GVEA) of Fairbanks, Alaska,
had a problem. GVEA’s transmission system is radial in nature, leading
to power quality difficulties for customers when there are problems
because of swings in the load, environmental forces, or a problem with
the intertie (inter-tie) or power facility in Anchorage. After evaluating
a number of options, GVEA installed in August 2003 the first four
strings of a $35-million, six-string battery energy storage system using
Saft’s NiCd battery, which provides 40-MW, 10-MWh (15 minutes)
(power electronics from ABB). When completed, this facility will be
the largest battery system ever installed. The unit is designed to help
improve the reliability of service to GVEA members and operates in
several distinct modes, all of which help GVEA reduce its fuel costs.
First, the unit provides frequency regulation by repetitively cycling
from charging to discharging. This is a fundamentally different tactic
from that of balancing a mismatched system load from existing
generation facilities, which must constantly change their output
level to provide this balancing role. Second, the unit reduces the
high spinning reserve requirements common among Alaskan utilities
because of their independent and divided nature. Finally, the unit is
designed to cover the 15-minute period between loss of generation
and the start-up of backup generation. To date, the facility has been a
success; as of June 2004, 26 events totaling 336 minutes of activity were
already encountered.
Example—Eskom, South Africa.44 Eskom, the national power

utility for South Africa, was having trouble with the backup LA batteries at
its Lethabo power station for one of its generator units. These LA batteries
provided backup power for a number of critical systems, which supported
the local control systems and ensured a safe shutdown of the unit in case
123
of the unit tripping off-line. These batteries were designed to have a lifetime
of 20 years, but they were proving to only have a life span of 12 years. The
cause for this was traced to the elevated temperature, which often reached
95°F for extended periods of time. Rather than simply replacing the
batteries, Eskom decided to consider alternative battery technology and
investigated a number of alternatives through a life-cycle cost evaluation.
Through this analysis, Eskom decided to switch to Alcad NiCd rechargeable
batteries for these critical control and protection systems. In addition to
their ability to withstand the demanding temperature, the NiCd batteries
also provided a number of other benefits over LA batteries. One of the
biggest problems with the LA batteries was their unpredictable lifetime.
They tend to have a nonlinear life and die suddenly; when their internal
components corrode, their capabilities quickly deteriorate (hence, the
need for constant monitoring). NiCd batteries, however, fail linearly and
predictably; they have a much longer lifespan and require only infrequent
maintenance. Since the NiCd batteries were installed in September 2000,
they have proved successful, and plans are underway to replace the control
backup systems for the other five units—and the backup power system for
the entire plant’s control system—with NiCd batteries.

Nickel-cadmium batteries will remain an important energy storage
technology in a number of existing market applications for the foreseeable
future. Although slightly more expensive than LA batteries, their operating
capabilities and excellent reliability span a wider operations envelope
and longer cycle life, allowing for lower ownership costs. In particular,
they suffer no sudden-death events common among older LA batteries.
Because of their low maintenance requirements, they continue to be a
favorite for longer-term deployment in remote telecommunication and
renewable energy applications. However, prospects for this technology
to be used in the expanding role of energy storage technologies are
limited because the requirements of many of these applications would
significantly curtail the life of the unit.
124
Unfortunately, a number of challenges will continue to restrain

significant further deployment of NiCd batteries for stationary
applications in the electric utility market. For many standby power needs,
customers still place significant emphasis on up-front costs rather than
total ownership—although this is slowly changing. As these markets
(UPS and remote power needs) continue to grow, NiCd batteries must
continue to capture additional market share, although they will always
remain more expensive than LA batteries for energy storage (but better
for power delivery). Because of their improved operating capabilities,
larger-scale NiCd-based battery energy storage facilities are somewhat
more practical than LA systems, but not enough for widespread use in
the wholesale power market. Although some advancement continues,
some recent advances in the manufacturing process fall short of the
breakthrough desired. As a mature technology, further development
will continue, but much additional research into nickel-based battery
technology to reduce manufacturing cost will continue to be directed
toward other nickel-based battery technologies. As environmental
concerns about toxic metals continue to grow, standards and regulations
concerning the manufacturing, handling, and disposal of these batteries
will continue to rise.
Major developers
There are a number of manufacturers of NiCd batteries for industrial
and electric power market applications. Some of the larger manufacturers
of NiCd battery systems include Alcad Limited, Hoppecke Batterien
GmbH, Saft, and Tudor Batteries.
125
Flywheels
Summary
Flywheels have been an essential tool for leveling power flow in
and out of a spinning mechanical device for a very long time. Flywheels
store energy through accelerating a rotor up to a very high rate of speed
and maintaining the energy in the system as kinetic energy (fig. 3–11).
For this technology, there are two main veins of development, low- and
high-speed rotors. Low-speed steel rotor flywheels predominate and are
used primarily in uninterruptible power supply (UPS) devices. In high-
speed flywheels, advanced composite materials are used for the rotor to
lower its weight while allowing for the extremely high speeds. Although
they could be used for UPS applications, the increased energy storage
capability allows for other market applications, such as regenerative energy
storage. For both types, energy is stored in the rotor in proportion to its
momentum, but at the square of its surface speed—hence the desire
to develop high-speed flywheels with correspondingly higher energy
densities to allow for other market applications. The flywheel releases
its energy by reversing the charging process and using the motor as a
generator; as the flywheel releases its stored energy, the flywheel’s rotor
slows until it is fully discharged. Flywheels are suitable for applications
requiring frequent and deep discharges that generally prove too taxing
for standard battery installations and are also more compact and require
less maintenance. The development of flywheel systems with far higher
cycling capabilities supports a number of emerging applications such as
regenerative energy applications.
126
Fig. 3–11. Flywheel (Courtesy of Pentadyne Power Corp.).
Historical origins
Flywheels have been used for thousands of years to both store energy
and smooth out the variable-speed operation of rotating machines.
Early hand-powered strategies such as a potter’s wheel were expanded
on during the industrial revolution to balance the variable power from
such pulse sources as early steam engines. The flywheel, attached to the
rotating shaft, moderates fluctuations in the shaft’s speed by temporarily
storing excess energy to be used during the nonpowered stroke of the
engine; this more-manageable power source then, in turn, allowed for
the development of far more complex powered mechanical drives. The
subsequent use of steam and combustion engines to produce electricity
introduced mechanically based flywheels in the power industry, with
many still incorporated in small generators today. Development of
flywheels as a stand-alone energy storage unit for electrical power was
given support in the 1970s, when advances in power electronics allowed
for the efficient voltage and frequency control of the power output
regardless of the rotational rate of the flywheel. These electronics also
127
allowed for a higher utilization of the flywheel’s momentum, which,

coupled with subsequent advances in material science (carbon fiber,
magnetic bearings), allowed for greater energy densities and the potential
for usable energy resources over a wider set of applications.

The basic components of a flywheel energy system include a rotor,
motor/generator, bearing system, vacuum housing, and power electronics
(conversion and system control). The rotor is the most important aspect
of the system, because its design dictates the amount of energy able to be
stored. Flywheels store power in direct relation to the mass of the rotor,
but to the square of its rotational surface speed. Therefore, the best way
to increase the amount of energy in a flywheel is to make it spin faster,
not by making it heavier. However, it is important to understand that it
is the surface speed that is important—not simply RPM—so a larger-
diameter flywheel can have the same energy level as a smaller one that
rotates much faster. Flywheels store energy through accelerating the rotor
up to its operating speed and maintaining its rotational speed (and level
of energy) with a small but constant additional energy input. When power
is needed, the flywheel reverses this process and discharges its stored
energy by using the motor as a generator; as the flywheel releases its stored
energy, the flywheel’s rotor slows. Because energy is stored mechanically
and not chemically, it is possible to deeply discharge the unit repeatedly
without any damage to the unit. To maintain the energy in the system,
anything that provides resistance to the spinning rotor is minimized. Early
bearing systems were made from roller bearings, but many designs today
use either passive or active magnetic bearings to reduce frictional losses.
To further reduce resistance, most high-energy designs also maintain the
spinning rotor in a vacuum housing. Although necessary to increase the
surface speed of the rotor, these last two design changes unfortunately
also minimize the transfer of heat out of the unit (a constant concern to
extend the life of the unit), so modern systems frequently have some type
of cooling or chiller system included as well.
128
Because the power and energy components are decoupled in

flywheels, these systems can be loosely classified into two categories,
optimized for either power or energy. Whereas optimizing for power
requires a greater emphasis on the motor/generator and power
electronics, optimizing for higher energy densities requires a larger, high-
speed rotor. Depending on the material used and design speed, the
rotor’s diameter can vary greatly. Low-speed flywheel systems generally
have a heavy, solid steel rotor and rotate with speeds below 10,000 RPM.
Having less energy, they are geared toward shorter bursts of power
in such applications as power quality in a UPS. High-speed flywheel
systems spin a lighter rotor at much higher speeds, potentially up to
100,000 RPM, but generally in the 20,000 to 60,000 RPM range. Because
of the increased stresses at these speeds, high-speed flywheel rotors are
normally constructed from composite materials, such as fiberglass or
carbon fibers, impregnated in an epoxy and wound into a thick cylinder.
Although high-speed flywheels can also be used for UPS applications,
the increased energy storage capability allows for other, emerging market
applications.
Flywheels used for energy storage have three basic applications:

power quality, regenerative energy, and frequency control. First, flywheels
have provided power quality enhancements as a substitute for batteries
in UPS systems for many years now. Besides helping to ensure high
quality, usable power service, they are able to provide a vital ride-through
energy source until a backup generator can be brought online. Most
notably, flywheel-based systems have proved to be far more capable
in harsh environments where frequent and sometimes deep discharges
would shorten the useable life of the common LA battery installation by
50% or more. Second, flywheels are capable of capturing wasted energy
in repetitive motion systems by converting it to kinetic energy in the
rotating mass of the flywheel rotor. Here, they are becoming increasingly
popular in industrial crane or light-rail systems that undertake repetitive
motions and frequently need short bursts or pulses of power. This not
only protects against excessive wear and tear on the existing equipment,
129
but also improves the economics of the application. Finally, the repetitive
cycling capability of a flywheel is ideal to dampen frequency variation
on power systems caused by moment-to-moment imbalances in power
supply and demand. Here, support can be provided with a relatively
small energy storage capability (as compared to the power grid), because
the repeated charging and discharging can be accomplished quickly
and continuously. Frequency regulation is of concern to all levels of the
power grid but is of special concern on smaller power grids or isolated,
self-generating industrial facilities. The variability of wind turbines is
becoming another area where high-cycle storage facilities can be used to
smooth the output of wind turbines.
As improved power electronics, vacuum housings, and magnetic

bearings have become more widespread, round-trip efficiencies of
flywheel systems have improved; many current production models are
in the 70% to 80% range, with some newer designs even higher. Because
these are mechanically based systems, their charge/discharge ratio is
1:1 and is capable of cycling tens of thousands of times. However, as
the components of the system can be optimized for either power or
energy, the needs of one application may make that design poorly
suited for the other.
Cost issues
According to the EPRI-DOE Handbook of Energy Storage, initial capital
costs of flywheel energy storage systems are in the range of $459 per kW.
Although flywheel installations generally cost at least 50% more than
LA batteries for similar (UPS) applications, their total life-cycle costs can
be vastly less expensive, even while providing better reliability. Driving
this competitive life-cycle comparison are the much lower operation
and maintenance costs, plus much greater life of the flywheel versus
the battery strings. These comparisons usually assume an average power
quality environment; however, flywheels can provide the same level of
capability in environments that would considerably shorten a lead-acid
130
battery’s life. Harsh temperatures and constant fast and deep charge/
discharge cycling dramatically increase the cost of choosing batteries for
these installations, but rarely affect the life of a flywheel. For regenerative
energy and frequency response strategies, batteries are not an option, so
flywheels must compete against standard business practices.45
Installations
Flywheels used for electrical energy storage are well established in the
UPS market, but recent high-speed flywheels are supporting a growing
number of alternative applications.
Example—STMicroelectronics, Rousset, France. 46 Normal

operations in the semiconductor industry are extremely susceptible to
power disturbances and thus require high-quality power at all times.
To reach zero shutdown from power disturbances, STMicroelectronics
installed, in 2001, a series of flywheel UPS systems from Active Power
in two areas of its semiconductor fabrication facility in Rousset, France.
The first application was to provide voltage support for variable speed
drives (VSDs), and the second provided a ride-through for wafer fad
production equipment. In the first placement, three 160-kW single
flywheels were installed (1 per VSD) to keep the DC bus above the cutoff
for the drives, as sags more than 22% above 10 ms were shown to turn off
the VSDs. Within the first year, these units experienced 200 discharges,
some of which would have resulted in shutdown of the drives. On the
second application, five 600-kVA dual-wheel flywheel UPS systems were
chosen to support the utility power for the wafer fabrication line. In
both locations, flywheels were chosen because of their reliability, small
footprint, and tolerance for temperature variations. Specifically, flywheels
were chosen over batteries because the potential for a battery to fail
within the 40-battery string was found to be quite high over the life of
the unit, leading to the greater probability of a battery UPS failing in
this application (10 times higher than flywheels). In addition, the extra
131
room required (for extra batteries) to achieve the goal of zero shutdowns
was not available because interior space is at a premium in a chip
fabrication facility.
Example—Deluxe Films, Toronto, Canada. Precision batch

processing of time-sensitive materials also relies on high-quality
power. Here, a motion picture processing facility in Toronto, Ontario,
Canada, had been feeling the impact of Toronto Hydro’s inability to
maintain high-quality reliable power during peak periods because
of high local load growth. The processing facility, owned by Deluxe
Film Laboratories, the world’s largest motion picture film-processing
laboratory, required high-quality power to operate the film-processing
equipment. Over the last few years, the facility experienced numerous
power spikes, sags, and even outages. Although many of these were
short-term voltage problems, predominately lasting for five cycles or
less, the company learned the true value of reliable, high-quality power
during the 2003 blackout, when the facility lost power unexpectedly.
Losing power like this—or even experiencing large voltage sags and
spikes—created an enormous problem for the film company because
it only takes one interruption to cause the equipment to shut down
and lose the batch of production. If such a shutdown occurs during
development of a critical piece of a large movie studio’s production,
the whole process must start all over. Not only does this produce
a direct cost of lost material (multiple copies of the film) that can
reach into the millions of dollars, future business may be lost from
demanding clients. For these reasons, the company needed a means
to ensure that the film production process could remain operational
even during power outages or service interruptions. Therefore, in the
fall of 2003 a SatCon Power Systems’ STARSINE 2.2-MW rotary flywheel
UPS system was installed at the facility. During installation, the unit
was functionally demonstrated by running the entire facility during
normal production without external utility system power to simulate
a power outage similar to the one Toronto experienced during the
2003 blackout. Since installation, more than 25 power quality events
132
sufficient to disrupt production have been recorded. The rotary UPS

was able to mitigate each of them and prevent a shutdown or process
failure, sometimes without the facility operators knowing—as was the
case one day in April 2004 when the power system experienced so
many problems that the unit engaged and remained in backup mode
for more than four hours (from correspondence with Mike Gibson,
Satcon Technologies of Boston).
Example—Lyon, France Metro.47 The Lyon France Metro installed

a 600-kW (4.3 kWh) Trackside Energy Management System from Urenco
Power Technologies (UPT) in 2003 (3 x 200-kW units) to solve some
chronic issues from overvoltage. During off-peak periods when not
enough trains were on the network to absorb the excess energy generated
by braking trains, the resulting overvoltage caused harmful wear and
damage to the equipment onboard the rolling stock. Several other energy
storage systems were evaluated but found to be not suitable because of
the cyclical demands of the subway environment. The flywheel-based
system acts as a short-term sink for the energy from the braking trains,
with the regenerative energy used to power auxiliary station load, such as
the lighting, ventilation, and escalators at the stop near the Hôtel de Ville.
This solution has reduced the wear and tear to the equipment onboard
the rolling stock and the brake maintenance costs.
Example—Usibelli Coal Mine, Usibelli, AK. Although most

flywheels have relatively little energy storage capacity because of their
size, some flywheels can actually be quite large. One such example was
installed at the Usibelli Coal Mine in Usibelli, Alaska, for $1 million in
1982 and is used to support a dragline (a very large bucket-loader) at
the coal mine. The electrical load from the dragline is an erratic 8-MW
band (plus or minus 4 MW) every 60 seconds, necessitating some means
to lesson the impact on the relatively lightly loaded GVEA power grid.
The flywheel itself is made from three one-foot thick, eight-foot diameter
steel disks (total of 40 tons), with a normal operational speed from
900 RPM to 1,100 RPM. During operation of the dragline, the flywheel
133
trades power with the dragline by slowing down during dragline loading
and speeding up during dragline unloading. To support this activity,
the flywheel can produce up to 5.2 MW for three-second (260 kWh)
bursts, with any additional power required coming from the 1.8-MW
motor/generator. With this flywheel system, the average power demand
from the utility is around 2 to 2.5 MW, and GVEA only sees around a
500-kW band of power fluctuation (from personal correspondence with
Tim O’Neil, Usibelli Coal Mine Inc.).

Low lifetime costs and the ability to survive in harsh operating
environments are the core strengths for the future success of flywheel
energy storage technologies. Rapidly maturing from an already
cost-effective base, life-cycle costs are already becoming a greater concern
to owners—especially those who have already purchased earlier battery
solutions for harsh applications. Flywheels currently represent 20% of the
$1-billion market for the energy storage component of high-powered UPS
market.48 With the greater understanding of space conditions of other
technologies, the minimal space and space-conditioning requirements
provide flywheel energy storage technologies a significant advantage in
lifetime costs and usability—leading to a growing penetration of this
existing market.
However, besides simply cannibalizing the battery share of the

market, the capabilities of flywheels can also extend into applications
not currently served by any solution because of the limited cycle life
of many storage technologies. These markets—regenerative energy,
frequency regulation, and the like—hold out the greatest opportunity
in the long run for a much wider integration of flywheel systems into
larger industrial settings. Coupling a flywheel with other power systems
in stressful environments can extend the life of these other components
by providing rapid injections/withdrawals to handle deep, cyclic events.
134
The main challenges toward furthering the introduction of flywheel

technology are costs and customer education. Continued development
of the technology will lower the initial costs of these units, but they will
always have a higher initial cost than the LA battery solution. Therefore,
increasing the understanding of the importance of life-cycle costs versus
simply initial costs will be important for flywheels to increase their
market share of the existing UPS market. Current pilot projects using
flywheels for regenerative energy solutions or frequency regulation are
also laying the groundwork for penetration into these other fragmented
markets over the coming years.
Major developers
Current manufacturing of flywheel units is dominated by traditional
steel rotor flywheel-based firms such as Active Power, Piller Inc. (rotary
UPS and flywheel), Vycon Energy, Hitec Power Protection (rotary UPS),
and SatCon Power Technologies (rotary UPS).
A number of other companies have begun developing higher-speed

composite rotor based systems such as Pentadyne Power, Beacon Power,
Boeing, and AFS Trinity.
135
Electrochemical
Capacitors
Summary
Electrochemical capacitors are similar to batteries in that they have
two electrodes immersed in an electrolyte and separated by a porous
separator. The goal in this design is to obtain the energy storage capacity
of a battery with the operating characteristics of a capacitor. They store
energy via electrostatic charges on opposite surfaces of the electric
double layer, which is formed between each of the electrodes and the
electrolyte ions (fig. 3–12). Ultracapacitors move electrical charges
between solid-state materials rather than through a chemical reaction;
therefore, they can be cycled tens of thousands of times more rapidly
and are not affected by deep discharges as are chemical batteries. Because
the total amount of capacitance in the unit is directly related to the
surface area of the electrode, energy stored increases with the square of
the applied voltage. Their wide-ranging capabilities and reliability make
electrochemical capacitors increasing favorites for retail markets, and
they are promising for use in solving some transmission system stability
applications in small niche applications. The prospects for electrochemical
capacitors are strong, partially because of their infrequent use in the
today’s electric power industry. Currently, these technologies have grown
into a $100-million market supporting mobile and communication
systems, with growing inroads into fuel cell, motor starting, and the
transportation market where size, weight, performance, and maintenance
costs are important. However, significant—but reachable—hurdles must
be overcome before electrochemical capacitors become widespread in
the electric power industry. As a developing technology, their operational
capabilities must be improved, their costs lowered, and their reliability
(both in manufacturing and operation) must be enhanced.
136
Fig. 3–12. Electrochemical capacitor (Courtesy of Maxwell Technologies, Inc.).
Historical origins
The first development toward electrochemical capacitors began
more than 100 years ago, with some of the most promising research into
a functional energy storage device occurring in the early 1960s at Standard
Oil of Ohio (SOHIO). After the decision was made to not continue with
the research program, SOHIO licensed all of its double-layer capacitor
technology to NEC of Japan in 1971. After a few years of refining the
technology and early manufacturing processes, NEC developed the first
successful commercial electrochemical capacitor in the late 1970s for
computer memory backup applications. Other companies continued to
develop their own products and processing methods during this time. One
was a Japanese electronics firm, Matsushita Electric Industrial Company,
which patented a manufacturing method for improved electrochemical
capacitor electrodes in the 1980s. By the 1990s electrochemical capacitors
were scaled up and commercialized for pulse-power applications, fuel cell/
engine starting applications, and specialty energy storage applications in
137
the electric vehicle market. Electrochemical capacitors are often referred to

by company trademark names including supercapacitors, ultracapacitors,
and electric-double-layer capacitors.49

The goal of the electrochemical capacitor design—two electrodes
immersed in an electrolyte and separated by a porous separator—is
to obtain the energy storage capacity of a battery and the operating
characteristics of a capacitor. It stores energy via electrostatic charges on
opposite surfaces of the electric double layer, which is formed between
each of the electrodes and the electrolyte ions. The electrodes are often
made with porous carbon material, chosen because of the extremely large
surface area available. There are four distinct design types of electrostatic
capacitors, designated as types I–IV. Type I electrostatic capacitors are
symmetric in design, with similarly sized positive and negative electrodes.
The electrolyte is an aqueous solution with a high concentration of
sulfuric acid. The maximum cell voltage is limited to 1.2 V, but it nominally
operates at 0.9 V. Type II electrostatic capacitors, which are the most
commonly used, are similar to type I, but with an organic electrolyte
(ammonium salt in organic solvent). The maximum cell voltage is higher
because of the electrolyte, with a maximum cell voltage of 2.3 V to 2.7
V. Type III electrostatic capacitors are asymmetrical in design. This design
has a high cycle life partly because of the depth of discharge for one
of the electrodes compared to the other. The electrolyte is an aqueous
solution similar to type I electrostatic capacitors, giving a maximum cell
voltage of 1.4 V to 1.6 V. Type IV electrostatic capacitors are new and
not yet commercially available. They are asymmetrical in design, with an
organic electrolyte similar to type II electrostatic capacitors, giving the
design the potential for a maximum cell voltage of 4 V.50
Because ultracapacitors move electrical charges between solid-state

materials rather than through a chemical reaction, they can be cycled
tens of thousands of times, cycled more rapidly, and are not affected
138
by deep discharges as are chemical batteries. Because the total amount

of capacitance in the unit is directly related to the surface area of the
electrode, the amount of energy stored increases with the square of the
applied voltage. Because of the cell’s charging properties, the voltage rises
linearly with time when charged at a constant current, or when charged
at a constant power. During charging, the electrolyte becomes polarized,
with roughly half of the electrolyte material transferring an electron to
the other half. These charged ions then migrate to the oppositely charged
electrode, forming a charged layer on the surface—although no electrons
are exchanged. These two layers of separated charges then form half of the
electronic double layer, with a similar layer on the second electrode forming
the other half. These are closely spaced, with the negative and positive
charges separated by only half the diameter of the electrolyte ions.
The ability of electrochemical capacitors to respond quickly and

reliably at full power is the key to their usefulness in the retail market,
and they have the promise to solve some transmission system stability
challenges. In the retail market, electrochemical capacitors are currently
used for engine starting and pulse discharges used for power quality and
bridging applications (UPS). For engine starting, large diesel generator
and locomotive engines use them because they take up less space and
weigh less than conventional batteries. They also start well in cold weather,
have a long battery life, and are low maintenance. For UPS applications,
electrochemical capacitors can either replace totally, or are currently
being investigated for supplementing, conventional chemical batteries. By
combining a front-end electrochemical capacitor with a chemical battery,
the life of the batteries can be extended. In this setup, the electrochemical
capacitors provide power for short duration interruptions and voltage sags,
significantly reducing the cycling duty on the batteries so that they only
provide power during longer interruptions.
As with all other energy storage technologies, environmental and

operating conditions will have an impact on the scope and type of
applications available to electrochemical capacitors. The average DC–DC
139
round-trip efficiency of an electrochemical capacitor is 80% to 95% during

normal operation, with variations because of the multiple design types.
Temperature can affect the cycle life of an electrochemical capacitor. The
normal upper operating temperature of an electrochemical capacitor
is generally 85°C, with self-discharges increasing as the temperature
rises; elevated temperatures decrease unit product life. Conversely,
electrochemical capacitors have very good lower temperature operating
characteristics, with the lower operating temperature for some variants
being as low as −50°C. Because of their design and lack of moving parts
or chemical reactions, electrochemical capacitors are capable of hundreds
of thousands of charge/discharge cycles.51
Cost issues
According to the DOE-EPRI Handbook of Energy Storage, total system
costs for an electrochemical capacitor are $456 per kW, with 40% of this
attributable to the storage module. This cost cannot be readily compared
to other energy storage facilities, as this is for a single-purpose facility
rather than a multifunctional one supported by other energy storage
technologies. However, for individual applications where space, weight,
and lower maintenance requirements are key, this may prove decisive. Life-
cycle capability becomes known as the unit ages, which is an important
contributor to the cost-effectiveness of these versus LA batteries that
can fail in sudden death—often requiring additional battery backups.
As this is a recently commercialized product, significant additional cost
reductions are envisioned, which will be primarily driven by manufacturing
advancements and scale benefits from larger production runs rather than
any new product breakthrough in material science. To date, automated
manufacturing techniques have reduced the cost of electrochemical
capacitors significantly. From the mid-1980s to today, manufacturers
have reduced their costs 95%. With continuing manufacturing process
improvements, it is estimated that the storage modules themselves can be
reduced in cost by anywhere from 33% to 50%.52
140
Installations
Siemens’ SITRAS SES System.53 The braking of light rail systems
remains an area of significant energy loss for city governments, and,
therefore, significant effort has been made to reduce it. In the 1980s,
many subway and light rail systems began replacing their friction wheel
brakes with systems that reversed the train’s electric motor and acted as
a generator (for braking); then the regenerated energy was fed back into
the rail power supply system. Unfortunately, much of this regenerated
energy can only be used if there is a rise in demand proportional to a
braking train—such as a train leaving a station; otherwise, only 66% or so
of the energy will be used in the system. Without such an applicable load,
the voltage of the system increases, so the remaining energy is dissipated
as heat through radiators on the cars—raising the temperature in the
tunnel and, thus, requiring greater power for air conditioning. To alleviate
both the loss of energy and the voltage instability caused by this practice,
Siemens developed the SITRAS Static Energy Storage (SES) system—a
stationary electrochemical capacitor–based system rated at 1 MW,
2.3 kWh. An important deciding factor in choosing these electrochemical
capacitors over batteries or other energy storage technologies is their
long cycle life and low maintenance requirements. The unit is made from
42 Maxwell Technologies’ 2,400 F cells (each with a capacity of 2,400 F).
The SITRAS SES can automatically switch from «regenerative energy»
to «voltage-stabilizing» mode, as needed. When the system is used
on short-distance traffic—common on light rail systems—the power
requirements at the station can be cut by approximately 30%. To date,
trials are taking place in Portland, Oregon, and Dresden, Germany, with
actual orders for systems received from Bochum and Cologne, Germany,
and Madrid, Spain.
141

The prospects for electrochemical capacitors are strong, in part
because of the current lack of use of these technologies in the electric
power industry. Currently, these technologies have grown into a
$100-million market, supporting mobile and communication systems,
with growing inroads into fuel cell, motor starting, and the transportation
market where size, weight, performance, and maintenance costs are
important. In particular, Russian companies continue to dominate this
field for large electrochemical capacitors (needed for cold-weather engine
and generator starting). Over the coming years, each of the different
types of electrochemical capacitors is expected to improve the level of its
energy density from 50% to 100%, potentially raising the energy density
of some electrochemical capacitors to that of LA batteries—leading to
even greater market penetration.
Significant—but reachable—hurdles remain to be overcome before

electrochemical capacitors become widespread in the electric power
industry. As a developing technology, their operational capabilities
must be improved, their costs lowered, and their reliability (both in
manufacturing and operation) enhanced. Core to all of these will
be required progress along three avenues: capacitor design, material
selection and usage, and manufacturing processing. Improvement
along each of these avenues will lower the cost of each unit, and also
improve reliability and the unit’s ability to cycle heavily—a prerequisite
for potential industrial and utility applications. For instance, large-scale
systems with hundreds of electrochemical capacitors would require very
uniform qualities of each cell (construction and operations) to prevent
unbalanced electrical voltage and current flows over the life of the
system. Although larger-scale manufacturing will provide significant cost-
reduction opportunities, cost reductions through manufacturing process
improvements will also be critical.
142
Major developers
There are a number of electrochemical capacitor manufacturers
based in a number of countries, including Russia, Germany, Japan,
France, Korea, and the United States. Some of the larger manufacturers
and installers of these systems include ELIT, ESMA Joint Stock Company,
NESS Capacitor Company, Maxwell Technologies, and Saft.
Superconducting
Magnetic Energy
Storage
Summary
Superconducting magnetic energy storage (SMES) systems store
energy in the magnetic field created by the flow of direct current in a coil
of cryogenically cooled, superconducting material (fig. 3–13). An SMES
system includes a superconducting coil, a power conditioning system, a
cryogenic refrigerator, and a cryostat/vacuum vessel to keep the coil at
a low temperature—required to maintain the coil in a superconducting
state and, thus, allow it to be highly efficient at storing electricity (more
than 99%). These units can respond within a few milliseconds, and very
high power output can be provided, but only for a brief period of time—
therefore, these units are best suited to provide repeated, short-interval
discharges. Although their costs are high compared to other storage
technologies in respect to the cost per unit of energy stored, they are
cost competitive with other flexible AC transmission systems (FACTS)
equipment or transmission upgrade solutions, which are normally the
competing choices. These facilities currently range in size up to 3 MW, and
143
are generally used to provide power-grid stability in a distribution system

and power quality at manufacturing facilities with critical loads highly
susceptible to voltage instabilities.
Fig. 3–13. Superconducting magnetic energy storage (SMES)

(Courtesy of American Superconductor).
Historical origins
Based on the development of high-powered magnets in the early
1960s, the original concept for an SMES facility came in 1969 with a
proposal for a large-scale unit able to provide commodity arbitraging
capabilities—primarily diurnal opportunities associated with nuclear
power. For the next two decades, a number of major programs for study
of SMES began, centered on research in the United States, Japan, and
Europe. Early U.S. research programs like the one begun at the University
of Wisconsin in 1971 were supported and expanded on by groups such
as the DOE, DOD, and EPRI. As mentioned earlier, much of this early
focus and research on SMES was for large-scale energy storage, with the
hope of storing hundreds of MWs or more as a competing technology
144
for PHS facilities. However, the need for pulsed-power delivery for fusion
power research led (among other expanding research opportunities) to
further development of smaller SMES units for transmission system volt-
age stability and industrial power quality during the 1980s. One early
installation in the United States occurred when Bonneville Power tested
a 30-MJ SMES facility to provide area control and frequency regulation
on long-distance power lines along the West Coast. Other installations
included some in Japan, where developers such as Hitachi tested a variety
of units to evaluate the system’s capability of providing distribution line
stability, one being a 5-MJ SMES unit at the Hitachi Works, and a second
1-MJ unit with Chubu Electric.

SMES systems store electrical energy in the magnetic field created
by the flow of direct current in the coil of a cryogenically cooled super-
conducting material. An SMES system typically includes a superconducting
coil, a cryogenic refrigerator, a cryostat/vacuum vessel to keep the coil at
a low temperature, and PCS equipment. The superconducting coil can be
developed in either a solenoid or torus shape. Although the torus design
uses twice as much superconducting wire as the solenoid, its shape pre-
vents the magnetic field from penetrating into the surrounding space, in-
creasing efficiency. However, these same magnetic fields in the torus cause
strong outward forces in the magnet and must be countered with strong
reinforcements to prevent stress from degrading the magnet. For this rea-
son, the solenoid design is used in current commercial designs.
Typically, the low-temperature superconducting coil is made of

niobium-titanium, cooled to 4.2°K by liquid helium. The SMES system
also uses high-temperature (77°K) superconducting components as an
interface between the cryogenic area of the SMES and the conductors
145
to the PCS. With the use of low-temperature superconducting material,

the loss of the magnetic energy from resistance (heat) in the wire is
avoided, enabling the superconducting coil to carry a very large current
with little loss of power almost indefinitely—if the superconducting
material is kept cold. By maintaining such a state, the energy storage
capacity of the unit is enhanced because energy stored within the coil
is proportional to the square of the current, and the coils vary in size
depending on the energy storage capacity. The material properties of
these superconducting coils are obviously of great importance, because
temperature, magnetic field, and current density are all variables
that trade off against each other in their design. In an ideal case, the
material would maximize all three, but the physical properties and the
manufacturing process normally degrade the performance of the coil
from an ideal balance. During operation of the SMES unit (once charged),
a small trickle charge is required to replace the power lost in the non-
superconducting part of the circuit.
Although applications for just rapid power delivery are limited,

coupling an SMES unit with sufficient power electronics gives the
capability to rapidly inject both real and reactive power, providing relief
in two markets: power transmission and the industrial sector. In the first
market application—transmission voltage support—the SMES/power
electronics unit provides fast response to voltage fluctuations. This
protects the power grid from the destabilizing effects of short-term
events such as voltage dips caused by lightning strikes and downed
poles, sudden changes in customer demand levels, and switching
operations. When the system detects a disruptive event (voltage drop)
on the connected transmission line, these systems are able to respond
within a few milliseconds, and in most cases can restore voltage
stability to the power grid within one to two seconds (far less than the
five seconds used as a guideline by many utilities). For these voltage
stability roles, the first second is the most crucial, making response
time a more valuable capability than gross energy storage capacity.
146
However, besides providing reactive power, experience with SMES units

on the Wisconsin Public Service system (see example following) has
shown that providing some real power significantly improves system
performance and increases the rate at which voltage stability is restored
during a voltage instability event. Finally, by providing voltage stability in
this way, SMES power electronics units are able to increase the power
transfer capacity of a locally congested power grid, often by upward of
15% depending on the layout of the system.
A second market where SMES units can be used is in the indus-

trial power quality market to protect manufacturing operations from
momentary sags on critical loads to prevent motor stalling or shutdown.
A typical installation can react to voltage sags within a few milliseconds
and can protect critical processes at the facility from voltage sags up to
60% for times exceeding one second.
SMES units are highly efficient at storing electricity, leading to a

round-trip efficiency of more than 99%. For units targeting the utility
market and the industrial market, the charging time for a unit is generally
on the order of 90 seconds, with a full discharge taking less than 1 second,
giving the system roughly a 90:1 ratio for charge/discharge cycle time.
The main difference in these time requirements is the need to keep the
temperature of the equipment in a superconducting state; more heating
occurs during the charging than in the discharging event, necessitating a
longer charging period. However, because these systems are not designed
to cycle rapidly, this timing requirement is not of much concern as the
more valuable market capability is the short discharge time—charging
can easily occur after the system has returned to a stable state. Because
of the solid-state makeup of the equipment, most SMES units have a
cycle life of thousands of charges/discharges without degradation to the
magnet, giving a unit a design life of 20 plus years.
147
Cost issues
According to the EPRI-DOE Handbook of Energy Storage, a multi-
megawatt, single-functional SMES system has a system cost of $509 per
kW, with a little more than 60% from the battery module. However,
these costs are difficult to compare to other storage technologies
because of the scale and purpose of the SMES units. SMES costs also are
difficult to compare to other storage technologies in respect to the cost
per unit of energy stored, and their market applications are somewhat
different as well, so a direct comparison with other energy storage
technologies is not always valid. More importantly, SMES systems are
cost competitive with other FACTS equipment or transmission upgrade
solutions, which are normally the competing choices. For example, in
the Wisconsin Public Service example (described in a following section),
American Superconductor’s D-SMES unit, which costs $4 million, was
found to cost far less (with a faster installation time) when compared
to other solutions, which would have cost anywhere from $6 million
to $15 million. Besides these up-front costs, operational expenses for
these units are also comparable to, or less than, alternative technology
solutions. With advancements in superconducting material capability and
processing costs, the cost of the storage component has the potential to
decline an additional 30%. Finally, the PCS equipment associated with
SMES units is generally more sophisticated and is integrated as a larger
component of the overall D-SMES system than in the other storage
technology installations in order to handle the rapid discharge capability
of the SMES unit and because the focus of the D-SMES’ capabilities is in
VAR production.54
148
Installations
SMES systems have been in use for several years at utility and industrial
sites in the United States, Japan, Europe, and South Africa to provide both
transmission voltage support and power quality to customers vulnerable
to fluctuating power quality. In these two markets, more than 100 MW of
these units (with the average unit being 3 MW or less) are estimated to be
currently in operation around the world.
Example—Wisconsin Public Service.55 In 2000, Wisconsin Public

Service Corporation (WPS) installed six of American Superconductor’s D-
SMES units to handle voltage disturbances on the WPS Northern Loop
system. Inductive motors powering the region’s numerous paper mills
comprise a large portion of the system’s peak load, which is added to
during the summer months by the significant air conditioning load of
tourists in the region. Inductive motors pose special stability challenges
to utility networks, as induction motor stalling at paper mills in northern
Wisconsin sometimes precipitates the voltage collapse of the utility
transmission system serving almost 33% of the state. To solve this
instability problem, WPS has planned a transmission line upgrade but
requires an interim solution until the upgrade can be moved through
the planning process and brought into service. After evaluating other
options (smaller power line extension, Static VAR Compensators [SVC],
etc.) the D-SMES unit was chosen as having the least cost and lowest
installation time. Once installed, the D-SMES units were able to solve the
existing voltage instability problem while increasing power grid capacity
by 15%. The installation has proved to be so successful that a seventh
unit has been moved to the area to continue deferring the need for the
additional transmission investment as the construction plans for the
upgrade continue.
149
Example—Stanger, South Africa.56 SMES units can also provide

significant power quality benefits in the industrial market outside the
oft-mentioned microelectronics industry. For example, a 1-MVA (3.0 MJ
storage) Power Quality Industrial Voltage Regulator (PQ IVR) (with
SMES unit) from American Superconductor (through Eskom, the local
utility and partner) was installed at a Sappi paper mill in Stanger, South
Africa, in 1997 to protect the plant from voltage sags (a voltage sag of
only 250 milliseconds can shut down a paper machine). In this facility, a
continuous paper web runs through several independently driven, speed-
synchronized units. Any interruption in a drive snaps the web, resulting
in downtime (measured in hours) for cleaning, rethreading, possible
repair of damaged equipment, and finally restarting. These disruptions
are significant, with production losses and damages at a typical paper
mill averaging $50,000 per sag. The SMES unit has provided significant
protection to the facility, protecting the mill from more than 70 voltage
sags during its first year of operation—one-half of which could have
caused the plant to shut down.

SMES technology has found limited success to date, but real
prospects for this technology exist along with opportunities for greater
market acceptance and use going forward. The largest potential market
available for SMES technology currently is to support utility transmission
voltage levels against sudden disruptions. Here, two trends point toward
growing market success for these units. First, current SMES systems
compete well on price and capability against existing FACTS solutions and
transmission upgrades for areas with growing transmission instabilities.
Second, as additional investment capital is directed toward bolstering
the existing but strained transmission infrastructure, SMES technologies
stand to capture additional sales as this multibillion-dollar market
expands. With the increasing constraints imposed on the transmission
150
system from a growing load and scant system expansion, tools to provide
system stability and increase the available transfer capability of the
power grid are of increasing importance to utility power grid managers.
As Charles Stankiewicz, vice president and general manager of American
Superconductor notes, “Transfer of power that is limited by voltage
stability can be effectively increased by solutions anchored by dynamic
devices such as our D-SMES system. The end result is a more reliable
network with an ability to import more economical electricity from
outside the service territory.”
The second target market for SMES technologies, industrial power

quality, also has significant possibilities for added sales because high-speed
and precision equipment continues to make inroads into manufacturing
processes, with these machines being particularly inflexible to voltage
instabilities. If an increased use of high-temperature superconductors
becomes possible in the construction of these units, additional cost
reductions from a vastly simpler and less expensive cooling system could
also assist the competitive position of these units.
However, for all of these prospects, significant challenges still

remain because SMES technology currently possesses only a limited
installation base, leading to a greater level of misunderstanding of its
actual installation requirements and operational capabilities. Although
technologies competing to provide power grid voltage support and
alleviate congestion also continue to meet a general reluctance, there
is greater emphasis toward using these more traditional concepts.
Essentially, the greatest issue of concern by utilities is the long-term
reliability of the SMES unit in a utility setting rather than any cost or
efficiency level of the unit. As is well understood, most utilities are
skeptical of such technologies until their reliability and maintenance
profile is proved elsewhere to be similar to comparable equipment.
Solving this problem of customer understanding will only be possible
through continued success and promotion of installations such as the
ones at WPS and Entergy.
151
Major developers
The only major manufacturer of SMES products is American
Superconductor of Westborough, Massachusetts. The firm sells its
SMES product in cooperation with General Electric to utilities under
the D-SMES line and to industrial customers under the PQ IVR name.
The D-SMES unit is housed in a 50-foot truck trailer for easy installation
at utility substations. As a mobile unit, the system provides transmission
and distribution support that can be deployed as needed to reduce utili-
ties’ changing transmission constraints on their existing transmission and
distribution power grids.
Thermal Energy
Storage
Summary
Not generally thought of when discussing energy storage tech-
nologies, thermal energy storage (TES) systems are already well
established as a means to reduce peak-cooling loads for commercial and
industrial firms. Actively developed since the early 1980s, TES systems
have advanced steadily so that now they are easily manufactured and
installed as modular units, with nearly 7,000 systems active around the
world that displace nearly 5 GW of peak load requirements. TES units are
designed to work with the existing building’s cooling system (a chiller),
which chills either water or an ethylene glycol solution for the heat-
exchange air conditioning of the building. The TES system simply uses
the chiller to make ice or chilled water during off-peak hours and stores
it in the insulated storage tanks (fig. 3–14). During the day, TES systems
supplement the chillers by providing the cooling load for the commercial
152
buildings, allowing for smaller chillers and substantially lowering air

conditioning operating costs—both the amount of on-peak energy
used and peak demand charges. On average, a retrofit installation has a
one-to-three-year payback, and units integrated into new construction
can often pay for themselves by reducing the amount of cooling system
required for the facility.
Fig. 3–14. Thermal energy storage (TES) (Ardour Capital Investments).
Historical origins
The development and use of TES—often referred to as off-peak
cooling—for medium and large commercial cooling systems began in the
late 1970s and early 1980s, when utility-sponsored conservation and load-
sifting strategies were encouraged to avoid the need for new power plants
as the cost of these assets was rising quickly. As the number of utilities
wanting more load-shifting grew, state regulatory commissions enacted
demand side management (DSM) programs to let utilities offer incentives
for additional TES installations. By the early 1990s, many new manufacturers
153
were producing TES systems for the commercial market. Unfortunately,

many of these products were developed only to obtain the DSM rebate,
without a clear ability to operate in the long term. This left many early
adopters of the technology skeptical and left a misunderstanding of the
technology in general. The early 1990s saw continued growth in the market,
and realignment and maturing of the rebate strategies led to a decline in
the weaker product designs. Since then, continued work toward improved
coolants and validating the long-term economic savings have improved
the standing of the technology. Although the move toward retail market
deregulation in the mid-1990s initially hurt the industry with promises
of cheap electricity and the loss of rebates, the reality of state-level retail
market deregulation has shown that the basic economics for supporting
TES—both for the consumer and the utility—are still valid.

A TES unit generally consists of a heat exchanger system of helical
coils (metal or plastic) placed inside an insulated storage tank, a
refrigerant pump, and an air-cooled condensing unit. Being well insulated,
these storage tanks can normally be located anywhere, either indoors or
outdoors. TES units are designed to work with the existing building’s
cooling system (a chiller), which chills either water or an ethylene glycol
solution for the heat-exchange air conditioning of the building. The TES
system simply uses the chiller to make ice or chilled water during the
night and stores it in the insulated storage tanks. Besides being available
because building cooling is not needed at night, chillers also perform
more efficiently at night when the outdoor temperatures are relatively
low (improving efficiency). When cooling is needed during the day, the
TES system can then either assist or replace the chiller for meeting the
day-time cooling load of the facility. If designed into new construction
from the outset, the TES provides additional flexibility in the sizing of
the entire cooling system needed. Normally, a building’s chiller system
operates to produce cold air for the building during the hottest part of
154
the day, which happens to be during the peak daily load for the utility.
For sizing purposes, these chillers are normally designed to be run on
the hottest possible day, all day, at full load. For most usage needs,
therefore, the unit is far oversized, so incorporating a TES system allows
the building’s chiller system to be sized and run more efficiently.
For these reasons, a TES system operates in an energy management

role in the retail energy market for commercial users with medium to
large cooling loads. The TES then essentially shifts building cooling from
expensive on-peak electricity periods to off-peak periods. One important
determinant of the true value of TES is the difference in peak versus off-
peak utility rates. Overall, by operating in this manner, TES systems not
only reduce the peak energy used, but also its cost. First, off-peak power
is cheaper because of time-of-use rates for large energy users; and second,
reducing peak demand allows for a reduction of the demand charge.
As TES systems are truly system optimizers for existing building

chiller systems and not stand-alone units connected directly to the
electrical load, comparing their efficiency directly to other energy
storage technologies is difficult. On average, TES systems lose through
thermal loss only approximately 1% of their energy during inactive use.
This efficiency is comparable to or much higher than other technologies.
Another way to look at the effective efficiency improvement offered by
TES systems is the effect of shifting the load to off-peak periods. For
instance, an earlier California Energy Commission (CEC) report explained
that the summer peak heat-rate for an average steam unit in California
was 11,744 Btu per kWh, compared to its off-peak idling heat rate of 7,900
Btu per kWh.57 Using TES systems effectively arbitraged that difference
for the utility and reduced the delivered price for the consumer. Generally
the units operate in charging mode (making ice) for the same length of
time as they do in discharging (supporting cooling load).
155
Cost issues
Typical costs for TES facilities can vary greatly depending on whether
a unit is installed as a retrofit or as part of new building construction.
For retrofit installations, exact costs are difficult to establish as each
cost is project specific, but generally they range from $250 per peak kW
shifted to $500 per peak kW shifted.58 TES systems can be justified as a
retrofit for customers with existing air conditioning equipment based
on the lower costs incurred by using off-peak instead of on-peak power.
Typically, however, the decision to proceed with a retrofit installation is
not simply due to the cost savings in the energy bill, but also some other
capital asset decision (or a combination), such as an aging chiller plant,
building expansion, or limited electrical supply. Typical payback periods
for these retrofit TES systems normally range anywhere from one to three
years. This is possible through savings from reducing the demand charge
and lowering peak power purchases. The demand charge in particular is a
target for reduction because for a typical commercial facility, the demand
charge can often equal 50% of the entire cost of service. With lower on-
peak use of power, these facilities can reduce peak power demand by
50% and can reliably reduce cooling load costs up to 30% compared to
standard cooling equipment.59
Integrating a TES unit with new construction (medium to large

construction) provides for even greater benefits—in most cases, the
decreased design requirements on the building’s cooling system will pay
for the cost of the system, plus provide the additional operating cost
reduction described in the retrofit installation. In particular, because TES
systems allow a building’s air conditioning system to operate at lower
temperatures, duct sizes can be reduced by 20% to 40%, and chillers can
be 40% to 60% smaller than normally required.60 Because the cooler air
also requires a smaller volume of air to cool the building, fan motors,
air handlers, and chilled water pumps are smaller and less costly. Based
156
on the reductions in capital equipment needed when a TES system

is integrated into the original design, EPRI has estimated that overall
ventilation and air conditioning costs are reduced by 20% to 60%.61
Installations
TES systems have one of the widest application bases of all consumer-
side energy storage systems, with roughly 7,000 units installed worldwide
and a combined capability of shifting nearly 5 GW of peak-demand load.
Although most of these units are for individual facilities, some are for
multifacility systems and can become quite large. For example, in Chicago,
Illinois, Unicom Thermal Technologies has installed a central 66,000-ton
cooling unit that serves its commercial customers in the downtown area.
(A ton is the standard metric for cooling load requirement—12,000 Btu
heat removal per hour—and represents the amount of cooling energy
melting one ton of ice over a 24-hour period.)
Example—Grossmont Hospital, La Mesa, CA.62 Hospitals have

large cooling loads, and Grossmont Hospital in La Mesa, California, is
no exception. In looking for way to lower the cooling load cost of the
490,000-square-foot medical facility, the management of the facility
became interested in incorporating off-peak cooling strategies. A
feasibility study to determine the cooling load of the hospital found that
the facility had a cooling requirement of 740 tons. It was found that by
using a TES facility made up of 22 Calmac ICE BANK tanks, the chiller
required could be reduced in size to a 320-ton unit. Besides allowing
for a smaller chiller, the use of the TES unit qualified the hospital for a
significant up-front subsidy from San Diego Gas and Electric to defray
a portion of the system cost—which helped convince hospital officials
to proceed with the retrofit project. The installation of the ice-storage
system enabled Grossmont Hospital to shift 232 kW of load to off-
peak and avoid paying a demand charge. Once operating, the cooling
cost savings from the unit approached $17,000 per year, providing a
payback period of approximately three years. Using the modular Calmac
157
ICE BANK system had other benefits; the unit could be easily expanded
to handle additional cooling loads from planned additions to the hospital
complex, such as a 60,000-square-foot Woman’s Center complex that
was slated to be added the following year.
Example—San Jacinto College District, TX.63 TES projects are

also important to state governments for the savings they promote—
especially for the state-run colleges and schools. In this example, the
Texas State Energy Conservation Office (SECO) helped the San Jacinto
College District in East Harris County, Texas reduce its cooling load
expenses by more than $1.5 million over seven years through the
installation of a chilled-water TES unit at each of the three facilities. The
original plan was devised in 1991, with the first unit installed in 1994 and
the other two units following soon thereafter. The three TES systems at
San Jacinto generate a total of 1.7 million gallons of chilled water during
off-peak electric-use hours which is then stored until it is needed to cool
facilities during peak hours. The $2.7-million loan for the entire project
was financed through SECO’s Texas LoanSTAR program, which allows
borrowers to repay loans with money saved by the energy retrofits.
Projects in this program usually are expected to produce enough
savings to pay the loan back within eight years. Also benefiting from the
reduction in peak demand, Reliant Energy gave the college a $500,000
rebate for installing the unit (for load repayment).
Example—Villa Julie College, Baltimore, MD.64 The real value

of a TES unit often is realized through its variety of benefits. In this
example, Villa Julie College was adding more than 135,000 square feet of
space to its campus in 2000, effectively doubling the size of the college,
which is located outside of Baltimore, Maryland. One of the primary
objectives of the new facility’s $12-million design was energy efficiency.
To support the decision for the inclusion of TES units, Baltimore Gas and
Electric provided a rebate of $52,400 toward the units’ installation. For
these reasons, the college chose Baltimore AirCoil Company’s Modular
ICE CHILLER® thermal storage units to be part of the new installation.
158
In this system, two 300-ton chillers supply ethylene glycol at 19.2ºF from
10:30 p.m. to 6:30 a.m. to build ice in the storage units. During the day,
only one of the chillers and the ice storage system are then required
to support the cooling load of the new facilities. Besides enhancing the
energy efficiency of the new construction, the TES units were essential
because the building’s design called for a much-reduced internal space
allotment for mechanical equipment. The TES units worked well with the
smaller ductwork and piping and did not affect the internal aesthetics
of the design. The results of this new system have been impressive. By
shifting 262 kW of on-peak demand, the units have provided a $44,700
annual cost savings, and over the expected 20-year life of the units, this
system is expected to provide more than $460,000 in value compared
to a conventional cooling system. Plans also exist for further savings by
integrating the two cooling systems on campus. Currently, two chillers
meet the cooling demand of the existing buildings. When the two
systems are finally connected, the cooling requirements of the campus
will be met by the ice storage system and by operating only the two
new chillers.

TES is a mature technology that provides real potential saving for
customers—especially those with large cooling loads—as either a retrofit
or in new construction, because central chillers are used in approximately
80% of buildings greater than 200,000 square feet. Continued develop-
ment will help improve the economics of the technology, but the real
strength in the argument for using TES facilities is the growing body
of evidence for successful installations of these units, which can more
directly supplant existing hesitation toward these systems. Another
important aspect for the continued deployment of this technology is
the ongoing utility support because of the savings potential with the
continued installation of these units. Shifting peak demand load has
long proved to be less expensive than building new power facilities
159
or ensuring power from independent generators is available. Striking

closer to home for these utilities is the ability to slow or reduce peak
demand on a particular distribution line and thus prevent the need of
an upgrade for some time. Both of these benefits point toward utili-
ties’ continued support through small but important up-front grants to
customers to defray the installation costs of these units. As these grants
are also approved through state PUCs, continued government support
is also important. For these PUCS, their continued support stems from
not only the lower end-user costs these systems can provide, but also
their support in reducing peak-power demand, which helps to lower the
environmental impact required to meet the peak load.
Challenges for these systems will continue to be present, however, as

their benefits are misunderstood to only be long-term and, as with any
piece of capital equipment, they require some up-front capital outlays.
Continued changes and uncertainty in the makeup of retail energy
market rules and their resulting retail consumer price structure are also of
concern—not for how they will affect the economics of the installations,
but how they affect the perception of future retail prices. Finally, other
building efficiency improvements such as improved lighting, windows,
and the like will continue to compete for any available funding in the
overall conservation strategies to reduce the cooling load costs.
Major developers
There are a number of manufacturers of TES systems based on vary-
ing technologies. Some of the larger manufacturers and installers of these
systems include Baltimore AirCoil Company, Calmac Manufacturing
Corporation, Cryogel, Dunham-Bush, Inc., FAFCO, Inc., Evapco, Henry
Vogt Machine Company, and Paul Mueller Company.
160
References
1. Ogelthorpe Power. Rocky Mountain Pumped Storage Hydro Plant.
http://www.opc.com/opccom/power/rocky.jsp?menu=power,
http://www.opc.com/ (accessed May 1, 2005).
2. Dinorwig and Electric Mountain, First Hydro Company.
http://www.fhc.co.uk/DIN.htm, http://www.fhc.co.uk/INTRO.htm
(accessed May 1, 2005).
3. Japan Commission on Large Dams. Seawater pumped-storage
power plant. http://www.jcold.or.jp/Eng/Seawater/Seawater
(accessed May 1, 2005).
4. Callahan, T., J. Degnan, and D. Miller. 2003. Upgrading the Taum
Sauk pumped storage project. Paper presented at Waterpower XIII
meeting, Buffalo, New York: HCI Publications.
5. Electric Power Research Institute (EPRI). 2003. Compressed air energy
storage. EPRI-DOE handbook of energy storage for transmission and
distribution applications, 15-1–15-42. Palo Alto, CA: Electric Power
Research Institute.
6. Crotogino, F., K. Mohmeyer, and R. Scharf. 2001. Huntorf CAES: More
than 20 years of successful operation. Paper presented at American
Society of Mechanical Engineers, Spring 2001 Meeting, Orlando, FL.
7. van der Linden, S. 2002. CAES for today’s market. Paper presented
at Electrical Energy Storage Applications and Technologies
meeting, San Francisco.
8. Iowa Association of Municipal Utilities. February 2003. Transforming
wind power into a reliable resource (stand-alone 2-page handout).
9. CAES Development Company. 2002. Norton Energy Storage Project.
(Presentation).
10. Electric Power Research Institute. 2003. Vanadium redox batteries.
EPRI-DOE Handbook, 10-1–10-27; Lotspeich, C. 2002. A comparative
assessment of flow battery technologies. Paper presented at the
Electrical Energy Storage—Applications and Technology (EESAT)
2002 Conference, San Francisco, CA.
161
11. Electric Power Research Institute. 2003. Zinc bromine batteries.

EPRI-DOE handbook, 9-1–9-26; Lotspeich, C. A comparative
assessment of flow battery technologies.
12. Lotspeich, C. A comparative assessment of flow battery technologies;
Electric Power Research Institute. 2003. Polysulfide bromide batteries.
EPRI-DOE handbook, 11-1–11-25.
13. Electric Power Research Institute. 2003. Vanadium redox batteries.
EPRI-DOE handbook, 10-1–10-27; 9-1– 9-26; 11-1–11-25.
14. Clarke, S. 2004. Observations on building a flow battery company.
Paper presented at the 2004 Electrostatics Society of America (ESA)
Conference, Columbus, OH.
15. Lotspeich. A comparative assessment of flow battery technologies.
16. Sumitomo Electric International. User: Tottori Sanyo Electric Co., Ltd.
http://www.sei.co.jp/redox/e/index.html (accessed October 2004).
17. Blackaby, N. 2002. VRB technology comes to the fore. Power
Engineering International 10, no. 3.
18. Williams, B. 2004. A case study on the use of VRB energy storage
system (VRB-ESS) as a utility network planning alternative. Paper
presented at the 14th Annual Electricity Storage Association
Conference, Columbus, OH.
19. ZBB Energy. Project status-United Energy 400 kWh system.
http://www.zbbenergy.com/status.htm (accessed September 2004).
20. ZBB Energy. Project status-Detroit Edison/Department of Energy
and SNL 400 kWh system. http://www.zbbenergy.com/status.htm
(accessed September 2004).
21. ZBB Energy. Project status-Australian Inland Energy 500 kWh system.
http://www.zbbenergy.com/status.htm (accessed September 2004).
22. Regenesys Technologies. Project status-Little Barford Power Station.
http://www.regenesys.com (now offline) (accessed November 2003).
23. Regenesys Technologies. Project status-Tennessee Valley Authority.
http://www.regenesys.com (now offline) (accessed November 2003).
24. Electric Power Research Institute. 2003. Sodium sulfur batteries.
162
25. Ibid.
26. Ibid.
27. Hyogo, T. 2003. Commercial deployment of the NAS battery
in Japan, Takayama, Tokyo. Paper presented at the 2003 EESAT
Conference, San Francisco, CA.
28. Baba, Y. 2004. Electricity storage applications for electric utility
business. Kyushu Electric Power Co., Inc., Paper presented at the 2004
ESA Annual Meeting, Columbus, OH.
29. Nichols, D., B. Tamyurek, and H. Vollkommer. 2003. Sodium
sulfur battery (NAS) applications. Paper presented at IEEE Power
Engineering Society meeting, Toronto, Ontario, Canada.
30. Electric Power Research Institute. 2003. Lead-acid batteries. EPRI-DOE
handbook, 6-1–6-51.
31. World lead-acid battery markets. 2002. San Jose, CA: Frost & Sullivan.
32. Linden, D., and T. Reddy, eds. 2002. Lead-acid batteries. Handbook of
batteries, 3rd ed., 23.1–23.88. New York: McGraw-Hill.
33. Electric Power Research Institute. 2003. Lead-acid batteries. EPRI-DOE
handbook, 6-1–6-51.
34. Ibid.
35. Ibid.
36. Faber De Anda, M., and J. Boyes. 1999. Lessons learned from the Puerto
Rico battery energy storage system. SAND1999–2232. Washington DC:
U.S. Department of Energy.
37. Sandia National Laboratories. 2000. Energy 100 awards: Metlakatla
energy storage system. Sandia, NM: Sandia National Laboratories.
38. World lead-acid battery markets. Frost & Sullivan.
39. Linden and Reddy. Industrial and aerospace nickel-cadmium batteries.
Handbook of batteries, 26.1–26.29.
40. Electric Power Research Institute. 2003. Nickel-cadmium and other
nickel electrode batteries. EPRI-DOE handbook, 7-1–7-40.
41. Ibid.
42. Ibid.
163
43. Golden Valley battery energy storage system comes online.

September 2003. Electricity Storage Association Newsletter.
44. ALCAD. Lethabo power station switches to nickel-cadmium batteries
for back-up power for critical systems. http://www.alcad.com/about_
news_detail.asp?PressID=21 (accessed October 2004).
45. Electric Power Research Institute. 2003. Nickel-cadmium and other
nickel electrode batteries. EPRI-DOE handbook, 7-1–7-40.
46. Active Power. 2001. Flywheel energy storage for quality power in the
semiconductor production industry. brochure. Austin, TX.
47. Urenco Power Technologies. Case Study—Lyon Metro. http://www.
uptenergy.com/eng/applications/traction/casestudy/lyon.htm
(accessed May 2005).
48. World UPS markets. Frost & Sullivan.
49. O’Brien, D. 2001. EC capacitors deliver high capacitance in a small
size. Power Electronics Technology 27, no. 3.
50. Electric Power Research Institute. 2003. Electrochemical capacitors.
51. Ibid.
52. Ibid.
53. Schneuwly, A., J. Auer, and G. Sartorelli. June 2004. More power for
the rails. Power Systems Designs Europe 1, no. 5: 26–27. AGS Media
Group, Laboe, Germany.
54. Electric Power Research Institute. 2003. Superconducting magnetic
energy storage. EPRI-DOE handbook, 12-1–12-3.
55. American Superconductor Case Study. 1999. Wisconsin Public Service
Corporation. brochure. Middletown, WI.
56. American Superconductor Case Study. 1998. Sappi Paper Mill, Stanger,
South Africa. brochure. Middletown, WI.
57. California Energy Commission. 1996. Source energy and environmental
impacts of thermal energy storage. California Energy Commission,
Report #500-95-005. http://www.energy.ca.gov/reports/reports_
500.html (accessed October, 2004).
164
58. Thermal energy storage—Economics and benefits. 2002. E3 Energy

Services, LLC., Arlington, VA.
59. Silvetti, B., and MacCracken. 1998. Thermal storage and deregulation.
ASHRAE Journal 40 no. 4.
60. MacCracken, M. 2003. Thermal energy storage myths. ASHRAE
Journal 45 (9):36–42.
61. Electric Power Research Institute. July 1991. Cold air distribution with
ice storage. Brochure CU-2038. Palo Alto, CA, EPRI.
62. Calmac Manufacturing Corporation. November 2001. Hospital
market brochure (Englewood, NJ).
63. Jones, D. January 2000. Hoarding hertz. Texas: Fiscal Notes, Texas
Comptroller of Public Accounts, State of Texas, Austin, TX.
64. Baltimore AirCoil Company. 2000. Ice thermal storage—PRJ42.
Baltimore, MD. http://www.baltimoreaircoil.com/english/products/
ice/tsum/index.html (accessed May 2005).
4 APPLICATIONS
Energy storage technologies will have far-reaching impacts

throughout the electrical power industry. They can conceivably provide
benefits (fig. 4–1) in all three market segments: wholesale power,
transmission and distribution, and the retail market. Within each market
there are similar needs: to improve asset usage, increase flexibility
and optionality, and prevent power variations from interrupting
operations or damaging equipment. Storage technologies provide an
extra dimension to make these improvements—time. Because they
bring flexibility, energy storage technologies are useful in many market
roles. They are best thought of as enabling technologies, either raising
system optimization or promoting a market change, such as the faster
introduction of renewable energy resources. Because they impact the
market’s margin when costs are highest and/or stress on the system is
at a maximum, even a relatively modest action—especially a fast-acting
one—could dramatically impact the market.
166
Fig. 4–1. Energy storage applications (Ardour Capital Investments).
However, no energy storage technology is suitable for all market

roles. Although most storage technologies could be used on innumerable
applications, any energy storage technology is best suited for only a
few related applications (fig. 4–2), where its technical capabilities can
be leveraged for maximum economic benefit. This matching of the
technology to the application is balanced between technology capability
and application requirements. These roles are often grouped (fig. 4–3)
according to how long the facility is expected to discharge power quality,
bridging power, or energy management. Generally, short-duration storage
roles are focused on damage- or loss-prevention, whereas long-duration
storage roles are focused on commodity arbitrage. Because they can
operate across a wide spectrum of uses, a technology can be optimized
for one of the following applications above another by altering such
attributes as the capacity of the storage medium, the conversion
capability, and the like:
APPLICATIONS
167
Fig. 4–2. Application capability sizing requirements (Courtesy of the

Electricity Storage Association).
Fig. 4–3. Energy storage market roles (Courtesy of the Electricity

Storage Association).
168
• Short term—power quality: Prevents poor power quality

from damaging sensitive electronics or other electrical
equipment from unpredictable and momentary changes
(milliseconds to seconds) in the frequency, voltage, and so
forth. Depending on the facility’s size, these systems can
provide voltage and frequency control (wholesale power/
transmission market) or power quality (retail market).
• Medium term—bridging power: Assures a continuity of
service (seconds or minutes in length) when switching from a
primary generation source (utility service) to another (backup
generator/alternate power feed). Depending on the size of
the facility, these systems can act as contingency reserves
(wholesale power market) or as a UPS (retail market).
• Long term—energy management: Allows arbitraging
between two market periods (hours) by decoupling the timing
of generation and electricity consumption, or load leveling.
Depending on the facility size, these systems can be used for
bulk power sales (commodity arbitrage), minimizing capital
outlays (transmission asset deferral), or reducing individual firm
energy costs (peak shaving).
Key Design Issues

As with any engineered system, the final design of every energy
storage facility is essentially a compromise between attributes to
optimize the facility for a particular application. There are three key issues
in particular: energy versus power, cycling, and usage costs. Adjusting
one part (i.e., greater power deliverability) frequently impacts the overall
design and affects other attributes, most prominently the economics of
APPLICATIONS
169
providing the service for a particular application. For instance, power

quality requires an extremely fast reaction time to be of any value to
the consumer, with this requiring not only a high deliverability from the
energy storage medium, but also from the supporting equipment, such
as power electronics. Thus the importance of each capability is evaluated
and balanced with the application in mind, along with the incurred cost
of promoting one capability over the other. Other factors outside the
direct operation of the unit can also have impacts on the design, even
acting as a limiting factor, such as a physical footprint (limitations) or
environment control requirements (space conditioning).
Energy versus power

The first aspect of an energy storage technology is the unit’s capability
to store and release energy at different power ratings. Outwardly, these
aspects of the technology are responsible for determining the market role
of the technology. For instance, the energy rating (kWh) is sometimes
thought of as the volume or scale of the facility and is usually the prime
determinant in how long a unit can operate. The power rating (kW)
of an energy storage facility describes the rate at which it can absorb
and discharge energy, and largely establishes in what part of the market
the unit can operate—MW-sized units for the wholesale power market
versus kW-scale units that are more suited for on-site installation at
commercial or industrial facilities.
Inwardly, these application requirements for a specific size and

capability will define both the type of technology capable of fulfilling the
market role and the geometry of the storage facility. For instance, the ratio
of energy to power is largely fixed for chemical batteries, whereas other
technologies have separated these two, providing a degree of freedom in
system design. Therefore, fitting a chemical battery to a particular market
170
role requires adjusting the number and length of battery strings to get the
requisite power and energy rating. For those technologies (such as flywheels
or flow batteries) where these aspects are separate, fitting a technology to
a market role entails scaling each aspect individually to what is needed. For
instance, changing the power rating is normally accomplished by altering
the conversion component (motor, cell stack, etc.) to improve the rate of
power transfer of the storage medium. To increase the energy capacity,
one would simply add (or enlarge) storage modules or increase the energy
density of the storage medium. Although adding additional storage
modules is generally the easier and cheaper option, outside issues such as
space constraint can limit this choice.
Cycling issues
The second issue concerning the design of an energy storage
technology is its cycling capability: the unit’s ability to be repeatedly
charged and discharged. All storage facilities have a finite useful life
based on the number of times the unit is used; this plays directly on
the applications they can support. This cycle life also varies depending
on the physical attributes of the storage medium and how the energy is
stored. For example, chemical batteries generally have a shorter cycle life
(hundreds or a few thousands) than a mechanically based flywheel unit
that can sustain hundreds of thousands of cycling events. This is the case
as the chemical battery frequently has a small but cumulative chemical
by-product during each cycling event, whereas the flywheel suffers little
or no residual stress to the rotor. For these higher-cycle-life technologies,
often the limiting factor is not the storage medium but its associated
equipment. As these physical characteristics affect the cycle life of the
technology, the cycle life itself then plays a central role in the usage cost
of the system (fig. 4–4).
APPLICATIONS
171
Figure 4–4. Cycling issues (Courtesy of the Electricity Storage

Association).
Many factors affect the design cycle life of these technologies; three
crucial factors are depth of discharge (DOD), the rate of discharge, and
the environmental conditions in which the unit operates. Probably the
greatest determining factor affecting the cycle life of storage technology
is the DOD to which the unit is repeatedly cycled. Again, one can see
differences between the technologies as some are electrical or mechanical
(capacitors and flywheels) and are able to discharge almost completely
without much effect, whereas the chemically based (lead-acid batteries)
suffer significant deteriorations in their cycle life from full discharges.
Below an 80% DOD, the cycle life of many of these chemical-battery
systems deteriorates dramatically; for this reason, most comparisons
between technologies are based on an 80% DOD. Second, forced, rapid
recharging can also damage the storage medium and lower the cycle
life of the unit or even its ability to charge. This is of greater concern
to many of the chemically based storage technologies but can easily be
172
avoided by limiting the rate of flow through the accompanying power

electronics. Finally, environmental conditions can also negatively affect
a storage medium’s cycle life, especially chemically based technologies.
These impacts also can be avoided or greatly reduced through the use
of space conditioning, but this adds to the parasitic load of the facility,
lowering the unit’s round-trip storage efficiency.
The efficiency in each charge/discharge cycle, the round-trip efficiency,

is also an important aspect of the cycling operation of the unit. These
losses stem from resistance, heat, and so forth, as well as in the associated
power electronics (although some comparisons of the technologies do
not include this component). This round-trip efficiency determines to
a large degree with what type of market application the facility can
compete. Low round-trip efficiencies effectively increase the average cost
of operating the unit, for example, so such a technology would be better
suited for a power-quality role above commodity arbitrage—because
the value of power quality is to prevent a catastrophic loss, whereas
arbitraging power is based on the marginal difference in the price of
a given amount of power. Other losses (float) occur while the unit is
charged but inactive, and play a role in the economics of the unit to
provide a competitive service for an application requiring long-term
storage of energy.
Ownership costs
Although technology capabilities are important, for most potential
owners the cost is the real deciding factor. At the time of purchase, this
would consist of the up-front capital costs, but the total ownership cost
should actually be used for any decision-making purposes. Different
technologies have different operational life spans and operational
economics that will greatly affect this calculation (fig. 4–5). The structure
of these costs is important, especially when comparing different storage
technologies. For some technologies the operational cost can end
up an appreciable part of the total ownership costs. Although some
APPLICATIONS
173
technologies may be cheaper initially, the requirements of the application

make the real usage cost far higher than a competing one, such as the
technology’s cycle life or round-trip efficiency. If purchasing a UPS unit,
for example, simple lead-acid batteries often have the least initial cost. If
the total ownership costs of the unit are evaluated, however, especially in
a hostile environment (concerning power quality), flywheel systems are
frequently the cheaper option.
Fig. 4–5. Ownership costs (Courtesy of the Electricity Storage Association).
Capital and operating costs of a technology rely heavily on the

design of the technology. Capital costs are based on many factors,
including the technology’s complexity and the scale of the facility; for
this reason, storage technology capital costs are evaluated in dollars per
kW like other power technologies. Depending on the technology in
question, these costs can further be broken down into three subsets:
costs that scale with energy (storage capacity of the unit), costs that
scale with power (conversion module), and the balance of plant costs
(system housing, monitoring system, etc.) that often only vary slightly
with the size of the unit. Operating costs, normally measured in
174
dollars per kWh (or dollars per kWh per cycle), originate from a number
of sources, including required maintenance, consumables (for some
chemical storage mediums), and replacement part costs. Included in
these costs is the energy needed to maintain the unit in a ready but
inactive state, and the parasitic load from supporting systems such as
space conditioning. All storage mediums slowly lose slight amounts of
energy—float losses. This is one aspect that determines whether some
technologies are appropriate for long-term storage applications.
Choosing one technology over another should be done with care.

This is especially true for finding a comparative metric, which holds biases.
Any metric—such as leveled dollars per kW, dollars per kWh, and dollars
per kWh cycles—holds limitation as a comparative metric for dissimilar
technologies (pumped-hydro versus capacitors) in disparate market
applications, yet is appropriate in comparing technologies with similar
capabilities targeting the same applications. For instance, dollars per kW
is an obvious guide for up-front costs. Similarly, dollars per kWh is useful
to evaluate the cost of usable cycling energy, but dollars per kWh per
cycle costs is more useful to compare technologies for applications with
high-cycling needs, but not low-cycle applications. Therefore, the final
choice should be made based on a variety of comparisons in relation to
the application desired.
Wholesale Power
Energy storage technologies can play an important—and profitable—
role in the wholesale power market both by supporting the efficient
operation of existing power facilities and providing ancillary services
to enhance system reliability and security. These facilities can also help
promote the greater use of renewable resources (wind) and solve such
issues as its unpredictable and noncoincident power production.
By decoupling electricity supply and demand, storage units can play a
APPLICATIONS
175
role in addressing many existing problems in the wholesale power market,

such as the daily load curve, which creates two major problems for
power facilities: chronically low usage and increased cycling stresses. As
discussed previously, the current power facility fleet design and operation
was crafted around the classic baseload, mid-merit, and peaking regime,
which although it provides power in the most economical way, keeps
overall usage of many fossil-fueled generation facility assets at 60% to 70%.
Although peaking facilities may have been built to operate on a short-term
basis, coal-fired units consistently operate below their optimal level. Raising
the use of these units would not simply give them additional operating
revenue. They would also operate more economically because the stress
from cycling the units would be reduced, prolonging the replacement
cycle of many hot-zone components of these boilers. Even combined-
cycle, gas-fired facilities could benefit; many of the early units suffered far
greater and faster cycling degradation than first envisioned. With storage
units supplanting power production facilities in support roles, such as
load following or contingency reserves, existing facilities can be redirected
toward more stable power production.
This market is dominated by large power facilities; for that reason,

only large-scale energy storage facilities measuring in the 10s, or more
likely 100s, of MWs with discharge endurance of many hours will most
likely be successful. Storage technologies like existing pumped-hydro
facilities or even new compressed air energy storage (CAES) units can
then be used to act as both a sink and source of energy in ever-changing
market conditions. These larger units will cycle (inversely) on a daily
basis with the system load or act to provide a ready reserve of power in
the event of an incident. Smaller storage technologies can also operate
successfully in these roles if the market they participate in is a smaller
or isolated power grid on a much smaller scale. Like other assets in the
power market, their capability is only one determinant of their role.
Where the unit is situated will also impact greatly as to how and when
the facility will best be able to leverage its capability.
176
By acting as both a sink and source of power, energy storage facilities can
play a role in the wholesale power market along three broad applications:
1. Commodity arbitrage: Absorbs low-cost, off-peak power and

resells it during peak demand periods when its value is highest.
Besides earning revenue for the storage facility, this can provide
benefits to both underused baseload units during off-peak
periods and mid-merit units being damaged from excessive
ramping to meet rapidly changing load requirements.
2. Contingency reserves: Provides additional resources in the
event of an incident at a power or transmission facility. Besides
providing these services directly to the independent system
operator (ISO) in competition with generating facilities, these
units could also represent a callable resource during peak
periods to free up generation units held at standby.
3. Blackstart capability: Helps reenergize the power grid in
the event of a blackout. By using centrally dispatchable and
strategically placed storage units during these episodes, these
facilities could hasten the restoration process.
Commodity arbitrage
Commodity arbitrage is the act of absorbing low-cost, off-peak
power and selling it during peak demand periods when its value is
highest (fig. 4–6). Facilities with the ability to dispatch large quantities
of power on demand in a competitive market, such as pumped-hydro
units, are already highly valued in the market as evidenced by the prices
they were able to command during 1990s auctions. By definition, only
storage facilities can provide this service because other power sector
technologies cannot hold energy in a ready-reserve state. Because these
APPLICATIONS
177
units will operate in the wholesale market, the scale of these units would
range from the tens to hundreds of MWs. Depending on the need, either
the power from these storage facilities can be supplied to the market
in one full-power discharge, or the power can be delivered to the grid
with a variable output to provide desirable balancing. The frequency of
a storage facility performing this application would obviously depend on
the area’s load profile but could easily consist of multiple discharges (of a
few hours) per day. For this reason, the endurance capacity of the facility
is key to determining how often the unit can operate in the market, and
larger is usually better.
Fig. 4–6. Commodity arbitrage (Ardour Capital Investments).
The key to the profitability (and hence success) of commodity

arbitrage obviously rests with the relative difference between peak and
off-peak power prices in a region. For example, areas dominated by a
single fuel source like the Pacific Northwest (hydro) may not have
enough of a price differential in their daily cycle to provide sufficient
profitability for a storage facility alone, requiring the unit to be used
for other applications as well. Therefore, the relative mix of generation
178
prime movers in the region is a key indicator of success. Regions with

particularly high potential for success would have significant unused
baseload coal or nuclear generation assets and peak daily prices set by
natural gas peaking units. This type of dispatch and price structure would
provide significant opportunities to profit on the daily arbitrage cycle.
Three energy storage technologies have the capacity to operate

successfully in this market: pumped-hydro storage, CAES, and flow
batteries. Only flow batteries do not yet have active installations
operating in the market, although the technology has the potential to
be scaled to the required size. Market flexibility of these technologies
quickly lends them to new roles for the units. For instance, Oglethorpe
Power’s 848-MW Rocky Mountain (GA) pumped-storage facility was
intended to run only in a peaking mode, but in its first calendar year
of service (commissioned 1991), the unit ran 364 days in a variety of
roles because of its wide operational capability to take advantage of
profitable market opportunities. Similarly, Alabama Electric Cooperative’s
110-MW Macintosh (Macintosh, AL) CAES facility has also found many
uses, operating throughout the day and offering critical support during
peak power. All told, the CAES unit can store 2,600 MWh, allowing the
unit to operate at full power for up to 23 hours. With the continued
volatility and wide spread between on- and off-peak power prices,
storage technologies have many opportunities in the evolving power
market. However, their siting will remain a central concern, lending
support to additional CAES facilities. As opposed to pumped-hydro
technology, CAES has much less strenuous siting requirements, allowing
these facilities to be placed in relation to the grid’s needs rather than its
geographical realities. Finally, CAES facilities hold out the potential for
large storage reservoirs (tens of hours or more). Such reservoirs could
be filled both at night and on the weekend, extending the amount of
run-time during the week for the unit.
APPLICATIONS
179
A final benefit of arbitraging also can be seen from the coal facility’s
point of view. By acting as a comparable sink for underused power
facilities during off-peak periods, these units allow the coal facilities to
operate in a more level and less stressful and expensive manner. Because
many power facilities are required to ramp up and down on a daily
basis to match power-demand levels between on- and off-peak periods,
they currently operate inefficiently, increasing their operating costs and
environmental emissions. These additional costs can be significant for
coal units and especially for gas-fired combined cycle units.
One way to see the potential benefits of additional large-scale

storage facilities is to show the impact on a typical coal facility. This
example will only deal with annual average costs to show in a broad
stroke how large-scale energy storage can assist baseload facilities. It
should be remembered, however, that storage can have many other
comprehensive benefits improving the coal facility’s operations. In this
example, we will look into coupling a 1,000-MW coal facility operating
at a 71% usage rate with a 405-MW CAES facility (80% system efficiency
utilizing a 300-MW compression motor). Operating the compressor
for eight hours each night, five nights per week, results in more than
568,000 additional MWh of demand from the coal unit during off-peak
periods. This could increase the coal facility’s use by more than 6% (to
77%), lowering its production cost by roughly $1.40/MWh (based on
$25/MWh, a 2-to-1 fixed/variable split). Therefore, besides the additional
$6.7 million of additional power sales the coal unit could capture during
off-peak periods, the facility could receive an additional $8.5 million from
its current power sales with the new, lower production costs.
Other examples of large-scale storage facilities show that for these

units to be truly successful, they must be able to perform a single market
role as well as a variety of roles that may change over time. For instance,
180
the 290-MW E.ON CAES plant near Huntorf, Germany, has now been
operated successfully since 1978. This plant’s market roles have changed
according to the situation on the power grid, including as a blackstart
unit, a peaking unit (especially in the evening when no more pumped-
hydro capacity is available), and as a contingency reserve (because slow-
starting coal units are the primary option currently). This ability of the
facility to adapt continues, as the CAES unit is now being tasked with
compensating for sudden and unexpected wind power shortages as the
number of wind turbines increases in northern Germany.
Contingency reserves
To prevent a sudden failure of one or more power facilities or
transmission lines from impinging on system stability and security,
utilities and ISOs maintain various levels of reserve generation capacity.
Different control areas and utilities hold varying amounts of capacity in
reserve, as some target a proportion of the system load, whereas others
specify a MW level. Overall, these reserves can account for more than
1,000 MW that are reserved and prohibited from generating power.
Depending on the response time needed, these reserves are classified
as either spinning or nonspinning. Spinning reserves are the first line of
response and generally are required to equal the largest unit in the areas.
Units providing this service are online, synchronized, and unloaded but
ready to serve customer demand immediately should a need occur; all
are required to be at full power within 10 minutes. Nonspinning reserves
are similar-type units that are held at standby (unsynchronized) or units
that are able to start up quickly, with both groups required to be fully
online within 10 minutes (nonspinning reserves are simply those not
immediately available). Generating units providing these contingency
reserves are compensated through capacity payment based in part on
the market-clearing price.
APPLICATIONS
181
As with other ancillary markets, although the technical

requirements for contingency reserves have existed since the
beginning of the industry, the ISO/RTO competitive market for these
services is still evolving. As the definitions of the markets vary across
the country, these markets in total can easily represent upward
of 5% of the total energy market’s value (depending on the ISO).
Previously, utilities attempted to harden the power grid to prevent
any disruption from occurring. They accomplished this strategy
primarily by providing a significant redundancy to essential generating
and transmission capacity so that the system would always retain a
significant reserve margin. Policy leaders are adjusting that goal slightly
to focus on providing a level of survivability in the event (not in case)
of a disruptive event. Of course the utilities will continue to strive
to thwart an outage in the first place. Significantly more effort is
being placed on preventing small, inevitable, localized problems from
becoming grid-spanning events, hence the need to find new, better-
dispersed resources with greater responsiveness that can be included
for these services. The need for these units is especially acute during
peak times because units providing frequency regulation services are
not permitted to provide contingency reserves at the same time. One
increasingly popular alternative solution for supplemental reserves
is demand-response programs as they mature from utility-run,
dispatchable load–curtailment programs to competitive, opt-in ISO-
led price response programs tied to real-time prices and loads.
Energy storage technologies can act as another source of spinning-

reserve capacity (fig. 4–7); in fact, hydro and pumped-hydropower
facilities are prized because of their superior reaction capability. For
instance, if the 1,800-MW Dinorwig (Wales, UK) pumped-hydro facility
is held as spinning reserve, the entire plant can reach maximum output
in less than 16 seconds. Other existing storage technologies currently
provide similar spinning-reserve requirements; for instance the Macintosh
182
CAES unit can come online within 10 minutes. This unit can also remain
operating at only 10 MW (maximum 110 MW), extending its ability to
supply a flexible resource. With such responsiveness, additional large-
scale CAES storage facilities should also find a receptive market for their
capability. In a curious twist of fate, one of the largest proposed energy
storage facilities—Ohio’s Norton CAES plant (potentially 2,900 MW)—
has been proposed for where the 2003 blackout began in Ohio. Had
this unit been operating, its supporters suggested the facility could have
provided system operators some additional time to respond, helping to
prevent or reduce the scale of the resulting event from spreading to such
an extent. If other regions follow California to establish reserve adequacy
requirements of 15% to 17% planning reserve by 2008, this could provide
an additional need for storage facilities to meet these requirements.
Fig. 4–7. Contingency reserves (Ardour Capital Investments).
Finally, because of the storage facility’s ability to shift from charging

to discharging quickly, this effectively doubles the unit’s system-balancing
capability, making it exceedingly useful for reacting to sudden changes in
the system load and supply. Pumped-hydro facilities have been used in this
way for a long time and remain one of the seldom-realized, but significant,
APPLICATIONS
183
values of these large-scale storage facilities. CAES units can also provide
this capability, but compressors and generators are actually separate units,
and by operating independently, they provide an additional level of control
and responsiveness. Although it may seem counterintuitive to operate
the compressors while generating power, this provides the facility with
the vital flexibility to provide a number of ancillary services with a lower
response time.
Blackstart capability
If the regional contingency reserves are not sufficient to maintain
the system and a blackout occurs, the grid operators must undertake
the process of restoring power to the grid using units with blackstart
capability. These blackstart generating units are able to self-start (or are
capable of operating at reduced levels when disconnected from the
grid) after the system power is lost; these units are generally combustion
turbines and hydroelectric facilities and are the starting point for the
system’s restoration after a complete system power loss. These specialized
plants are needed because once shut down, most generation facilities
require system power to begin operation (provide auxiliary plant-load
and cranking power for the generator) and reenergize the power grid.
Besides self-starting, these units must have the capability to maintain
frequency and voltage under varying loads while the system is restored
because most of the system inertia will not be online. The entire process
can take many hours, so regional transmission organizations (RTOs) such
as the PJM require that blackstart generation units have the ability to
operate for many hours to restore the system.
The locations of these units are important. To get the main generating
units back online as quickly as possible, many are directly tied to the
blackstart units or located nearby. Others are dispersed throughout the
system at important transmission points for redundancy lest a unit or
transmission line failure block the restoration attempt. As utilities have
184
divested their generation assets, the assurance of sufficient blackstart

capability in the system is of issue because it is difficult to maintain the
capability (necessary equipment, training, and fuel), and there is always
the threat that the new owners may not continue providing this service.
Without sufficient blackstart resources, system restorations will take
much longer, which raises health, safety, and security issues. For these
and other reasons, this is not a competitive market but a cost-based
service with payment to these suppliers from all customers; there is a
benefit to all.
Larger energy-storage facilities (multi-MW, multihour endurance) can

assist with system restoration after a blackout or similar event (fig. 4–8).
No conventional generating facility is designed exclusively for this role,
nor is any storage facility. They would provide this service in addition to
their other market roles because there would not be a conflict; also, those
storage facilities providing this service would need to have sufficient
power and discharge endurance to be of value in the system restoration.
Although the location of the units would also be a crucial consideration
as to whether they would be suitable as a primary blackstart unit, once
the system collapsed, all centrally dispatchable storage facilities would be
a resource; those units in less-strategic locations or with lower output
capability would simply serve a supporting role for other blackstart units.
Therefore, although storage units would in no way displace current power
generators providing this role, additional blackstart-capable resources
could potentially hasten the system power restoration. Storage facilities
already do so in the market (even existing blackstart power generation
units rely on battery packs to provide power for their own starting).
Besides pumped-hydro facilities, the 290-MW Huntorf, Germany, CAES
facility has provided a blackstart capability to some nearby nuclear
reactors (besides its other market roles) since the late 1970s. In addition,
the recently cancelled Regenesys 15-MW (eight-hour) sodium bromide
flow battery was to provide blackstart capability for Innogy’s 680-MW
Little Barford Power Station in Cambridgeshire, UK, besides providing
frequency and voltage regulation for the area.
APPLICATIONS
185
Fig. 4–8 Blackstart (Ardour Capital Investments).
Transmission and
Distribution
Energy storage technologies have the opportunity to play a
significant role in the transmission and distribution (T&D) market
by enhancing the reliability and stability of normal operations and
increasing their efficiency by reducing or deferring capital investments.
A number of market trends have conspired to make the management
of the power grid more difficult in recent years, as well as an area of
growing concern for the future. Most recognize that the essential billions
of dollars to upgrade the power grid are not simply for additional power
lines but for power electronics and other supporting equipment, such as
storage technologies, that provide rapid and self-correcting capabilities.
Generating capacity expansion is also increasingly oriented toward
incremental merchant mid-merit and peaking facilities or distributed
resources instead of baseload power facilities, thus requiring far more
actionable control capability to be embedded into the power grid to
186
balance the increasing variations and sources of power. With merchant

power providing much of the new capacity, however, there is concern
that planning the location of these new units is increasingly optimized
more for profitable power production (fuel, land, etc.) than grid stability.
Storage systems placed at strategic locations could act in concert with
flexible AC transmission systems (FACTS) equipment to provide stability
and enhance the power grid’s operation by providing additional flexibility
in operations and planning. In addition, small mobile storage units could
offer significant relief until a more permanent solution is developed.
Because their role is to enhance system stability, energy storage

facilities designed to support transmission and distribution networks
are more concerned with the ability to swiftly inject energy rather
than replacing the power generators’ role of providing bulk energy to
the grid. Consequently some storage facilities could have a relatively
small storage capacity while still having a major impact because some
roles call for repeated cycling of power in and out of the unit quickly.
In the transmission network, large-scale storage facilities (MWs) can
provide flexible grid frequency regulation to widespread areas. In the
distribution market, however, smaller but faster responding storage
units (MWs-kWs) can provide targeted frequency regulation and
voltage stability (from their power electronics) as rapidly changing
conditions create localized areas of instability. For both markets, the
relief provided by the storage at or beyond constrained points in the
power grid can reduce or delay required capital investments needed to
serve the growing load in a reliable manner.
By acting as both a sink and source of power, energy storage facilities

can play a role in the transmission and distribution network along three
broad applications:
• Frequency regulation. Frequency regulation acts as a

stabilizing force to balance supply and demand as the system’s
load fluctuates. If these changes are not compensated for,
APPLICATIONS
187
damage may occur to both the power facilities and the

customers’ electrical equipment. Required reaction times span
from seconds-level corrections for sudden changes to minutes-
long responses for changes in the daily load.
• Voltage regulation. Voltage regulation is needed to maintain
power flows across the power grid and for the operation
of customer power equipment. Reactive power, the force
controlling voltage levels, is produced by power facilities and
specialized equipment placed throughout the network. Poor
voltage control results in additional power loss, especially at
the end of a long distribution power line.
• T&D asset deferral. Although the infrastructure of the
power grid must be constantly upgraded to address the
always increasing system load, traditional means of doing so
are inefficient, leaving significant capital assets underused.
Providing additional resources—both commodity and grid
support—in areas with limited transfer capability can help
the system ride through peak demand periods, increasing the
usage of the network and allowing for a postponement of the
required system upgrade.
Frequency regulation
Frequency regulation stabilizes the power grid by managing the
moment-to-moment changes in the demand or supply balance of
the power grid (fig. 4–9). As load changes, excess generation causes a
frequency increase above 60 Hz; insufficient generation causes a decline.
Small shifts in frequency (load) do not degrade reliability, but large ones
can damage equipment, degrade system efficiency, or even lead to a
system collapse.
188
Fig. 4–9. Frequency regulation (Courtesy of Brendan Kirby, ORNL &

Consultant).
These changes are first counteracted by the rotational inertia inherent

in the connected synchronous generators. As the variation continues,
regulation can also be provided through generating units operating under
automatic generator control or participating in manual frequency control,
both of which can change output quickly (on the order of MWs per
minute). Automatic generator governors restore frequency if it deviates
more than 0.036 Hz from normal. If the variance in frequency continues
to grow, the power facility may lose its synchronization from the grid
at ±1.5 Hz to prevent significant equipment damage, which can begin
to occur at ±3.0 Hz. Besides the obvious costs of damaged equipment,
variations in power output can also produce raised operating costs in
plants constantly altering output, plus the increased operation and
maintenance (O&M) costs from the added wear on the unit, which also
reduces its operating life. Frequency regulation is often a greater issue on
small or isolated utility systems, because system inertia on a larger power
grid provides more time for generator governors to respond. Having less
system inertia, the frequency will fall faster on a small grid system in the
event of power supply imbalances. Therefore, these isolated power grids
APPLICATIONS
189
must be capable of a far more rapid response to frequency deviations,

or the frequency change will begin to affect equipment throughout the
system more rapidly than on a larger, integrated system. For some smaller
utility systems, regulation is combined with spinning reserves to make
the most of scant excess capacity.
Load following is similar to regulation but on a longer time horizon.

Rather than focusing on the minute-to-minute load variations required of
regulation, load following provides energy correction on an hourly basis
for any mismatch between energy supply and the system load. Although
useful throughout the day, this service is obviously more valuable during
the morning pickup and evening drop-off, when the greatest changes
occur. As the load for the area changes, the system operator must bring
on more generation to meet that rising demand, which can change
10% per hour. These power facilities are forced to ramp up generation
according to the demand level.
Frequency regulation services, as well as load following, are

currently produced by power facilities. For example, PJM currently
designates 1.1% of the projected peak capacity to provide regulation
and load following for the system. RTOs like the PJM have developed
the service into a competitive market so anyone could bid in to provide
the service; however, frequency regulation is a technical requirement of
the system, so all utilities need a means to provide it. In other areas
not under an ISO/RTO structure, the utility provides this capability to
itself, but as it is rolled into operations, breaking out the cost can be
difficult—even for the utility.
Some energy storage facilities are well suited to provide frequency

regulation; in fact some large-scale storage facilities are already key
providers of these services. Pumped-hydro and CAES facilities have
provided these capabilities for years (decades) and are sometimes even
the primary providers of regulation for the entire utility. These facilities
190
counteract the frequency variations by increasing their discharge to meet

a negative imbalance in the system load and decreasing their discharge
to meet a positive imbalance in the system load, similar to how power
generation facilities currently provide the service. Many of these storage
units can also provide regulation services during their charging episodes
by reversing the previous actions. Because of the large scale of these
facilities, these units can also provide longer charges and discharges for
load-following duties.
Other storage technologies are able to provide frequency regulation

in multifunctional facilities. An early example was the Puerto Rico Electric
Power Authority’s 20-MW, 14-MWh lead-acid battery storage plant at its
Sabana Llana substation, which was installed in 1994. This $21-million
facility provided spinning reserves, frequency regulation, and voltage
control services for the island. These capabilities proved extremely useful
during the fall of 1998 in the aftermath of Hurricane George, when
unreliable power supplies and transmission reliability caused a number
of load-shedding events and increased spinning reserve requirements—
requirements that the battery facility reduced during this time of scant
resources. A more recent example is the recently inaugurated 46-MW
NiCd battery facility for the Golden Valley Electric Association (Fairbanks,
AK). This unit supports the radial makeup of the Golden Valley’s power
grid, providing frequency regulation and lowering the larger generator
spinning reserve requirements common among Alaskan utilities because
of their nonconnected status. These energy storage units provide
frequency regulation by repetitively cycling from charging to discharging
in order to balance a mismatched system load. This is a different tactic
than existing generation facilities that provide this balancing role through
a constant change in output level. Because these smaller units are not
sized to the storage capacity of these other units, these smaller storage
facilities must have fast reaction capability and a high cycling limit to
effect a similar result. To maximize the effectiveness of the storage facility,
the facility should also have as deep a discharge capability as possible.
APPLICATIONS
191
Although these multifunctional storage facilities have been useful

(especially in small power grids), providing frequency regulation requires
a high cycling capability of the storage facility, limiting the number of
storage technologies actually available to provide this service. By nature
frequency regulation is comprised of essentially a constant need to
charge or discharge the storage unit, and across the 20-year life of a
storage facility, providing the service can easily require more than 500,000
full charge/discharge cycles (based on a 15-minute cycle) if constantly
engaged. Power generation facilities of many types currently provide this
service, but their performance is much slower than optimal because their
response time is not capable of reacting to all of the constant changes;
and attempting to match what is actually needed imposes a life penalty
from additional wear (in fact, the current level of operation does as
well). Therefore, these units’ control schedules are purposefully slowed
to reduce the life impact at the expense of performance. One solution
to provide a much-needed prompt-response capability for regulation
service is the proposed 1-MW, 250-kWh Beacon Power Smart Energy
Matrix (SEM) grid stability facility, which is designed to operate purely in
the regulation role. Matt Lazarewicz, vice president and chief technology
officer of Beacon Power, notes in a personal correspondence, “Beacon
Power’s SEM is particularly well suited for frequency regulation, and is
much more effective than generators used today. The SEM is purposely
designed for fast response (milliseconds versus minutes for generators)
with high reliability, much lower operating costs, and operates without
any emissions.” The flywheels in the unit possess many advantages for this
type of operation; they have the required life cycle, deep discharge, and
fast response rate needed for this role. Because the system is designed
only for this high-cycle role, its performance and operating cost can be
much better than that of a generator for this role. Finally, because this
unit is enclosed completely in a typical container shell, the unit is mobile
and able to be reassigned to different areas of the grid as warranted when
conditions change.
192
Voltage regulation
Maintaining and regulating the system voltage level is important
for efficient transmission of power and customer requirements. Low
voltage conditions arise on power systems from two main sources:
highly loaded power lines and long, unsupported distribution lines.
Uncompensated for, both of these situations can negatively impact
the amount of power able to be transmitted through the power line.
As with frequency regulation instabilities, swings in customer power
use impact system voltage, particularly with the use of heavy motive
equipment. Unfortunately customer requirements for voltage stability
are increasing, in part because of the increasing penetration of electrical
equipment into the commercial and industrial setting. According to EPRI,
most distribution-system problems are voltage sags of 10% to 30% below
nominal, extending from 3 to 30 cycles in duration.1 These types of
disturbances are responsible for causing motor-controlled manufacturing
lines to shut down or trip off-line. Although some consumer electrical
equipment can operate within a 10% range of the rated voltage, much
electronic control equipment or information technology produced
now has a significantly lower limit (5% or less) before the equipment
is affected or damaged. With continual distribution-system expansion,
maintaining sufficient voltage control will remain a challenge; ongoing
budget constraints limit the ability of the utility to install all the needed
voltage control equipment.
To control voltage levels, reactive power is produced and managed

throughout the system. Reactive power, measured in mega volt ampere
reactive (MVAR), helps control voltage levels and is required to power
consumer electrical equipment as electric motors and the like for creating
magnetic fields. Reactive power primarily is produced at power facilities
through voltage regulators, but it is also controlled with specialized
equipment (capacitors/inductors, Static VAR Compensators [SVC],
transformers, etc.) spaced throughout the network that adjusts the
voltage level as needed. Spacing the equipment throughout the network
is one way the industry currently manages the reactive power levels
APPLICATIONS
193
to prevent small incidents from growing into larger, area-wide events.

Unfortunately, by producing reactive power, the generating facility lowers
its ability to produce real power for sales—crucial during peaks when
reactive power demands rise along with the need for power.
The reactive power needs of the system operate dynamically

with the use of the transmission system; if they are not addressed, the
imbalance in reactive power levels will negatively affect the carrying
capacity of the power line. When a transmission line has only a low load,
the capacitive effect dominates, requiring absorption of reactive power.
Conversely, at high loading the inductive effect dominates, requiring the
production of reactive power. This relationship is nonlinear, so periods
when the power line is fully loaded require significantly more reactive
power injections to prevent a voltage loss than when it is only at an
average load. There are also problematic, significant drops in voltage at
the end of long distribution lines, especially during high-load periods.
Both of these situations require the utility to inject reactive power during
high-load periods. With improved carrying capacity at these congested
points, the entire carrying capacity of the power grid can be improved.
Fast-reacting energy-storage technologies are well suited to

counteract the variations and instabilities in the voltage level that occur
every day in the power grid. This is accomplished by providing direct
reactive power from their associated power electronics and real power
(because of the dynamic relationship between reactive and real power).
Energy storage facilities designed to provide support for the grid are
concerned with quickly injecting stability into the grid. They are therefore
designed with a smaller endurance capacity but contain a faster reaction
time and deliverability capability instead of long-term endurance
(fig. 4–10). Similar to other voltage control equipment, these facilities are
not envisioned as replacements of power facilities (the main source of
reactive power) but as distributed providers of reactive power to correct
a momentary imbalance in localized areas of voltage instability. This role
is especially critical during periods of peak power demand when voltage
instability will be worse in the remote areas of the power grid.
194
Fig. 4–10. Voltage regulation (Ardour Capital Investments).
Early examples of energy storage technologies providing voltage

stability were limited to a handful of battery-based, multifunctional
systems such as Puerto Rico’s Sabana Llana substation facility, and PG&E’s
substation facility in Chino, California. More recently other storage
technologies with faster reaction times and a far higher cycle life have
become more common and are expected to become more widespread
as utilities often simply want a more direct alternative to voltage control
equipment. Another aspect of these alternative energy storage solutions
is that they are trailer-mounted mobile units. This can be an important
attribute to the unit; areas of voltage instability normally grow in areas
previously not suffering from these problems because of load growth
and usage change by consumers. With its significant experience with
energy storage technology, Detroit Edison installed a demonstration
zinc-bromide-based, trailer-mounted 200-kW, 400-kWh flow battery
from ZBB Energy Corp. The unit was able to provide relief for a number
of load management roles, including frequent voltage sags. Because the
system was truck mounted, it was extremely mobile, allowing for the
quick redeployment of the system to other locations.
APPLICATIONS
195
Raising the usage and reliability of an existing transmission network

through improved voltage control equipment is frequently a more cost-
and operational-effective option for expanding the carrying capacity
of the transmission line. Besides such experience at the Wisconsin
Public Service installation of the American Superconductor D-SMES
(Distributed Superconducting Magnetic Energy Storage) units, Entergy
has continued to operate a number of similar D-SMES units since 2002
in different areas throughout the Houston, Texas area. These units were
installed in substations to solve voltage-related problems and to increase
the power-transfer capability for critical transfer paths in congested
areas, improving the carrying capacity of the entire area. The need for
these units is driven by seasonal peaking loads from increasing summer
cooling requirements, and it was a cheaper solution than siting additional
generating capacity—the other option considered.
Transmission and distribution

asset deferral
Providing utilities a means to better control the timing of their
distribution-system upgrades would be a boon to firms struggling to
provide increasing reliability with diminishing resources. One strategy
is to selectively defer a portion of these required upgrades—especially
in the lower-voltage transmission and distribution where existing
problems from underinvestment increase in incidence. By providing an
additional five or more years of usability to the existing transmission
equipment, scant resources would be available for other purposes; but
more importantly, it would allow the system upgrades to be planned
around utility’s schedule, not the other way around. The need for finding
alternative solutions to this problem is apparent because the drives for
them—continued demand growth and aging equipment—show no
likelihood of abatement; in fact, these point to even greater demand for
alternatives. Based on data from 1999, upwards of $12.7 billion are added
to the capital infrastructure base each year: $2.3 billion in the transmission
196
market and $10.4 billion for the distribution market.2 This is needed to
keep up with demand; over the last 10 years, demand growth in the
United States has increased 20%, including a shift toward higher reliability.
Meanwhile, the existing equipment continues to age, with much of the
existing equipment approaching or surpassing its design life.
With this growing demand, distribution lines and their associated

substation equipment require continued upgrades. Here again the
reason is for reliability, but long-run economics dictate how the upgrades
are completed. Because upgrading power lines entails expensive
construction costs, the conventional approach for utilities is to upgrade
only when a particular power line nears its transmission capacity more
than a few days per month. Unfortunately, demands on these lines are
often extremely variable, leaving such a power line with an overall low
usage. At this time, the utility will raise the carrying capacity of the line
by about one-third, a level sufficient to preclude another upgrade for a
while. Operating in this way without a storage component, the power
transmission and distribution system must obviously be built out, making
it capable of meeting the greatest demand under harsh conditions. It is
evident, however, that it is not the entire system that reaches constraint
but single points in the system.
Storage technologies can help alleviate both the lack of transmission

capacity and the stability problems that ensue during periods of
high demand (fig. 4–11). Sited past a bottleneck, they can provide a
prepositioned source of energy, which will allow the system to ride-
through a few short-term peak demands during the month—postponing
the need for an expensive upgrade on the line until a more sustained
level of demand warrants an upgrade. As an added benefit, the power
lines will experience greater, but less intense, use and wear through
reducing overloaded lines (peak power transfers that rise above rated
limits incur far more wear and tear on a line than pushing additional
power through the line at night when the usage is low). A key strength
storage facilities have over traditional solutions to counter reliability and
APPLICATIONS
197
capacity problems is the multifunctional capability of the unit. Frequently,

it is not simply real power that is needed, but also voltage control and
even frequency regulation services that are needed. By providing such
system stability support, storage technologies can also help avoid
the need for more expensive power stability and control equipment.
Evaluation of storage technologies for this purpose has been undertaken
for years, with countless surprising findings in storage’s favor. PG&E even
concluded back in 1994 that a 1-MW, 2-hour duration battery storage
system (with a 10-year operating life) priced at $700 per kW could enable
the profitable deferral of one additional substation each year.3
Fig. 4–11. Transmission and distribution (T&D) asset deferral (Ardour

Capital Investments).
At the distribution level, other utilities such as PacifiCorp are

beginning to use other storage technologies to postpone remote power
line upgrades—a constant challenge that is both costly and lengthy.
Long-distance, low-voltage distribution lines are especially prone to
low-voltage problems. Here, frequent upgrades are either impractical
or impossible, giving greater support for an energy storage technology
solution. In this case, PacifiCorp has a 200-mile, 25-kV distribution feeder
198
near Moab in Utah, where increasing episodes of high loading caused

periods of severe voltage problems. The problem had become so severe
that new connects were delayed, and power quality problems were so
numerous that many consumer complaints were lodged with the Public
Service Commission. Several potential solutions were examined, including
conventional planning options such as line reinforcement, additional
power lines, substation upgrades, and increased reactive compensation.
However all were found either impractical (due to environmental
restrictions) or expensive. The use of energy storage as an alternative
proved to be the most economically attractive solution and resulted in
the installation of a 250-kW, 2,000-kWh vanadium redox battery energy
storage system (VRB-ESS) with a vanadium-redox flow battery from VRB
Power Systems in late 2003 on a distribution feeder near Moab, Utah.
Continuous full-power, daily cycling operations began in March 2004.
Feeder voltage deviations have improved by 2%, and feeder power factor
improvement from VRB-ESS have reduced line losses by 40 kW, more
than offsetting the parasitic losses of the VRB-ESS. The unit is capable of
providing a variety of functions, including peak shaving, load following,
dynamic frequency control, voltage support, and islanded operations;
and it provides premium power for industrial and commercial customers
on the power line. Although the final solution has always included
developing a power line upgrade, by using the energy storage facility,
the utility now has the ability to schedule the work at its convenience,
while power quality along the line is assured. According to Brad Williams,
director of Power Delivery Business Technology at PacifiCorp,
The VRB application buys time to work with communities

and regulators on difficult siting and aesthetic issues that new sub-
transmission lines and substations always seem to bring. Once the
permanent sub-transmission line and substation solution is built
in a number of years, we can relocate the VRB to another site for
additional benefit.
APPLICATIONS
199
Retail Market
Although energy storage technologies are already widely used in the
retail market, they stand to play a much wider role protecting consumers
from growing power-quality issues or volatile power prices. In this way, as
opposed to the revenue generation opportunities of the wholesale power
market, the retail market for energy storage technologies is more focused
on providing cost-saving and loss-prevention capabilities. Commercial
and industrial firms have enacted programs to reduce their cost of using
electricity, both the direct cost of the service and the impact it can have
on their business operations, for many years. U.S. firms spent $133 billion
for electricity service in 2000.4 Unfortunately, those firms with larger and
more unpredictable demands are normally subject to a demand charge
and multistage tariffs, increasing the propensity for a greater variability in
their costs; much as these firms dislike large power bills, however, they
like unpredictable bills even less. Conventional strategies to reduce these
firms’ exposure have picked much of the low-hanging fruit, but some
strategies threaten to interrupt production schedules, costing more to
the firm than what can be saved from lower energy bills. Protecting
ongoing operations from poor power quality is also a growing issue for
many firms, as increasingly sophisticated IT equipment and computer-
controlled process machinery grow less tolerant of voltage or frequency
variations. In fact, there are some estimates suggesting that poor power
quality now costs the U.S. economy $150 billion annually from damaged
equipment or lost production and work time.
Just as changing market conditions drive the need for more storage
technologies, improving storage technologies expand the number of
applications able to be addressed. Many of these technologies will be
increasingly implemented into firms’ operations where existing storage
solutions do not currently exist, but the need is growing. For shorter-
term power-quality problems, fast-reacting energy storage assets like UPS
units can improve the quality and usefulness of the customer’s power
200
to protect critical loads from any disruption that would impact their
normal operations, or even provide short-term ride-through supplies
to bridge power supplies until on-site power supplies can start up
and protect critical loads. To support cost-minimizing strategies, these
technologies can also provide multihour discharges to reduce demand
charges and manage energy usage. Growing interest is also emerging to
take advantage of the cycling capability of storage technologies so as to
balance the load from short, repetitive motion.
By acting as both a sink and source of power, energy storage facilities

can play a role in the retail market along three broad applications:
• Power quality: Provides a means to protect sensitive

information technology or high-speed industrial equipment.
On-site UPS equipment can also act as a bridging power
source until backup generators come online in the event of an
interruption of system power.
• Energy management: Provides an on-site cache of energy
to modify demand profile. This cache can either support a
cost-of-service minimizing strategy through some form of peak
shaving, or provide a regenerative energy charge/discharge
capability for heavy cyclic loads to reduce wasted energy.
Power quality
As mentioned previously, poor power quality is a growing problem
for U.S. businesses. Although the exact value is difficult to determine,
many current estimations of the annual cost to the U.S. economy from
poor power quality agree with the DOE’s value of $150 billion from
interruption or loss in operations, work time, or damage to increasingly
expensive equipment. This problem has been acknowledged for many
years, prompting many utilities to estimate the impact of this poor
power quality on the economy in the areas they serve. Duke Power
APPLICATIONS
201
(now Duke Energy) developed a detailed estimate of these customer

impacts in 1996 and found that voltage sags and interruptions cost $4.3
billion per year to their customers. The DOE extrapolated this value to
the national level, and derived a value in the range of $150 billion in
total costs.5 Unfortunately, these estimates will always remain subjective,
with some estimates pointing to an even larger problem. For instance,
one early survey of 450 information systems executives at Fortune 1000
companies in the late 1990s reportedly revealed that power quality
problems resulted in lost data and productivity that were estimated to
cost U.S. businesses $400 billion a year.
Both industrial and commercial firms suffer from poor power

quality conditions; while transient power fluctuations disrupt and
damage industrial batch processing (lost work product) and continuous
processing (schedule disruption), commercial firms are also feeling the
impact as fault-intolerant information technology spreads throughout
the firm. Voltage sags and swells are the most common power quality
problem, and these events can originate from obvious external sources,
such as lightning strikes, but also from less-obvious internal sources,
such as large electrical loads (motors, etc.) suddenly switching on and
off. Surprisingly, most power disruption events do not last long. An EPRI
study analyzing utility disturbances found that 98% of all events last less
than 30 seconds, and more than 90% of power quality disturbances last
for less than 2 seconds.6
In the industrial market, the effects of poor power quality are

well known. The same EPRI study on chronic U.S. retail power quality
issues revealed that more than 90% of all manufacturing facilities would
experience utility voltage sags of greater than 20%, with more than
30 dips per facility (greater than 10% each year). Serious voltage sags
lasting only a few cycles with a deviation of 30% from normal have been
known to trip motor drives, controllers, and other industrial equipment,
causing significant interference with the manufacturing line. Although
most discussions about power quality have dealt only with power from
202
the utility, power quality within a large manufacturing facility can also
vary and be affected by the alternating uses of electrical equipment
throughout the plant. Motors and drives for precision-controlled
equipment are increasingly susceptible to voltage disturbances, even
from those stemming from within the facility itself.
Commercial firms are also requiring an increased level of power

quality protection. First, as mentioned previously, information technology
in the commercial market is also increasingly susceptible to poor power
quality; other findings from EPRI show that a typical computer system
annually experiences around 300 power disturbances outside the
manufacturer’s voltage tolerance limits, potentially disrupting millions
of dollars of transactions for some financial institutions. Secondly, the
type and location of the commercial building can have an impact on
the power quality available. For instance, buildings in older parts of a
city may experience a higher failure rate because of congestion and
aging equipment. Unfortunately, the suburbs do not offer a much better
solution because many commercial office parks are widely distributed,
with older ones built to carry a far lower load than is currently demanded.
With a greater distance from the local substation (compared to an
industrial facility), older equipment on a longer, smaller circuit poses a
growing need for stabilization solutions.
What these firms want—either from their service provider, or

increasingly from their own energy management strategy—is a means to
provide a capability for loss-prevention from these transient frequency
variations or voltage surges and sags. UPS systems, a long-standing use of
energy storage technologies, have provided both enhanced power quality
and reliability to consumers for many years (fig. 4–12). Its first use is to
protect critical loads from momentary power sags that damage sensitive
equipment. As the impact of poor power quality grows, the global market
for UPS systems was estimated to be $7 billion in 2002, with growth over
the foreseeable future to average 7% a year.7 Because most line instability
incidents that disrupt production lines or computer equipment last only
a brief instant, storage facilities can provide clean power by providing
APPLICATIONS
203
repeated, shallow discharges to combat such issues as voltage dips and sags.
Often these systems are made up of lead-acid batteries, but other storage
technologies are extending the capabilities of this class of equipment.
A second important use of a UPS system is to provide a ride-through or
bridging-power source of power in the event of an interruption of utility
service to critical loads until an on-site power generation is running and
stable. Generally, 15 to 20 seconds are needed to either bring a backup
power generator online or a switch to a different feeder line.
Fig. 4–12. Power quality (Ardour Capital Investments).
Finally, once a backup generator begins supplying power in an off-

grid environment, power quality is still a concern. Simply put, firms want
similar power quality from their distributed generation setup as they get
from utility service. Unfortunately, most do not. Fuel cells, some natural
gas-fired units, and other smaller-generation technologies are unable to
reliably react quickly to a changing load. If the generators serve only a
nonvarying load for the facility, the generator may handle most variations
in load. If the load covered by the backup generator changes because of
processing requirements, however, some storage technologies such as
flywheels show promise to provide a stable (frequency regulation and
voltage stability) environment while the facility is isolated; in essence, the
204
storage unit would provide a load-following capability because the single

generator has a far more difficult time responding to a changing load
than the utility system power. To perform this role, the storage asset will
need the ability to operate in dynamic conditions with many charge and
discharge cycles during this period of interrupted utility service. Because
the storage unit can react more quickly, these combined systems will be
able to perform over a much wider range of operating conditions than
just the distributed generation asset itself.
An outsized example of the need and usefulness of a UPS system

can be seen in the power quality problems solved at a semiconductor
wafer fabrication facility in Phoenix, Arizona. Utility power disturbances
are a great concern in the wafer fabrication industry because producing
the final product can take 30 to 50 days, and disruptions at any point
along the process can ruin the entire batch. To add an additional
level of protection, in August 2000 STMicroelectronics installed a
12,500-kVA PureWave UPS system (five 2,500-kVA units) from S&C
Electric Company operating at 12,470 volts (V). Based on the analysis of
Arizona Public Service’s system events, three options were considered:
numerous conventional 480-V UPS systems installed within the facility
to protect critical process; a medium-voltage, high-speed source-transfer
system installed between the plant’s two incoming utility feeders; or an
off-line PureWave UPS system installed at the utility source. The single
medium-voltage UPS system was ultimately selected because it provides
protection for the entire operation and ensures a higher reliability
because a fault on an unprotected aspect of the facility could affect the
entire process. This off-line system operates only during a utility system
disturbance and switches to the battery bank within 4 milliseconds (less
than 1/4 cycle). When the system power returns to normal, the load
is then switched back. Because the duty cycle is only expected to be
100 seconds of run time, such a high power rating can be obtained
from a relatively small system. The system has proved its worth since its
installation; it has protected the facility more than 90 times in the first
four years of operation—the first a mere 10 days after installation.
APPLICATIONS
205
Energy management
Managing energy usage at commercial and industrial firms has
become far more involved than simply signing up for an energy audit. As
mentioned previously, the original desire by commercial and industrial
firms to simply have lower energy bills has been joined by the aspiration
to prevent volatile power prices from affecting the bottom line. For many
firms, even a small rise in energy costs could wipe out hard-won efficiency
improvements; worse yet is simply an unpredictable price. Conservation
programs have been used for many years to reduce enterprise energy
costs, but they have two drawbacks: many do little to counteract the
specific high-use periods that drive demand charges, and those load-
shifting programs that do quickly begin to interfere with production
operations. These demand charges are commonly a much greater factor
in many industrial heavy users of electricity, especially if their energy use
fluctuates from batch job processing.
Like conservation programs, energy-management programs using

storage technologies would target reducing power-intensive activities
that occur during peak periods; unlike most conservation programs,
there could be little interruption in the operation of the facility. These
technologies can help commercial and industrial firms reduce their costs
along two main routes: peak shaving and providing a sink to capture
otherwise wasted energy in repetitive motion activities. For both uses,
storage augments and expands the capability of any existing energy
management program. Because customers with more stable energy use
patterns are generally more welcome to a utility than those with highly
variable loads, the wider introduction of these technologies into the
marketplace could even gain support from the local utility.
Peak shaving. Peak shaving allows customers to directly reduce

their utility bills by arbitraging power stored from off-peak to on-peak
periods. By doing so, energy storage technologies allow firms to engage in
load shifting without the subsequent effects on the production schedule.
The total use of energy would not be affected, but the firm’s costs would
206
decrease because of the lower use during higher-priced periods of the

day (if the firm’s tariff has multiple levels). However, the main focus for
such actions is generally to reduce the firm’s demand charge (fig. 4–13).
By maintaining such a cache of energy on-site, the energy storage facility
can also be integrated with other energy management strategies gaining
favor, such as a demand response program, to provide the firm with greater
flexibility to reduce its usage cost of energy. To date, thermal energy storage
technologies already have a well-established track record of successfully
reducing demand charges at industrial and commercial firms, providing an
example for other technologies to follow.
Fig. 4–13. Peak shaving (Ardour Capital Investments).
Although heavy energy use increases the commodity energy charge,

highly cyclical loads such as cooling or batch processing can increase
the demand-charge component of the bill dramatically. For this reason,
even for many firms where overall energy costs are only a small part of
general costs, often certain components of their energy usage may be
well-suited to incorporating an energy storage component. For example,
the DOE notes that 25% of commercial usage for electricity is for cooling
load, a highly cyclical operation. One current example where storage
technologies are already used for peak shaving roles is thermal energy
APPLICATIONS
207
storage (TES), which assists commercial chillers to reduce the cooling

load of the building. These central chillers predominate in medium- to
large-scale buildings (the larger the building, the more potential benefit
from TES), and 80% of buildings greater than 200,000 square feet use
them, about 25% of all buildings.8
For customers with existing air conditioning equipment, thermal

energy storage systems can be justified as a retrofit based on the reduced
costs of using off-peak instead of on-peak power. Typical payback periods
for these retrofit thermal energy storage systems can range from one to
three years. This is possible through savings by reducing the demand
charge and lower peak-power purchases. Following this strategy, thermal
energy storage facilities can reduce peak power demand by 50% and
reliably reduce cooling load costs anywhere from 20% to 30%. Because
each retrofit installation is project specific, exact costs are difficult to
establish, but generally range from $250 per peak kW shifted to $500 per
peak kW shifted.9
Integrating a thermal energy storage unit with new medium to

large construction provides for even greater benefits. In most cases the
decreased design requirements on the building’s cooling system will pay
for the cost of the system, plus provide the additional operating cost
reduction described in the retrofit installation. In particular, because TES
systems allow a building air conditioning system to operate at lower
temperatures, duct sizes can be reduced by 20% to 40%. Because the
cooler air also requires a smaller volume of air to cool the building,
fan motors, air handlers, and chilled water pumps are smaller and
less expensive. Based on the reductions in capital equipment needed
when a TES system is integrated into the original design, EPRI has
estimated that overall ventilation and air conditioning costs are reduced
by 20% to 60%.10
Besides these costs, energy storage technologies can provide

additional support and savings. Hotels, hospitals, and other commercial
buildings with large cooling loads require significant power demands
208
during start-up—events that can occur many times per day. In these
situations, power constraints often require staggering the start-up of
cooling equipment; an inconvenience that costs time and reduces the
flexibility of operations. With additional energy storage technologies
to provide much-needed short-term pulse-power to reduce these
intermittent demand spikes, anecdotal evidence suggests that peak
demands could be cut by upwards of 25%.
Regenerative energy. Another area of retail energy management

gaining significant attention of late has been the use of an energy storage
technology to improve the efficiency of repetitive motion operations.
These operations are often extremely wasteful; for some operations,
strategies have been devised, such as the counterbalance in elevators,
to recoup otherwise wasted energy, but many operations remain
without satisfactory solutions. Energy storage technologies with high
cycling capabilities hold out the potential to play a much wider role
in capturing and reusing otherwise wasted energy in repetitive motion
processes (fig. 4–14). This role of acting as a dynamic sink and source for
power fits well with transportation sector applications, where repetitive
starts and stops produce inefficient use of energy. Although container
port lifting cranes and light-rail/subway systems all exhibit these usage
patterns, many other applications could benefit from such an addition
to the process, even including the manufacturing floor, where voltage
fluctuations can be caused by the many motors turning on and off
constantly throughout the day. It should also be noted that if an energy
storage solution is used to lower the operator’s costs for the customer,
even the local utility stands to benefit (especially on larger installations)
because the load swings are reduced on the power lines feeding these
installations, especially in the case for urban subway/electric bus systems
where the extremely variable (and short-duration) demand requirements
of the starting and stopping often place unacceptable strains on the
city’s power system and actively work against the goal of providing
reliable, stable power to other customers. Although battery technologies
generally dominate the retail energy market, traditional solutions such as
APPLICATIONS
209
lead acid batteries are not able to support these applications because of
the high cycling requirements. For many of these more demanding roles,
flywheel systems have seen some recent success.
Fig. 4–14. Regenerative energy (Ardour Capital Investments).
For instance, urban rail systems represent a key opportunity for

storage technologies to capture and recycle energy usually lost during
braking. A prominent example is England’s Urenco Power Technologies
(UPT) deploying a 400-kW flywheel system on the San Francisco MUNI
rail system, enabling upward of a 20% reduction in energy purchases
through capturing the braking energy of the trains instead of allowing
that energy to be dissipated through braking. Besides the energy saved,
the stored energy will reduce the energy demand spike required during
the acceleration phase, creating less wear on the equipment. Besides
San Francisco, New York City’s transit system has also investigated
the use of flywheels from UPT. Here, a 1-MW UPT Trackside Energy
Management System has operated since January 2002 at the New York
City Transit (NYCT) Far Rockaway Test Track, providing voltage support
and optimizing the use of a new generation of trains by reduction of
peak power demand through recovery of the braking energy. A solution
was needed because voltage sags of 10% or more were common during
210
acceleration of these new trains. The candidate technologies included

batteries, super capacitors, and flywheels. Batteries were rejected because
they could not meet the arduous cycling requirements of the application.
Super capacitors were also considered unsuitable for the application as
the cyclical demand greatly reduced their life to 5 years or less under
these conditions. Flywheels were selected because of their high cycling
capability, long life (20 years), and minimal maintenance. The flywheels
were able to stabilize the voltage of the system and also supported the
adjacent operational lines during normal operation.
In the previous example, lowering demand spikes helps reduce

power fluctuations on a utility’s highly congested urban power grid.
Using energy storage technologies has benefits for rural utilities as well
(besides the industrial firm). For example, in the Usibelli Coal Mine in
Alaska and its utility, the Golden Valley Electrical Association (GVEA),
the mine installed a $1-million, 40-ton flywheel (in agreement with
GVEA) to support the mine’s dragline load. During the unit’s operation,
the mine’s electrical load is greatly affected by the dragline because its
demand alone can swing by as much as 4 MW (across an 8-MW band)
and cycle every 60 seconds. To provide a more manageable power level
to the utility (and hence, lower their costs), the mine installed a flywheel
system that can produce 5.2 MW for 3 seconds (plus a 1.8-MW gen-
erator for longer-term energy needs), thereby reducing the worst of the
demand spike. This set-up was necessary because GVEA’s turbines could
not handle the large power swings of the dragline, causing the power
frequency to fluctuate with each dragline cycle. With the use of this fly-
wheel system, the average power demand from the utility is now only
around 2 MW to 2.5 MW, and GVEA only sees about a 500 kW band
of power fluctuation. As a precaution, there is a radio link between the
flywheel building and the dragline, so the dragline operators are warned
when the flywheel shuts down for any reason. Without the flywheel
in operation, each use of the dragline can be watched on the utility’s
meters in Fairbanks (100 miles away), and even noticed in Anchorage
(350 miles away).
APPLICATIONS
211
References
2003. Palo Alto, CA: Electric Power Research Institute; Knoxville, TN:
Electrotek Concepts.
2. Edison Electric Institute. 2001. Getting electricity where it is needed.
Washington, DC: Edison Electric Institute.
3. Abbas, A., S. Swaminathan, and R. Sen. 1997. Cost analysis for energy
storage systems for electric utility applications (SAND97-0443), 14.
Albuquerque, NM: Sandia National Laboratories.
4. Energy Information Agency. 2004. Annual energy review 2002.
5. Swaminathan, S., and R. Sen. 1998. Review of power quality
applications of energy storage systems (SAND98-1513). Albuquerque,
NM: Sandia National Laboratories.
8. Thermal energy storage—Economics and benefits. 2002. Arlington, VA:
E3 Energy Services, LLC.
9. Ibid.
10. Electric Power Research Institute. July 1991. Cold air distribution with
ice storage. brochure CU-2038. Palo Alto, CA, EPRI.
5 RENEWABLE
ENERGY AND
STORAGE
Increasing the use of domestic renewable energy resources is
one of the primary goals of U.S. energy policy makers. The U.S. federal
government has led this effort with its continued support of the
National Renewable Energy Laboratory’s National Wind Technology
Center research and the 1.8¢ per kWh Production Tax Credit (PTC).
Unfortunately, the sustainability of this support has been less than
reliable—consistency of support is crucial for market development, and
it has been lacking at this level. Luckily, state governments are joining
the fray through their Renewable Portfolio Standards (RPSs), which are
increasingly seen as the primary driver for the deployment of additional
wind turbines. Increasingly, these RPSs are setting more aggressive targets,
as in California, where 20% (possibly moving to 33%) of all generation is
required to come from renewable sources of energy by 2017. Building off
this newfound support, there are increasing calls for renewable energy (in
Europe and the United States) to provide 10% of all electrical generation
(or all energy usage) by 2010; 20% by 2020; and 50% by 2050.
Unfortunately, there is an underlying problem to this growth target.

Renewable energy market penetration continues to lag expectations, and
significant efforts on many fronts will be needed to expand the level of
214
renewable energy generation to the point where it can begin to account

for such a significant segment of the supply base. According to the U.S.
Department of Energy (DOE), renewable energy sources (nonhydroelectric)
provided only slightly more than 2% of the nation’s electric power in 2002,
holding even or slipping each year since 1991, when it peaked at roughly
2.5%. Including hydroelectric power simply moves these percentages
upward to 9% and 11%, respectively, but the trend continues to decline.
Shifting to a total energy accounting makes little difference, as renewable
energy resources account for only 6% of all energy consumed in the United
States in 2002, down again from more than 7% in 1991.1 These problems
extend into the future. As can be seen in figure 5–1, there is a growing
deficit in the level of expected renewable energy production and the
previously stated goals for such renewable energy. Much of this shortfall
stems from lack of additional resources from traditional renewable energy
sources, such as hydroelectric, biomass, landfill gas, and the like. Although
power production from these resources will continue to expand, only solar
and especially wind power are flexible enough and can be integrated most
easily into the existing infrastructure to hold out the prospect of meeting
this or any other significant renewable energy production increase.
Fig. 5–1. Renewable energy shortfall (Data: DOE, Annual Energy

Outlook–2004).
R E N EWA B L E E N E RG Y AND S T O RA G E
215
Wind power has long been sought after as a source of renewable

energy, but its cost and operational issues have retarded significant growth
in the United States until recently. Fortunately, the wind energy potential
in the United States is immense, with proponents of wind energy quick to
point out that the continental United States has the potential to produce
nearly 11,000 billion kWh of electricity—three times current demand of
the country.2 For wind energy project developers, however, even more
interesting is that areas with the highest sustainable winds are highly
localized, with only a small percentage of the contiguous United States
containing the majority of the addressable energy resource. Developing
these resources is accelerating, and more than 6,000 MW of wind
turbines were installed in the United States by the end of 2003—half
of this total within the three preceding years. These advances have only
become possible as wind energy has grown to be far more economically
competitive—essentially matching (or beating) the production cost
from competing power sources if located in favorable wind conditions.
However, even as production costs continue to decline, inherent
operational disadvantages will continue to put these wind power turbines
at a significant disadvantage in the broader power market.
Coupling energy storage technologies with wind turbines can solve

many of wind power’s operational issues and support the continued
expansion of wind energy production. Many types of renewable
energy production already benefit from energy storage technologies.
By decoupling the production and delivery of energy from renewable
resources, storage technologies can make the generated energy more
useful and more valuable—resulting in a better standing in the market
for these facilities. To date, the wind power industry has made great
strides in enhancing the capability of wind turbines and how they are
integrated into the overall power market. Unfortunately, although the
direct production cost may now (at certain locations) be competitive
with other power generation resources, its effective usage cost is still
higher because of inherent qualities of the wind resource. Storage
technologies can provide additional flexibility to mitigate these issues. For
216
example, storage can promote system stability to encourage further wind

penetration of the market in both small and large power grid systems.
Although on small power grids the activity will focus around providing
a high-cycle dampening capability (high wind penetration), storage
technologies can provide more of a commodity balancing capability for
wind on larger power grids, giving system operators a greater level of
reliability (through lower variability) for the growing quantity of wind
energy. Without addressing these and other negative qualities of the wind
resource, these challenges will continue to restrain the development of
wind resources and their integration into the market.
Resource Utilization
Challenges
Certain qualities of the wind resource itself will continue to hinder
future development of wind power generation. In particular, three
systemic challenges must be overcome to ensure the widespread
adoption of this renewable energy resource:
1. Its noncoincident peak

2. Its nondispatchable nature
3. The stability of the system as wind gains greater market
penetration
Wind turbine technology has made great advances in improving the

conversion efficiency of this resource, but being competitive on price
does not also mean that wind is as useful an energy resource to the
system administrator as other more reliable generation technologies.
Now that wind-generated power is becoming competitive with other
generation technologies, it needs to be made equally useful (for
217
planning purposes) to gain wider market acceptance—especially as its

market penetration grows.
Energy storage technologies stand to improve the utilization (and,

hence, profitability) of wind projects both now and in the future. It is
true that in the near term, there are still many opportunities for wind
turbine–only projects to succeed simply by extending the current
technology and market strategies. However, even the most profitable of
these projects could benefit from the inclusion of a storage component.
With the storage component providing a cache for the wind energy,
a wind project could approach the sale of its output to the market
differently, reducing or even removing the utility-imposed charges that
impact its profitability. Storage facilities could even relieve temporary
transmission congestion from remote areas with established wind parks
that are already threatened by transmission availability issues.
In the long term, these resource utilization challenges will bear

directly on the level of market penetration wind power can attain. So
far, the existing low level of wind penetration has not been difficult to
account for on utility power grids. However, power system operators
are hesitant to increase the penetration of such intermittent resources
indefinitely. Some way to balance their output must be found, and
the cycling of fossil power plants to both demand and wind output
variations might not be a workable solution—especially to the owners
of the fossil power plants. In fact, small island power grids installing wind
turbines already have such problems and are finding storage a solution
to many of their needs. Finally, the vast majority of U.S. wind resources
are in the West, especially the Great Plains, far away from the major
cities and industrial areas. If plans to export wind power to these major
demand markets are ever to be considered seriously, it must be done
efficiently to reduce the required transmission investment—something
already in short supply. Storage technologies could assist in bringing this
power to market—a crucial step for reaching many of the various states’
renewable energy targets.
218
Noncoincident peak
The first market challenge in delivering wind energy to the utility grid is
that the wind resource is noncoincident with the peak power demand—it
simply does not blow the hardest at the best times to produce electricity.
For many sites, upward of 67% of the total wind power resource can be
outside of the peak demand period (i.e., 9 a.m. to 5 p.m. Monday through
Friday), as much of the wind power is only available in the morning and
evening. Besides this daily variation, the average wind speed for a particular
site also follows a seasonal pattern, with maximum winds occurring (for
most locations) in winter and spring, and minimum winds in summer
and autumn. These trends in the resource potential present a problem for
many utilities because summer is the peak demand period for most of the
United States. In fact, for much of the Great Plains (where the majority of
domestic wind resources exist), the summer wind power resource can fall
anywhere from 20% to 30% below the wind power of winter.
This mismatch of wind resources to power demand has an obvious

impact on the profitability of the wind turbine or wind park. Because
wholesale electric power prices vary throughout the day, when power is
sold matters just as much (or more) as how much is sold in determining
the total value of a wind turbine’s output. Unfortunately, unlike other
power generation facilities, wind turbines cannot choose when to
produce, leaving control over any market strategy the installation may
have not in the hands of the owners. Therefore, as the existing strategy
is simply to produce all that one can—whenever one can—and hope
for the best, any strategy that could maximize higher- over lower-margin
production would be well received.
Nondispatchable
The second market challenge in delivering wind energy to the utility
power grid is that the wind resource is intermittent—it is unreliable from
a scheduling point of view. Not only is the timing variable, but also the
219
intensity is correspondingly variable—leading to uneven power production

throughout the day. This variability in the resource combines to leave many
sites with an average utilization of only 33%, although siting in exceptional
locations or using the latest wind turbine design is pushing this toward 40%
or higher. This unpredictability relegates wind energy to a different status
than power from conventional power sources—still welcome for a number
of reasons, but with definite reservations from a scheduling point of view.
In a small power grid environment, for example, this unpredictability
requires the system operator to maintain additional dispatchable resources
to back up the wind generators, whereas in the broader wholesale power
market, system operators account for their additional balancing cost in part
through a reduction in the capacity payment to wind developers—often
by 80%. Besides losing this capacity payment, unpredictable wind resources
do not allow wind generators to maximize their guaranteed delivery of
power during peak pricing periods.
The wind industry has undertaken significant efforts to overcome

this dilemma by extending the scope of existing weather forecasts into
effectively highly detailed wind speed forecasts. Most of this effort has
taken place in Europe, where reliability for hour-ahead wind speed (and
hence, potential wind energy) forecasts has reportedly reached a 95% or
more accuracy rate, allowing wind resources to participate in short-term
energy imbalance and even load-following roles. Unfortunately the day-
ahead forecasts continue to fall short of this mark, requiring additional
generation resources to be maintained in more traditional standby
availability in case of a larger energy imbalance.
System stability
The two previously mentioned market integration challenges for wind
resources conspire to produce system stability issues as the magnitude of
wind resource penetration reaches approximately 15% of an area’s supply.
This is not a hard-and-fast rule, as power grid geometry, transmission
capability, and responsiveness of existing generation facilities all play
220
a role in determining at what point the variability of the wind power

begins to affect the stability of the power grid—especially in areas where
transmission constraints exist. Although there is not a significant amount
of experience in the United States on this issue, in a number of places in
Europe (especially in Denmark), wind energy accounts for 20% or more
of the energy production for the power grid. Their experience is that this
high of a percentage of wind energy produces unstable conditions on the
transmission power grid, akin to (according to some European power grid
operators) “driving an articulated lorry with no brakes and no steering.”
The setup requires additional management adjustment of dispatchable
generation resources—increasing production costs and environmental
output—in addition to adjustments for a changing load during the day.
As wind resources continue to be developed, especially in highly localized
remote areas, these problems will become more apparent in the United
States over the coming years.
Early evidence in the United States shows that as wind penetration

reaches 10% of an area’s load, the system operator has charged up to 10%
of the value of the electricity for regulation purposes. As the penetration
level deepens toward 20%, the additional charges increase only slightly,
but shift toward day-ahead energy imbalance costs instead of hour-ahead
regulation. Although the charges for this can vary greatly depending on
the location, timing, and market penetration level of the wind resource,
they are substantial. Studies on the integration cost of wind power
have ranged from $1.80 per MWh (Xcel utility study) to $6 per MWh
(PacifCorp utility study).3 Because most potential wind energy power
generation sites are located in remote locations where there is generally
a limited amount sufficient transmission capacity, a large amount of
localized wind energy production in these rural areas can strain existing
power transmission systems and potentially limit the transmission of this
power to the demand centers.
221
Finally, besides the daily balancing issues, adding significant wind

energy resources to a region’s power grid can also affect the regional
power flows and, therefore, must be taken into account when
accomplishing resource planning or transmission upgrade studies. The
problem could even put into question the ability of states to hit their
renewable energy penetration targets. For example, California’s mandate
for 20% (or possibly 33%) of its generation coming from renewables by
2017 may prove problematic, as even existing wind turbines can be
underused at times because of transmission constraints. This is of real
concern, as illustrated by Ireland; with only 7% of its generation coming
from wind power, the country imposed in early 2004 a moratorium on
the connection of new wind power because of power grid instability
issues. Other states with similarly aggressive goals as California’s may
also have similar experiences, making it difficult for them to reach their
mandates in their renewable portfolio standards.
Remote Power
Renewable energy technologies have long been a favorite for small,
remote power applications; adding a storage component to these
systems is not simply useful, it is often a necessity to make the setup
practical (especially for use after sunlight hours as seen in figure 5–2).
The desire to use renewable energy sources for these remote applications
is generally based on one of two main drivers: environmental impact or
cost. In these off-grid systems, many installations are in environmentally
sensitive locations without the benefit of system power, creating a bias
(or even mandate) toward using renewable resources over a diesel
generator with its accompanying emission and leakage issues. As for
cost, connecting the small load to the nearest power grid is also usually
impractical because the cost of running a distribution power line to a
remote site can quickly rise past any economic validation. This solution
has its own environmental impacts, besides the wildly expensive cost
222
per delivered kWh for a small, intermittent load. If self-power is chosen,

even the diesel option can be too expensive from a total-system cost
perspective, because reliable diesel fuel delivery can become impractical
or cost-prohibitive because of the transportation costs, which can easily
account for more than the fuel charge.
Fig. 5–2. Remote power (Ardour Capital Investments).
Because of the limited power availability (many systems are less than
1 kW in size), these renewable energy power units are generally restricted
to supporting remote communication equipment or a limited residential
network or other commercial electrical equipment. The renewable
technology most often coupled with a storage unit for these applications
is photovoltaic (PV), although some small wind turbines are also used
in the larger systems. In 2003, nearly 750 MW of solar PV panels were
delivered globally, with the market expected to maintain its 30% plus
growth for the foreseeable future; of this, roughly 25% goes toward the
off-grid market. To minimize operating and maintenance costs, standard
lead-acid batteries are most often used to store the power until needed,
because their operating capabilities and costs are well understood for
these applications, and the duty cycles are generally not substantial.
However, most batteries used in these installations were originally
223
designed for the automotive market, so, although many of these systems
are of minimal size, they must contain sufficient power electronics to
prevent either under- or overcharging that would reduce the operating
life of the battery. In general, the solar panel component of these remote
power systems makes up approximately 50% of the total system cost,
with the storage component often representing approximately 20% of
the total system costs. Although it adds to the cost, a storage component
does provide economic benefits. With storage, the solar component
can be undersized for the required load, thereby reducing the capital
costs of the system—possibly by upward of 10%—while expanding the
usefulness of the system significantly.
International prospects for these remote power installations far

outnumber U.S. domestic opportunities. Many of these projects are
concerned with bringing electricity to remote villages where reliable
power grid power is not available. In these locations, governments are
supportive of these programs because many of these areas may not be
served for many years by the expansion of the national power grid. A
number of good examples of these projects can be examined through
the National Renewable Energy Laboratory (NREL) Village Power
Program. Some are discussed in the following sections.4
Example—Malaysia.5 Remote PV systems from BP Solar Malaysia

and its joint venture partners—Projass Engineering and Projass Enecorp—
are cornerstones of the Malaysian Ministry of Rural Development’s
(MORD) rural electrification scheme. This program provides electrification
to remote communities in the peninsular and the East Malaysia states
of Sabah and Sarawak, in one of the largest rural infrastructure projects
in the world. These intrinsically simple, stand-alone PV system packages
were specifically designed to deliver small-scale AC power in an easy-
to-maintain package of solar panels, power electronics, and lead-acid
batteries (flooded & gelled). Since 1996, MORD has spent $39 million on
this ongoing program to install more than 11,600 PV systems (2,200 kW
in total for all systems). These units provide power to individual homes,
224
long-houses, rural clinics, community halls, schools, and churches for uses
ranging from basic lighting to vaccine refrigerator-freezers; these systems
are now responsible for generating 3,300 MWh per year.
Island Grid
Island grid is simply another name for a small, isolated power grid,
usually serving a single town and its accompanying commercial and
industrial load. Because of its small size, its load profile is generally much
more volatile than larger systems, and it is not uncommon—as is found in
many Alaskan isolated grids—for the system’s peak demand to be three
times the minimum load, and twice the average load. These systems can
be found on an actual island, but they also are small power grids in many
rural areas where the existing transmission capacity is severely limited,
effectively isolating the area from most balancing support of the main
power grid. Because of the need to supply the load over such a variable
range, these systems’ baseloads are generally small, requiring a number
of small additional power facilities to balance and stabilize the load as it
shifts throughout the day.
The most common power technology used in these systems is a

diesel-reciprocating engine. Although dependable, these diesel units
require fuel and maintenance, which are often very expensive because of
transportation issues (with the transportation charge much larger than
the fuel charge). As the units cycle on and off in response to changing
load conditions (some only lasting minutes), operating efficiency declines
and emissions increase; such activity also raises the maintenance costs
and shortens the unit’s life. To reduce the wear and tear on the units from
multiple starts, many systems impose a minimum run-time on the diesel
units, but this strategy then further lowers fuel savings, driving up costs.
For example, diesel generator sets (gensets) only partially loaded can use
nearly half-again as much fuel per kWh as when run at full power.
225
To reduce the run-time (and start-ups) on these units, wind power is

often sought as an additional, local energy resource. The benefits of using
wind energy can be quite high. A number of studies by U.S. government
laboratories (NREL, Lawrence Livermore National Laboratory [LLNL],
etc.), have shown that adding wind to a diesel-powered local grid can
reduce fuel consumption by 40% to 50% and total costs by 30% to 50%
for areas with plentiful wind resources.6 However, because of the small
size of these power grids (lack of system inertia in its limited generators,
etc.) simply adding wind turbines to small power grids cannot be done
haphazardly—a systematic review of the load and potential additional
wind turbines must be undertaken to ascertain potential benefits, and
to determine what level of wind penetration is best. For many of these
power grids, the opportunity exists to have wind resources well in
excess of 50% of the peak load. However, if that much wind capacity (or
more) is added, steps must be taken to have practical uses when output
exceeds demand. This can include space and water heating, as well as
simple thermal load dumps (radiators) if no other use is found.
To alleviate the need for such creative solutions for the problem of too
much low-cost renewable energy (and system optimization), energy storage
units should be added to these isolated grids for three basic reasons:
1. Reduction of diesel starts/run-time

2. System stability
3. Increase in the level and value of wind penetration for the system
Because the most variable cost of a small power grid is the diesel
operations, the goal is usually aimed at lowering their costs—not just their
fuel costs, but the overall usage costs. Probably the greatest single factor
in determining the amount of savings a storage component can provide
a hybrid system (wind/diesel/storage) is the duration of the storage
unit’s discharge during normal operation of the system. Storage units can
significantly lower the number of power generator start-ups needed to
226
meet short-term requirements for power (fig. 5–3)—essentially providing

a ride-through capability for the system. With a larger storage capacity,
load leveling becomes possible. In 1997 an evaluation for a wind turbine
combined with a battery energy storage facility for a small power grid in
Deering, Alaska, was undertaken. The storage unit was used to provide an
extended ride-through capability to prevent a peaking diesel unit from
running for only a short period of time. In this modeling study, it was
shown that the largest savings came from relatively short-term storage;
even a 10-minute ride-through capability from the energy storage unit
reduced diesel fuel use by 18%, the diesel running time by 19%, and the
number of diesel starts by 44%.7
Fig. 5–3. Island grid (Ardour Capital Investments).
Adding a storage component when adding wind power to these

small power grids is another way to balance the system and avoid
many integration problems. Without a storage component, the system
operator must increase the spinning and regulation reserves (incurring a
relatively greater cost than on a larger system) to prevent variable wind
production from impacting the customer base. As an island power grid
is small with little of the deep system inertia found on larger power grids,
227
changes in both power production and system demand have a greater

impact than on a major grid. Acting as a shock absorber during variations
in system stability by providing enhanced frequency regulation and load-
following services, these energy storage units have a beneficial impact on
the operating costs of these systems. As an additional ready-reserve of
power, the unit could also provide spinning or even restoration services
to the power grid, enhancing the security of the local utility.
Wind penetration can be very high on these island power systems,

easily more than 50% with the right wind resources. The same U.S.
government studies that showed that increasing the wind penetration
can lower the diesel fuel costs on these systems also showed that adding
a storage component can gain an additional 10% to 20% in system cost
reductions. Because some of the primary benefits to adding storage
surround the role of supplanting diesel starts, storage contributes very little
to improved system economics at low wind penetration levels. However,
the value of storage to the system increases as the wind penetration
increases, as there will be an increasing amount of time that the available
wind power exceeds the total system loads. According to one NREL study,
at 50% wind penetration, storage can provide 20% greater fuel savings and
20% less diesel run-time than nonstorage wind/diesel systems alone.8
Example—Hluleka Nature Reserve, South Africa.9 Until 2002, the

Hluleka Nature Reserve in South Africa relied on two diesel generators for
all of its power needs. Unfortunately, it was far from an optimal solution.
Billed as a natural retreat, the constant throbbing of the generators spoiled
the peace and quiet of this exquisite coastal spot. In addition, although all
of the kitchen and bath appliances supplied in each chalet were electrical,
transportation and cost issues kept the two generators operating only 10
hours each day. Even with such a reduced load, these units still consumed
around 90,000 liters of diesel fuel annually. Although a recognized problem,
the reserve needed the flexibility of the diesel engines, because with 12
chalets of guests (plus the management staff), it had to be able to provide
for possibly 90 people at any one time.
228
To improve the quality of the stay for the guests, the country’s
National Electricity Regulator (NER), the Council for Scientific and
Industrial Research (CSIR), Shell Solar, the Department of Minerals and
Energy (DME), and the Eastern Cape Provincial Government deployed
a stand-alone mini-grid for the reserve, which combined the existing
generators with a wind, sun, and storage component to produce the
most cost-effective and efficient energy solution. Besides solving the
electrification problem, the grid was also designed to fix related issues.
Prior to this project, the water supply for the camp was pumped directly
from the river into holding dams without any purification process, and
telecommunication in the area was, at best, problematic. The mini-
grid was designed to not only provide sufficient electrical power for
the reserve, but also to provide sufficient power for enhanced water
purification and communication systems.
The resulting hybrid system makes use of a 56-kW solar array, two
2.5-kW wind turbines, a battery storage unit, solar water heaters, and
liquid petroleum gas. On sunny and windy days, the power from the
generators is fed directly to the camp and to the battery storage unit,
which can supply five days of reserve electricity in case of no wind or
sun. As a backup, one of the existing diesel generators is being retained.
Because upgrading the telecommunication equipment was part of the
project, an Internet connection allows the batteries, wind chargers, solar
panels, and power consumption to be remotely monitored 24 hours a
day by Shell Solar.
Example—Metlakatla, AK. 10 One well-known example of

how a storage facility can improve an island power grid is a 1997
DOE-sponsored (through the Energy Storage Systems program at Sandia
National Laboratory) project in Metlakatla, AK. Here, the full cost of
the $1.5-million energy storage facility was recovered within three years
because the storage unit improved overall power quality and reduced
229
the operating costs of the existing generating assets. Metlakatla is an

800-resident island community in southern Alaska (common in this area
of the state), with an economy based on fishing (commercial cold storage)
and a lumber mill. Taking advantage of plentiful water resources, the
local utility powered its local grid with two reservoirs with four turbines
that produced 4.9 MW of hydroelectric power. In 1986, a chipper was
added to the sawmill—causing the hydroelectric dam to struggle against
brownouts, overvoltage, and frequency fluctuations when responding
to load swings of as high as 900 kW. Because the hydropower dam’s
response time of 10 seconds was too slow to follow the rapid swings
in demand from the sawmill, which could shift in as little as 1/20th of a
second, the village often suffered brownouts and sometimes blackouts.
A 3.3-MW diesel system was installed in 1987 to alleviate this problem;
the unit typically operated at 20% to handle load swings and provided
55% of the system’s power. With this diesel genset, generating capacity
grew to be twice the average system load; unfortunately, continued
inadequate spinning reserves and erratic system frequency combined to
produce continued poor power quality for the utility’s customers. It was
decided that a more workable long-term solution was needed, because
even this poor level of service cost on average more than $400,000 for
diesel fuel, a nontrivial amount of the community’s scant resources.
Sandia’s Energy Storage Systems program developed a lead-acid battery

(batteries from GNB Industrial Power, a division of Exide Technologies)
energy storage system that was installed in 1997 (with a 20-year expected
operating life). Surprisingly, the system matched generation needs very
well, almost totally replacing the diesel unit in normal operation. (The
diesel system was then operated primarily to charge the battery when
the hydroelectric system underwent maintenance, or during drought
conditions.) To minimize operating costs, the energy storage system
can operate autonomously for normal charge/discharge, standby, and
disconnect/synchronization for blackstart.
230
There were both direct financial and nonfinancial benefits to using

the system. One financial benefit was that the storage facility greatly
reduced and nearly eliminated the annual transportation of 500,000
gallons of diesel fuel (at a cost of $400,000 per year) from the mainland
by ferry and pipeline. Also, each of these diesel shipments also required
an initial outlay of $100,000 for handling costs—a serious burden for
such a small community. In addition to the fuel costs, interim diesel
unit maintenance cost $150,000 every three years, with major overhauls
of $250,000 required every six years. By reducing the constant starts
and extended operating time, the storage unit has increased the time
between required overhauls.
Nonfinancial benefits were substantial for the community. The

energy storage facility greatly reduced the environmental risks to the
community because diesel emissions and noise were greatly reduced
(or eliminated). Productivity improvements were evident as enhanced
power availability and quality became the norm for the community’s
commercial fishing operation (the cold-storage facilities) and sawmill.
Outages from the earlier power disruptions were virtually eliminated,
along with the financial losses for the corresponding product spoilage.
Example—Dogo Island, Japan.11 The installation on Dogo Island,

a small island just off the coast of Japan, is an excellent example of how
a flywheel energy storage system can provide the stabilizing capability
lacking on many island power grids, which had inhibited their use of
the readily available—but highly variable—wind energy around them.
In 2003, Fuji Electric installed a 200-kW Kinetic Energy Storage System
(KESS) from Urenco Power Technology (UPT) in conjunction with a
3 x 600-kW installation of DeWind D4 wind turbines to evaluate how
wind generators can be a viable source of power on remote islands with
weak links to the mainland power grid by smoothing their irregular
231
power output. Through incorporating the flywheel-based energy storage

unit into the installation of the wind turbines, Fuji Electric sought to
fulfill four goals; they wanted to
1. Stabilize the frequency variations stemming from the turbines

2. Capture excess energy from short-term wind gusts
3. Optimize the operation of (or eliminate the need for) diesel
generators on the island
4. Eliminate the need for additional spinning reserve because of
the introduction of the wind turbines
Results to date have been promising; by acting as both a dynamic

sink and source of energy, the UPT KESS improved the island’s power
grid efficiency and increased the penetration rate of the wind turbines.
The flywheel unit’s ability to provide a stabilizing capability to the highly
variable wind turbine power was found to be essential in allowing Fuji
Electric to connect the wind turbines to the island’s relatively weak
electrical transmission system. Because of this successful outcome, Fuji
Electric is now looking for further deployment opportunities of the UPT
KESS technology to provide truly reliable wind-generated energy as a
viable supply alternative in other locations.
Example—King Island, Australia.12 A more recent example of a

storage facility assisting an isolated power grid to use wind energy can be
found on King Island, a remote island located off the southern coast of
Australia. The local utility, Hydro Tasmania, installed five wind turbines
(ranging from 250 kW to 850 kW) over time to supplement and hopefully
supplant the power from four 1.5-MW diesel generators. As with many
isolated power grids, the demand profile was volatile, with the peak power
levels often reaching 80% above the average system load. Unfortunately,
the wind resources in the area are also extremely variable, with much of
232
the wind energy not coincident with the peak power demand of the local
residents. Therefore, the desire to add a renewable energy resource to the
island’s power mix actually led to a situation where the existing diesel unit
performance declined from constantly cycling to balancing the changing
consumer demand and variable wind turbine output.
Hydro Tasmania decided to add an energy storage unit both to

improve the balancing of resources with demand and to add another
component of system stability. The storage unit selected for installation
in 2003 was a 200-kW (800 kWh) flow battery from Pinnacle VRB Ltd
(a subsidiary of VRB Power Systems). Factors contributing to Hydro
Tasmania’s decision to choose the flow battery were the unit’s rapid
charging capability (1-to-1 charge/discharge) and its round-trip efficiency
of 77%. To assist in the deployment of the unit, the Australia’s Renewable
Energy Commercialization Program provided a grant of $700,000AU.
Since the vanadium redox battery energy storage system (VRB-ESS)

unit began operating, it has provided three main benefits as it lowered
operation costs for the utility and reduced the system emissions by 47%.
First, it allowed load shifting of the off-peak wind energy to increase the
contribution of the existing wind turbines to the island’s power needs.
Second, the storage unit improved operation of the existing diesel
engines. With the additional stored energy as a resource (and the ability
to absorb temporary excess wind power), the diesel units could be
dispatched more economically, resulting in both a reduced number of
starts and reduced run-time for the units. This provided both reduced
fuel usage and lower emissions of diesel exhaust. Overall, as wind energy
increased in its share of supplying the island’s power, the total diesel run-
time had been reduced 1,100 hours per year, reducing fuel usage by more
than 1.5 million liters/year. Finally, the storage unit improved stability
through better frequency regulation, load-following, and enhanced
voltage control for the system as the variable wind resources were used
to supply more power to the island customers.
233
Grid Connected
Although wind power is quickly becoming a real solution for
remote-power and island-grid environments, the real market for
wind power development lies with large grid-connected wind parks.
To improve their unit competitiveness, the focus of much wind
turbine development has been to reach cost parity with natural gas–
fired combined-cycle units. For large modern wind turbines, that goal
has been reached; however, being competitive on price does not mean
that wind is as useful an energy resource to the system administrator.
Currently, to ensure a wind turbine’s competitive position not only
requires that the turbine be placed in an advantageous wind resource
area, but also that the demand will be there for this power (rate and
quantity), and that demand will be within easy transmission distance.
For wind that only blows hard at midnight in a rural setting, the only
competitors a wind turbine using this resource will find are baseload
coal units idling at low power—a competitor no power generator
relishes. Although not a panacea, energy storage technologies can
provide assistance tailoring the wind turbine’s operating strategy for
different markets by decoupling the production of wind power from
its immediate sale.
There are three basic strategies for coupling storage technologies

with large wind projects: dispatchable wind, capacity firming, and
baseload wind. Interest in storage does exist in the wind community,
but what wind developers want most from any type of storage
technology is something that is cheap, responsive, and reliable. As
large-scale wind parks are relatively new compared to other power
generation options, storage will be introduced slowly, building from
experience in the island grid arena to increase flexibility and alleviate
existing integration costs. In the long term, storage can be more
234
systematically integrated into wind projects as developers better

understand how to leverage a storage technology’s ability to reduce
the risk associated with wind projects, enabling a greater level of
market penetration for wind overall.
Dispatchable wind
An early strategy to improve the economics of wind power was
simply to capture all of the power from a wind turbine and sell it as
a single block into the market. By using a storage facility to totally
decouple the production and delivery of the power, the wind energy
could be made dispatchable and sold during peak demand periods as
firm power—transforming the low-value, unscheduled output of the
wind turbine into high-value, on-peak green power and ensuring the
maximum revenue for the wind park. Beyond wholesale power sales, the
strategy held out the potential for the facility to sell capacity to ISOs for
ancillary services such as contingency services or restoration services.
Storing all of the power for a sale as a single block of power (fig. 5–4)
was designed to work around some of the less useful aspects of wind
power production. First, because the power delivery is firm, there would
be no capacity payment reduction nor any additional discounting by the
utility to make up for added ancillary service costs. Second, this strategy
allows for capturing the highest peak power prices, especially in the
summer months when output (especially in the Midwest) is noticeably
reduced at the time power is most desired. With the ability to obtain
high market prices, this strategy also holds out the potential to develop
marginal resources not normally considered cost-effective. Interestingly,
although the dedicated transmission power lines are totally unused
during the majority of the day when the storage facility is charging, the
overall usage of these assets would essentially be the same as if the wind
turbines delivered their power to the grid as it was produced (if the same
size power line was used).
235
Fig. 5–4. Dispatchable wind (Ardour Capital Investments).
Although straightforward, this strategy has only seen limited interest

recently, and it is doubtful that it will garner much future development.
The reason for this is that the added value of selling a firm block of power
into the market cannot usually make up for the loss in market flexibility
and added cost of the storage component. Part of this stems from the
recent advances in wind turbine technology, which have reduced much
of the lower efficiency issues of earlier wind turbine technology that
this strategy planned to correct. Also, selling capacity for contingency
reserves or restoration services would be problematic as the full power
capacity for these ancillary services would only be available daily for a
few hours just after the storage facility was fully charged but before the
unit would discharge. Also, selling any of these services would preclude
selling the bulk power into the market. Finally, this strategy relies on
having unfettered access to sufficient transmission capacity for anywhere
from four to eight hours per day. However, depending on the regional
transmission rules, it is possible that rural areas would have insufficient
transmission capacity (contractual and/or physical) during essential
hours, constraining the facility’s output off the grid, and thus losing some
or all possible revenue for that day. If for any reason the output could
not be delivered to market, penalties in the contract (for firm power)
would also significantly reduce the profitability of the project.
236
Example—Alta Mesa, CA.13 The Alta Mesa pumped storage project

(TenderLand Power Company), which was proposed in the late 1990s for
construction in the San Gorgonio Pass near Palm Springs, California, is a
good example of an early project designed to store off-peak wind energy and
sell it into the market during peak demand. The 70-MW pumped storage
facility was to be associated with a 54-MW wind park and have a round-trip
energy storage efficiency higher than 70%. By capturing the wind energy
produced during off-peak hours, 420 MWh could be delivered during on-
peak hours to Southern California Edison’s transmission system. The facility
was to consist of an upper reservoir on the Alta Mesa and a pair of large
concrete tanks at the bottom of the hill with a hydraulic head of 1,250 feet
and a storage capacity of 113 million gallons in a closed-loop system. Part
of the initial interest in using storage on this wind power project was that
at this location, the majority of the power generated is produced during
off-peak periods, limiting the economics of the wind-only development.
By incorporating a pumped-hydro storage facility, the combined project
could sell six hours of firm green power during peak demand (and pricing)
periods, increasing the value of the intermittent wind energy.
Capacity firming
If energy storage technologies are to play a significant role in
conjunction with wind power, it will be through firming the delivery
of wind power from grid-integrated wind parks. Rather than cycling all
of the output from the wind turbines through the storage facility, the
capacity-firming strategy focuses on providing sufficient support to the
output of the wind park to ensure maximum on-peak energy sales (fig.
5–5). Many of the current negative aspects of selling wind power into
the market—lack of capacity payments, additional ancillary services,
and energy reserve support requirements—are then reduced, improving
the actual economics of the wind park. Although this strategy cannot
provide full power-dispatchable green energy resources for the grid in the
237
same way as dispatchable wind, it can significantly improve the project’s

ability to provide some level of firm power from the wind park. In this
way, the wind developers’ customers—system operators—can begin to
view wind in a far more positive light, especially as the amount of wind
energy delivered to the grid continues to climb.
Fig. 5–5. Wind capacity firming (Ardour Capital Investments).
Determining the needed power rating for a storage unit to support

a wind park for this strategy requires knowing and understanding such
issues as the size of the wind park, the variability of the local wind
resource, the local power transmission capability, and the average load
profile of that local power system. With a volatile wind resource site, it
is not uncommon for the output of the wind park to change suddenly,
possibly exceeding the capability of the transmission system to absorb
the power. Surprisingly, the required power rating of the storage unit
only needs to be a small portion of the rating of the wind park to be
useful. This is because the power rating needs to only be as large as the
expected variations in the output of the wind park from some long-term
median desired output level. This often requires sizing the power rating
of the unit at only 20% of the size of the wind park.
238
More important to the strategy of the ensuing combined wind and

storage combination is the determination of how much energy storage
(MWh) is required. For areas of relatively constant wind speeds, high
cycling energy storage facilities with only a small storage capacity may be
useful. Akin to the frequency regulation service provided on small island
grids which are constantly absorbing and discharging a larger amount
of this excess energy, they can improve the delivery of a more stable
power output from the facility, benefiting the power flow on the grid.
This would especially benefit areas with weak transmission systems. On
a more common basis and for larger wind projects, wind energy during
off-peak periods or during any time when the transmission of the output
is constrained could be stored and then used later to supplement
existing wind generation during on-peak periods to ensure maximum
energy sales of firm green power. The marginal cost of adding additional
storage capacity for most energy storage technologies is relatively cheap
(compared to the cost of the entire system in general), opening up
additional strategies. This additional power could also be released during
low- or zero-output periods of the wind park to provide minimal energy
delivery. The amount of energy storage capacity would then determine
the length of time for which output could be supported without wind
generation and help with possible reserve requirements beginning to
face wind projects. Currently, the only way to increase the capacity
factor of a wind park is to oversize the number of wind turbines for the
transmission system (with the extra wind turbines providing the reserve
capacity), but developers are loathe to follow this strategy to provide
these contractual reserve levels.
Example—McCamey, TX.14 Texas ranks second amongst the 50

states for its wind potential, and McCamey, Texas, has been the center
of much development activity because of its wind resource potential.
Unfortunately because of its remoteness in West Texas, developers are
building more wind generation than the transmission grid can easily
handle, causing congestion problems. Although plans exist for additional
transmission upgrades, new wind park development is expected to match
239
or even surpass these upgrades. To alleviate this near-term transmission

constraint problem and provide room for additional wind generation in
the area, Texas’s State Energy Conservation Office (SECO) commissioned
a study (led by the Colorado River Authority) to determine what
benefits a large-scale energy storage facility would have for transmission.
To support this expected continued mismatch between wind power
and transmission capacity, SECO chose a CAES facility with 400-MW
compression, 270-MW generation, and an extremely large 10,000 MWh
storage capacity (at full power, a 25-hour capacity for compression, and
a 37-hour capacity for power generation). The facility would be used to
store power during periods when congestion on the transmission lines
constrained the growing wind resources.
Unfortunately, because of the wind pattern in the area—there are

times when the wind blows strongly and continuously for days at a time—
the modeling of the project showed that the CAES facility could become
fully charged and unable to provide additional compression even when
it would still be required. As providing total reduction of transmission
curtailment was the single desire for the study, it was found that the CAES
facility could not alleviate 100% of the transmission constraint; thus it could
not be evaluated as a substitute for transmission line capacity. However,
extending the evaluation of the CAES facility past a purely transmission
replacement role, it was found that it could provide several benefits:
• The facility would provide more wind generation (up to 400

MW) with only minimal curtailment, and better capacity firming
of the wind parks for the area—providing assurance that far
more power could be delivered during peak demand periods.
• Combining wind with storage would ensure that wind could
claim credit for operating reserves (equal to the amount
of CAES generation). Although this was not a significant
payment, having this capability added to the total value of
the facility (the value of any facility stems from not just one
revenue stream, but many), and it provided additional firm
240
capacity for the system operator to call on. This will be needed
in the near future, especially as the amount of wind generation
continues to grow.
• By operating the CAES facility to support wind generation in
the area, wind energy curtailment reduction totaling over 600
GWh annually was achieved (compared to the area without
the storage facility). This resulted in several million dollars of
profit annually above and beyond what would be required for
a positive return on the CAES facility investment. Because of
this outcome, work continues on siting a CAES facility for this
role in the area.
Baseload wind
The third market strategy—baseload wind—incorporates storage
most closely with the delivery of wind power to the power grid. It is
an extension of the capacity-firming concept and is envisioned for the
largest scale wind storage projects, with most facilities reaching hundreds
of MWs (or more) in size. Although most wind turbine technology
and development has (rightly) focused on the production of low-cost
wind power for direct sale, this concept (long championed by Alfred
Cavallo) focuses on obtaining the maximum wind energy production
from a particular site, and then using the storage facility to optimize
the delivery of power to the market competitively (fig. 5–6). This
then creates a choice of different wind turbine designs and maximizes
their site placement because there is no concern with their output
being constrained from delivery during peak periods. By balancing and
optimizing the wind turbines, storage component, and transmission of
the power, this strategy is designed to provide wind-derived power to
markets with the same capacity and dispatch capability as a mid-merit
or even a baseload power plant—especially over long distances.
241
Fig. 5–6. Baseload wind (Ardour Capital Investments).
During normal operation, the output of the combined facility would

provide power primarily from the wind turbines, supplemented by
energy from the storage unit. Wind energy could be stored both during
off-peak periods (evenings and weekends) and during periods when the
output of the outsized wind park could be constrained from delivering
power. Because of the storage capacity required and the flexibility in
siting, the storage technology most envisioned for this would be CAES.
In this design, both the size and the capacity of the storage facility are
somewhat flexible, with the power rating of the storage unit sized to
accept the variations in output from the wind park and the facility’s
energy storage capability (10s of hours at a minimum). In this way, the
storage facility is sized in relation to the amount of capacity support
envisioned by the project—with a larger storage component translating
into a higher overall reliable delivery from the facility. With this design,
instead of an average wind park’s roughly 33% utilization rate, the mix
of wind power supplemented with CAES output could provide double
that, or more. With the significant storage capability of the CAES facility,
even seasonal storage of the wind energy would be possible.
242
Besides providing system operators a means to incorporate wind

power into the market in a reliable manner, a baseload wind project could
also be the basis for transmitting wind power over long distances. Much
of the U.S. wind potential is simply too far away from demand centers to
ever be built out. However, the reliable output of a large baseload wind
facility could provide long-distance transmission through a dedicated
long-distance high voltage power line. Because delivery of the wind park’s
power would be regulated by the storage facility, a much higher utilization
of the capital-intensive transmission system could be achieved—making
the project economical, unlike the current setup. Large, long-distance high-
voltage direct current (HVDC) transmission of site-specific power exists,
whether for hydropower from the Pacific Northwest to California or lignite
coal power from the upper Midwest to Minnesota. Although siting large
power lines—for any reason—is extremely difficult, it is not out of the
question that more such lines could be built for wind power, especially if
those involved are serious about significantly increasing the penetration
level of wind power into the overall U.S. power market.
Example—Iowa Stored Energy Project (ISEP).15 As a first step

toward a large-scale baseload wind facility, the Iowa Association of Municipal
Utilities (IAMU) is developing a utility-scale wind park that will incorporate
a large-scale energy storage facility near Eagle Grove, Iowa. The Iowa Stored
Energy Project (ISEP) is a $200-million integrated power facility, composed of
an 84-MW wind park connected to a 200-MW CAES facility. The compressed
air for the CAES plant will be stored in an underground aquifer located 1,200
feet below the surface. This would be the first use of an aquifer for a CAES
facility, as both of the existing CAES facilities use underground salt caverns.
The aquifer has already been used for natural gas storage; in fact, because of
the structure of the underground rock, both air and natural gas storage are
envisioned for the site (held in different strata); the stored natural gas is to
serve both as a fuel source for the CAES plant and as a seasonal natural gas
storage facility for IAMU members. Storing natural gas on-site for the power
facility will also alleviate a potential pipeline capacity shortage during peak
243
operating periods. The potential storage capacity of the aquifer is huge, and
is estimated to be able to store anywhere from 8 billion to 30 billion cubic
feet of compressed air at 500 psi (the Macintosh, Alabama, facility can hold
19 million cubic feet at 1,100 psi).
The IAMU members have decided to proceed with the ISEP because
the project presents a unique way to generate additional power needed
to meet annual load growth, while capitalizing on a number of inherent
strengths. Iowa has tremendous wind resources, and increasing numbers
of these projects are finally finding acceptance in the regional power
market. CAES was seen by the IAMU members as a means to capitalize
on this growing local renewable energy resource base and provide the
flexible and reliable mid-merit capability needed for customer supply. By
using renewable energy and highly efficient natural gas power generation,
the ISEP is also designed to reduce the carbon emissions of the IAMU
utilities even as their power needs grow.
The ISEP will operate as an intermediate load power plant,

producing a mix of wind and power from the CAES facility 12 to 16
hours per day, 5 days per week. During off-peak periods, the ISEP will
compress and then store air in the underground aquifer using power
from both the wind park and off-peak system power (important in the
event of insufficient wind power). Overall, the facility is expected to
maintain a utilization rate of 50%, with wind power responsible for 33%
of the total output. Combining the competitive economics of both
wind turbines and the CAES facility, the ISEP project is expected to be
cost-competitive with other intermediate power plants in the region.
Gradually, with the addition of more wind turbine generators and air
compressors, the facility could evolve into a baseload facility in both
scale and capability. One option for expansion uses 400 MW of wind
turbine capacity with the 200-MW CAES power train, which would
allow the facility to operate with a utilization rate of 75%—similar to
surrounding coal-fired baseload facilities.
244
References
1. Energy Information Agency. 2003. Renewable energy annual—2002.
2. Pacific Northwest Laboratory. 1991. An assessment of the available
windy land area and wind energy potential in the contiguous United
States. Richland WA: Pacific Northwest Laboratory.
3. Parsons, B., M. Milligan, B. Zavadil, D. Brooks, B. Kirby, K. Dragoon,
and J. Caldwell. 2003. Grid impacts of wind power: A summary of
recent studies in the United States (NREL/CP-500-34318). Golden, CO:
National Renewable Energy Laboratory.
4. National Energy Renewable Laboratory. Renewables for sustainable
village power. http://www.nrel.gov/villagepower/index.html (accessed
March 26, 2005).
5. BP Solar. 2001. Mord Project—Phase 1&2 (Application Type, Rural
Infrastructure). brochure. Fredrick, MD: BP Solar.
6. Isherwood, W., R. Smith, S. Aceves, G. Berry, W. Clark, R. Johnson,
D. Das, D. Goering, and R. Seifert. 1997. Remote power systems with
advanced storage technologies for remote Alaskan villages (UCRL-ID-
129289). Livermore, CA: Lawrence Livermore National Laboratory.
7. Shirazi, M., and S. Drouilhet. 1997. An analysis of the performance
benefits of short-term energy storage in wind diesel hybrid power
systems (NREL/CP-440-22108). Golden CO: National Renewable
Energy Laboratory. (Also presented at the ASME Wind Energy
Symposium, Reno, NV, Jan. 6–9, 1997.)
8. Ibid.
9. South Africa: Sustainable energy for Hluleka Nature Reserve. 2002. Shell
Solar Project Brief. Amsterdam, The Netherlands: Shell Solar.
10. Butler, P. 1998. Energy 100 awards—Metlakatla Energy Storage System.
Sandia, NM: Sandia National Laboratories.
11. Urenco Power Technologies. Case study, Fuji Electric, wind power
study. http://www.uptenergy.com (accessed March 26, 2005).
245
12. Hennessey, T. The multiple benefits of integrating the VRB-ESS with

wind energy producers—a case study in MWH applications. Paper
presented at the American Wind Energy Association (AWEA)
Conference, Chicago, April 2004.
13. Chippas, S. J., P. R. Blaszczyk, and M. G. Jones. 1999. Wind and water,
the Alta Mesa project. Renewable Energy World. 2 (2).
14. Study of electric transmission in conjunction with energy storage
technology. August 2003. Prepared by Lower Colorado River
Authority for the Texas State Energy Conservation Office, Austin, TX.
15. Wind, T. 2002. Compressed air energy storage in Iowa. Wind Energy
Consulting. Report prepared for the Iowa Association of Municipal
Utilities, Ankeny, IA.
6 OUR NEW
ENERGY FUTURE
Energy storage technologies have garnered interest throughout the

power industry because of the growing evidence of their expanding
capabilities. For this interest to continue, however, these technologies
require a road map for their future development to better align
themselves with the current needs of market participants and to be
capable of adapting to a changing environment. This road map cannot
dictate which customers will use storage technologies and when, but
it should align the capabilities of storage technologies with interested
parties who will drive this technological development and deployment.
To understand how these energy storage technologies will work in the
market, much can be learned from the time and effort expended to
reach this point on the development curve; and storage technologies
have played a valuable role in reaching a variety of customer goals.
248
The Road to Here

Many different groups have supported the development of energy
storage technologies. Through previous research and demonstration
projects, these groups have been responsible for the commercialization
of many of the existing available energy storage technologies. Current
programs will continue to be instrumental in commercializing the most
recently developed storage technologies now just exiting the lab and
entering pilot-phase projects.
U.S. federal government

The U.S. federal government has long supported energy storage
technologies because of their promising capabilities. Since the
establishment of the industry, public policy makers have attempted to
protect customers from volatile prices or unfair trade practices while
improving operation conditions for businesses. To this end, they evaluate
the changing economic conditions of the industry to craft public policies
and support promising technologies that will promote their long-term
goals. Energy storage technology is one such promising technology.
Capabilities of energy storage technologies—improving the

efficiency, reliability, and security of the electric power industry—are
all goals embedded within the principles of the U.S. Department of
Energy’s National Energy Plan. This and other recent policy documents
reveal renewed efforts to craft an alternative market structure, where self-
correcting forces solve today’s (and tomorrow’s) challenges where former
command-and-control measures have not succeeded. To promote the
development and introduction of new technologies like energy storage
technologies, the U.S. Department of Energy (DOE) is following the
success of such environmental regulations as the SO 2 cap-and-trade
system. Rather than planning exactly how the goals of SO 2 reductions
would be reached, the government helped industry develop a number of
O U R N E W E N E RG Y F U T U RE
249
new technologies and strategies for market participants to choose from

and let the market sort out the most cost-efficient means of reaching the
goal. The government understands that using new technology within the
marketplace provides the means to reach these goals; hence, it looks to
innovative and flexible technologies such as energy storage.
The primary area of support for large-scale energy storage technology

research has been from the DOE, but other groups with the federal
government have seen the promise of these technologies and initiated
their own related programs, including the Department of Defense and
the Department of Transportation. In fact, the U.S. government has
supported large-scale research on energy storage technologies since the
mid-1970s; even prior to the formation of the DOE, the Energy Research
and Development Administration (ERDA) was evaluating applications
for energy storage technologies. By supporting the basic research and
proof-of-concept installations at both the ERDA and the DOE, these
federal programs have supported the eventual commercialization of
these technologies by the private sector.
Now the DOE channels its support for energy storage technologies
through the Energy Storage Systems (ESS) Program. The ESS program
endeavors to develop advanced energy storage systems in partnership
with industry to minimize costs incurred from power quality and reli-
ability problems, increase technology choices in deregulated, competitive
electricity markets, and increase the value of renewable and distributed
resources. The increased emphasis and visibility that the DOE is placing
on energy storage technologies are evident by the ESS program’s inclusion
into the recently formed Office of Electricity and Energy Assurance, which
formed to coordinate the modernization of the electric power grid.
The ESS program provides a framework for promoting storage

technology in the market. By identifying customer requirements and
analyzing operations experience of field systems, the program can better
target engineering research for storage system component research
and integration. Much ESS program research is conducted through
250
Sandia National Laboratories (Sandia, New Mexico), which focuses on

consumer- and distribution-side systems. The ESS program also works
closely with industry partners, often in a cost-sharing arrangement.
This allows the federal government to support research programs at
manufacturers with imminent promise that lack the financial capability
to finish prototypes of new technology. The DOE also cooperates with
major utilities to fund technology proof-of-concepts and validation
projects. Showcasing a successful integration and operation of the
technological platform is essential to allay customer reticence toward
being the first user of any technology. Besides proving the facility actually
works, operational data can be gathered under real-world conditions to
refine the most profitable market application.
State energy programs

Recognizing the potential benefits of energy storage technologies,
a number of state governments have developed programs to evaluate
and promote the adoption of commercially ready technologies in real-
world applications. These state governments have seen the potential for
energy storage technologies either in the marketplace or in development
by the federal government, but they have structured the programs to
focus on near-term issues facing their particular state economies, such as
mandates for renewable energy portfolios or protecting and enhancing
the environment for both businesses and electric infrastructure. The
hope is that this will help new and existing firms with their power quality
issues and also new technology firms that will develop new storage
technologies within the state. Two of the most prominent of these
programs are found in California and New York.
California. The first example of a state-sponsored energy storage

technology program is found in California, where the state government
has turned to energy storage technologies to help maintain the state’s
power system reliability while it tackles a number of challenges stemming
251
from the effect of deregulation on the state’s continued load growth.

The California Energy Commission (CEC), in conjunction with the ESS
program, launched a program in 2001 to provide matching funds for
projects to deploy market-ready storage technologies that can provide
economic value within three to five years. The goals of the projects
are: to improve the operation of transmission and distribution grids by
alleviating congestion or providing voltage and frequency regulation
support; to support the dispatchability of renewable resources; and
to help with peak load reduction and load management. The three
winning technologies were announced in 2003 and, once installed, will
be monitored for three years, with the units operating for a minimum
of 18 months. These projects will receive nearly $4 million from CEC
(through a utility CEC surcharge), and the participating firms will fund
an additional $4.6 million through cost sharing.
New York. Another example of a state-sponsored energy storage

technology program is found in New York, where the state government,
through the New York State Energy Research and Development Authority
(NYSERDA), initiated a cost-sharing program in 2004 to demonstrate the
practical use of emerging energy storage technologies, develop energy
storage technologies for commercial production within the state, and
perform feasibility studies for the use of energy storage technologies in a
number of commercial roles across the state. Similar in structure to the
CEC program, the ESS program will assist with technology evaluation
and support of the cost-sharing program. The two winning technologies
were announced in 2004 and will receive $3.5 million of support from
government sources. Once installed, these projects will be monitored
through their year-long operational period. Goals of the program include
electric service reliability improvement, grid voltage support, transmission
and distribution upgrade deferral, time-of-use energy cost management,
and renewable energy generating sources capacity firming. This project
follows on a series of other NYSERDA projects to enhance the electric
power service within the state.
252
International programs
Many other governments have seen the potential of these
technologies to solve unique and intractable challenges and have
undertaken research programs (collectively and independently) to
develop energy storage technologies. Just as in the United States, these
other public policy makers see energy storage technologies as a means
to protect customers from volatile prices or unfair trade practices while
improving the conditions for business operation. At the same time, they
are establishing a base of advanced technology manufacturing—and
potentially lucrative export capability—in fields that include power
electronics, electrochemistry, and others.
International Energy Agency. Acting as a technical clearinghouse

and driving agent, the International Energy Agency’s (IEA) Efficient Energy
End-Use Technologies program contains 14 different Implementing
Agreements (IAs), or areas of research. One of these programs focuses on
energy storage. Active participants in the Energy Conservation through
Energy Storage program include Belgium, Canada, Finland, Germany,
Japan, the Netherlands, Sweden, Turkey, the United Kingdom, and the
United States. Although the interest in these technologies is widespread,
the interest in electrical energy storage technologies is being driven by
four main issues:
1. Widespread reforms in electric utility regulations

2. An increasing reliance on electricity in national economies
3. Prevalent adoption of renewable energy sources
4. Attempts to reduce the environmental impact of power
production
Energy storage technologies have garnered such interest because

of the rapid rate of improvement in the basic technology, along with
anticipated cost reduction of each unit once production commences.
253
Although much of this research has been focused on thermal energy

storage, the IEA sponsored its first specific research program (Annex)
into electricity storage with the IEA Implementing Agreement on Energy
Conservation through Energy Storage (Annex 9)1 and a follow-up research
program that has been proposed but not yet acted on (Annex 15). 2
This next research program is designed to evaluate storage’s capability
assisting distributed generation assets to realize their full potential in
reducing atmospheric emissions and mitigating climate change.
Governmental programs. Besides the support for the collective

research and development (R&D) approach through the IEA, many other
governments have established individual domestic research programs
into energy storage technologies. Besides supporting the IEA research
program, countries such as Japan, South Africa, Israel, and Australia
have developed a network of domestic energy storage technology
R&D programs.
The goals of these countries are threefold. First, just as in the

United States, the governments of these countries see energy storage
technologies providing a means to improve the production and
efficient use of domestic energy sources. Second, these governments
are interested in developing a technological base of innovative energy
technology manufacturers with export potential. Finally, the availability of
capable energy storage technologies is crucial because they help provide
the clean and reliable power quality necessary to promote higher-valued
industry and commerce across many industries.
Industry programs
Besides governmental support, many types of firms active in the
industry have played an important role promoting and supporting the
development of energy storage technologies. In particular, the Electric
Power Research Institute (EPRI) has long promoted these technologies,
acting as a driver and clearinghouse for technical information.
254
EPRI has been instrumental in many of the first pilot projects of many of
the different storage technologies as they sought to show their value in
a number of different applications. EPRI was developed in 1973 (through
taking over previous R&D research conducted by the Edison Electric
Institute [EEI] and the Electric Research Council) as an industry-led
organization to conduct electricity-related research and development
programs. One of its latest, the Electricity Technology Roadmap
Initiative, represents a collective vision of the opportunities for electricity
to serve society in the 21st century through advances in science and
technology. In the technology road map, energy storage technologies
are highlighted as enabling technologies that strengthen the power
delivery infrastructure, foster a revolution in consumer services, and
boost productivity and prosperity. As part of its ongoing support, EPRI
launched a research program into energy storage technologies and their
ability to assist with ongoing transmission and distribution concerns.
One of the first stages in this continuing effort was the publication of a
detailed technical report on the impact energy storage technologies can
have on the transmission system entitled, EPRI-DOE Handbook of Energy
Storage for Transmission and Distribution Applications.3
Outside of the United States, other similar industry groups have

supported research into energy storage technologies and their applications
over the years. In Japan, the Central Research Institute of Electric Power
Industry (CRIEPI) was established in 1951 as a comprehensive research
organization for the electric power industry, with a mission to conduct
fundamental, pioneering research on the electric industry. In each of
the three central objectives of the group (cost reduction and ensuring
reliability, creation of integrated energy services, and harmonizing energy
and environmental priorities), research into energy storage technologies
has been conducted to provide benefits to those in the industry.
Besides these industry-wide groups, individual utilities across the

globe have provided key support through their proprietary programs to
evaluate a variety of storage technologies across a number of applications.
255
Utilities such as American Electric Power, Detroit Edison, Entergy, Eskom,

RWE, and Tokyo Electric Power Company (among many others) have
supported energy storage technologies and their promise as solutions to
some of the more intractable problems not currently being met by other
technologies. The programs underway promise significant savings and
additional flexibility in system operations. In previous years, this support
has had far-ranging impacts, so the expectation for today’s programs is
significant. Throughout the 1960s and 1970s, many U.S. utilities reacted
to the rapidly rising cost of fuels by searching for a means to arbitrage
some of their off-peak baseload power to daytime hours. They began
building pumped-hydro units, and pumped-hydro units expanded their
share of the U.S. peak capacity dramatically. As these technologies went
into service, additional applications were discovered that increased the
value of these installations even more. Unfortunately, as the market
changed, the capability of these utilities to build additional similar
units reduced—hence the interest in other large-scale energy storage
technologies that have a smaller environmental footprint. Without any
renewed support for large-scale storage projects, the proportion of these
facilities in the overall generation mix has fallen, reducing their beneficial
impacts (fig. 6–1).
Fig. 6–1. U.S. large-scale energy storage penetration declines

(Ardour Capital Investments).
256
User’s Goals for

Storage Technologies
Storage technologies have been gaining visibility and interest with
the R&D community as well as government policy makers, utilities, and
ISO officials. For storage technologies to penetrate different segments
of the market and be more widely used, however, their capabilities
must continue to align with the goals of the individual players in the
market. This is an important aspect for any technology type, but it is
particularly vital to technologies such as energy storage, where the
focus is on enabling existing operations and strategies to function more
efficiently and reliably in a volatile environment such as the electric
power industry. To date, energy storage technologies have been able
to provide support for the changing goals of market participants, and
their flexible operating capability ensures continued consideration as a
solution to many future challenges.
Commercial and industrial

As firms increase their use of information technology and precision-
controlled equipment, or just face a pared-down work environment with
little slack in the schedule, poor power quality is becoming a central topic.
The goal of commercial and industrial firms is to prevent their energy
usage from overtly affecting their operations. Energy storage technologies
currently assist by protecting against both power sags and outages in
UPS products as a ride-through resource. Increasingly, a number of other
energy storage technologies are following the lead of thermal energy
storage technologies and providing a peak-shaving capability to these
firms to reduce their energy costs as well as provide more control over
their energy resources, ensuring that cost-saving measures like shifting
activities to off-peak periods are kept to a minimum to reduce the
impact on operations.
257
Utilities
Utilities have supported the development and deployment of energy
storage technologies because these technologies promise to overcome
problems with added peak capacity, system stability, transmission
deferral, renewable integration, and customer power-quality issues. This
continued interest has been supported by the flexible capabilities of
these technologies to fit the changing needs of utilities. In the 1970s
and 1980s, the market was dominated by large, vertically integrated
utilities that needed a means to provide additional power resources
during peak times cheaply—especially as a diurnal sink for the growing
fleet of nuclear power plants. Although a number of technologies
were evaluated (superconducting magnetic energy storage [SMES],
compressed air energy storage [CAES], and pumped-hydroelectric
storage [PHS]), pumped-hydro storage is becoming the most widely
deployed. The 1990s brought the demand for more distributed facilities
with a multifunctional capability to provide transmission-system stability
and possibly asset deferral. Now utilities are increasingly evaluating other
storage technologies that not only follow this multifunctional approach
but increase their ability to be used as a working storage capacity (high
cycling capability) to help provide stability in an unbalanced environment.
Utilities are also envisioning storage technologies as a solution to the
challenge of integrating power from wind farms as their penetration of
their market increases.
Energy storage technologies hold great potential for their own uses
as well as for those of other parties dealing with transmission system
stability, such as ISOs and RTOs. Besides providing system stability for
normal operations, they are assisting in the formation of ancillary services
by creating much-needed price visibility. This came out of the experience
with pumped-hydro facilities, where once they were installed to provide
commodity energy arbitrage, they proved very capable of providing
additional capabilities for the vertical utility. These other capabilities were
soon highly prized by these utilities—including frequency regulation and
258
contingency reserves. Now (and in the future), ISO and RTO development
provide price discovery for ancillary services. This is important because
in a vertical utility, all of the value was absorbed into the overall cost
structure of the utility (that did not know the real value and costs of
these services or its own marginal cost of service).
Finally, utilities have continued to view storage technologies as a

solution to improving customer satisfaction. Beyond improving the
quality of the power, these technologies can also help reduce the cost
of delivering that power. For answers to both of these, utilities seek to
install storage technologies either on their distribution system or at the
customer site. An example of this is the continued support for thermal
energy storage. The utilities can reduce energy costs for customers and
help reduce peak demand, which allows them to prevent additional
generation, transmission, and stability issues.
Energy service companies

Just as utilities are looking to storage technologies to improve the
delivered power quality for customers, a growing numbers of energy
service companies (ESCOs) are gaining interest in storage technologies to
improve their customers’ service quality. Many of these companies have
three main goals for their efforts here. First, energy storage technologies
can help ESCOs extend their service offering to their commercial and
industrial end-use customers. Second, energy storage technologies
provide for a demand for power electronics from a subsidiary or partner
firm. Finally, energy storage technologies create additional demand for
the firm’s professional engineering services, whether it is site development
in the initial stages or a continuing revenue stream through follow-on
system maintenance contracts.
259
Regulators
Public-policy regulators also deem energy storage technologies to
be tools for promoting their goals within the electric power industry,
such as enabling a transition to a self-correcting market instead of one
based on command and control regulation regime. By acting as a shock
absorber, energy storage technologies provide optionality to all actors
throughout the industry. In this way, energy storage technologies solve
one of the regulators’ greatest challenges—the need to find solutions
for certain players that can be leveraged to other groups. For example,
by helping utilities find an easier and inexpensive means to provide
extended and enhanced transmission expansion, customers benefit
through lower cost and higher service level, such as in the wholesale
market by providing additional power resources during peak power. By
supplying power throughout these stressed periods, even a small increase
in resources can have a dramatic impact on power’s spot price (fig. 6–2).
The cumulative effect is that adding extra facilities spaced throughout
the system actually provides an even greater capability to react to short-
term power imbalances, which increases the effectiveness of large-scale
energy storage facilities.
Fig. 6–2. Energy storage reduces peak prices (Ardour Capital Investments).
260
A Road Map for

the Future
To continue the successful development of these energy storage
technologies, three core issues must continue to be addressed to ensure
the continued penetration of storage technologies into the power
market. These include improving the technology, easing its deployment,
and ensuring that the value in installing a storage unit continues to
outweigh alternatives.
• Technology. Because of the number of different technologies

included in energy storage technologies, the level of maturity
ranges from just becoming commercially viable to mature
with many years, even decades, of active use. However, all of
these storage technologies will need to continue advancing
their state of art, not only to improve their existing operating
capabilities but also to provide the capacity to operate in
new applications. Besides increasing the capability of the
technology, increasing the level of reliability in operation
is also vital.
• Deployment. Frequently customers are keenly interested

in a particular technology, yet lack the ability to justify its
initial costs or overcome other installation requirements.
Government support and incentives have been crucial for
driving the introduction of renewable energy and energy
efficiency technologies; including energy storage technologies
into these existing programs would yield significant results.
• Value. The first aspect of improving the value of any
technology is to reduce the initial and operating costs of
the technology, which have a large impact on life-cycle
costs. These costs are reduced through design engineering,
261
manufacturing processing, and the scale of production.

Other improvements, such as the ease of use, are important
for the value others place on the technology—not just
the consumer, but also system integrators contemplating
incorporating a storage component into their overall system
design. Finally, asset leveragability is important; when more
components of the electric power industry are able to improve
their own performance through coordination with a storage
asset, it increases the value that asset has in the market.
Although the effort toward these areas of development can be driven

primarily by the developer (in the case of technological development
and some improvement in the ease of use), finding improvements in
the deployment of the technology often will come through aligning the
capabilities of storage technologies with the goals of other groups in
the industry. This support can come from a variety of sources, including
federal, state, and industry groups that see the capabilities of energy
storage technologies furthering their causes, which will in turn help
storage technologies overcome eventual hurdles.
Federal support
Learning the lessons of what storage can offer from other markets,
federal policy makers are incorporating storage technologies into their
plans to first move the market past its current challenges, and then to
help maintain a self-correcting, competitive market once there. Because
of the interaction of storage technologies with the electric power
industry, this support will come in many forms as needed to assist with
the development of the technology, providing a better market and
reception for its deployment.
One has only to look as far as wind energy and energy efficiency
technologies to see the use of federal support for a technology that has
other attributes (in these cases, no emissions or fuel usage) that support
262
goals valued by the federal government. This support (such as pilot

projects to demonstrate the technology) has been crucial in developing
alternatives to the existing, limited energy options. By providing support
for commercialization of these and other advanced technologies, the
federal government hopes to provide the market with a variety of
tools to use their existing assets in new and cost-effective energy usage
strategies to meet tomorrow’s challenges.
The federal government has also had a significant role in deploying

energy technologies into the market. Besides helping with accelerated
depreciation of technology investment, or investment tax credits, the
federal government has even assisted many technologies with export
assistance. Other federal support, especially for the transmission market,
includes recognizing the immense obstacles that exist for alternative
technologies for existing T&D upgrades and promoting alternatives
evenhandedly, so as not to skew investment decisions away from
alternative strategies. Proponents of energy storage technologies suggest
a number of steps to encourage the introduction of energy storage
technologies into the transmission market:4
• Treat energy storage facilities the same as traditional

transmission facilities for capacity expansion purposes
(pricing incentives and planning processes).
• Have incentives to correct technical deficiencies separate from
the regulatory process of forming RTOs.
• Prevent discrimination against or imposed disadvantages to
alternatives to conventional transmission investments.
• Allow development costs for new technologies to
be recovered in transmission rates to recognize the
grid-wide benefits a strategically placed energy storage
facility can provide.
• Award firm transmission rights and congestion revenue rights
to parties that contract with energy storage service providers for
263
services that substitute for or augment transmission capacity.

It is also possible to award these rights to parties that invest
directly in energy storage systems that reduce congestion
or otherwise produce transmission system benefits, such as
permitting the deferral of transmission system investment.
• Do not prescribe the qualifying technology nor define
narrowly the categories of improved performance that the
Federal Energy Regulatory Commission (FERC) will reward.
State support
Although most attention points toward the federal government when
the subject of governmental support for technology commercialization is
mentioned, state governments have an important and increasing role to
play. For technology development, for example, California (through the
CEC) and New York (through NYSERDA) have supported near-commercial
energy storage technologies by providing R&D grants. Because of the
success of these programs in improving the quality of electric service to
businesses and residential customers (plus encouraging high-tech energy
technology firms to locate within the states), other state governments
are looking to emulate California’s and New York’s successes with similar
efforts of their own. Besides technology development assistance, state
governments are also helping to deploy these technologies by supporting
demonstration projects in real-world operating environments, and
by lowering installation costs though incentives such as property tax
reductions or other site cost advantages.
Even regulatory reform does not fall totally under the purview of
the federal government. Each of the individual public utility commissions
(PUCs) still controls much of the activity for serving customers with reliable
and low-cost power. By influencing and supporting the introduction
of energy storage facilities among utilities’ distribution systems, PUCs
potentially have even a greater influence on energy storage facility usage
for system stability and asset deferral than the federal government.
264
One important means of supporting the introduction of energy storage

technologies is to promote their use through state energy codes. In
California, Title 24, Part 6 of the California Code of Regulations deals with
Energy Efficiency Standards for Residential and Nonresidential Buildings.
These regulations were established in 1978 in response to a legislative
mandate to reduce California's energy consumption. The standards
are updated periodically to incorporate new energy technologies and
methods that promote these goals. The most recent changes, to take
place October 1, 2005, will consider time-dependent valuation of energy,
which in effect will favor peak energy saving measures over off-peak
measures. These changes will give energy storage technologies that
allow peak shaving significantly more support when being evaluated by
a contractor during installation or refurbishment of a building. These
building code changes have had a substantial impact on California’s
use of electricity. According to the state, California's building efficiency
standards (along with those for energy efficient appliances) have saved
more than $36 billion in electricity and natural gas costs since 1978. It is
estimated the standards will save an additional $43 billion by 2013.5
Industry support
Besides governmental support and incentives, industry groups
provide significant support and incentives to promote expansion of
energy storage technologies in the market. Besides the support from EPRI
for R&D and demonstration programs, other groups have been central
to the development of energy storage technologies for some time. In
particular, the Electricity Storage Association (ESA) has provided crucial
technical and developmental support toward the commercialization of
a wide variety of storage technologies for more than 14 years, and other
groups such as the Energy Storage Council recently formed to advance
public policy support for energy storage technologies.
To advance the presence of energy storage technologies throughout

the market, however, energy storage technologies must be taken up
265
by other industry groups and be seen as a means toward advancing

their own goals. One successful example of industry support for
energy storage technology lies in the commercial building market. The
Leadership in Energy and Environmental Design (LEED) program is an
excellent example of industry leaders recognizing the benefits of using
thermal energy storage systems in their building designs. At the heart
®
of LEED is the Green Building Rating System —a voluntary, consensus-
based national standard for developing high-performance, sustainable
buildings. Through the LEED program, green building is defined as a
common measurement standard to promote integrated, whole-building
design practices. It provides a complete framework for assessing building
performance and meeting sustainability goals, including energy efficiency.
With thermal energy storage systems in the design, buildings may qualify
for points in the “Energy and Atmosphere” section of the LEED’s Green
®
Building Rating System toward the overall total score of the building.
A higher score ensures not only lower overall operating costs of the
building, but corresponds to a healthier building environment, and thus
a greater demand for space in the building by potential tenants.
Many other potential areas exist for industry support for energy
storage technologies; some of the most prominent being renewable
energy proponents such as the American Wind Energy Association
(AWEA). Providing support for renewable energy production has long
been a goal of energy storage technology developers. Gaining the support
of groups such as the AWEA would greatly accelerate the introduction
of energy storage technologies to a host of potential installation
opportunities. However, just as with the LEED program, energy storage
technologies must continue to demonstrate that coupling an energy
storage facility with a wind turbine accomplishes some goals better than
what the turbine could achieve on its own. As explained in chapter 5,
“Renewable Energy and Storage,” there are many installations where wind
turbines operate successfully without a storage facility. Storage developers
should focus their initial attention on the technically or economically
marginal installations where renewable energy resources can only be
266
fully developed by adding energy storage facilities. If storage technology

developers are able to demonstrate a means for wind developers to
accomplish some goal that they were previously not able to accomplish
without a storage component, then wind developers will be only too
happy to include storage units in their planning, and will look for areas
where they can be profitably incorporated into their plans.
Hurdles to overcome
Going forward, energy storage technologies as a class will need
to overcome many obstacles to both continue developing their
competitiveness and increase their market penetration. Some of the
major hurdles include the following:
• Raising the visibility of the technology and its capabilities

• Proving the economic competitiveness of storage technologies
• Learning how to capture the value streams created
by the facility
• Avoiding the trap of having the hype surrounding the
technology overtake its realistic near-term capabilities and
disappoint potential clients and adopters
Visibility. Gaining visibility (for its capabilities, its potential impacts,

or even its existence) is important for any technology hoping to expand
its presence in the market. One reason storage technologies have rarely
garnered the spotlight so far is that most installations have operated
behind the scenes in a system support, or standby role. Outside of the
backup power roles of the ubiquitous battery, even the load-leveling
roles of the pumped-hydro facility or commercial thermal storage units
are not designed to operate as stand-alone sources of energy, but rather
to extend the capabilities of other components by providing additional
resources during the most capacity-constrained, highest-cost period of
267
the day. Even when a storage unit is designed to act as a bridging power
unit, it is only designed to maintain supply until a backup generator unit
can be brought online.
Economics. Another hurdle energy storage technologies must

overcome is justifying their investment economically. Proponents
of energy storage technologies must prove an investment in these
technologies is more valuable than allowing the current situation to
continue, either through increasing revenue through active working
storage or by reducing costs through T&D asset deferral or retail loss
prevention. This proof must also be explained through conventional
industry metrics used to evaluate any investment in technology, namely
that it can provide a three-year (or less) payback and provide a reduction
in life-cycle costs. Although up-front costs will always remain the main
decision point for purchases, life-cycle costs are an important metric for
firms able to take a variety of issues into consideration when purchasing
technology. Finally, proponents of energy storage technologies need
to show examples of follow-on value creation from the investment in
energy storage technologies. This can include things like finding new
or expanded uses for the technology once it has been installed. For
example, many pumped-hydro facilities were built to provide commodity
arbitrage, but once installed, an ancillary service market developed as a
means to monetize additional capabilities such as frequency regulation,
contingency reserves, and so forth. If the technology proves beneficial for
additional parties, these groups may even subsidize the purchase. Here,
thermal energy storage technologies were installed primarily to reduce a
building’s cooling load costs through peak shaving. However, this proved
very beneficial to utilities in their attempt to reduce their load growth, so
they now subsidize the investment in these storage technologies.
Multiple benefits to multiple beneficiaries. Another hurdle to

overcome is that energy storage technologies’ biggest strength is also
their greatest challenge to widespread adoption—they provide multiple
benefits to multiple beneficiaries. Most storage facilities (at all size levels)
are multifunctional because they can absorb or discharge power at
268
various rates. Because of the integrated nature of the power industry, all
of the benefits of operating an energy storage facility do not fall directly
to the owner. However, the final purchase or investment decision for
any power industry technology is generally made on its capability in one
function, not a basket of capabilities. Therefore, at least one capability
must be more valuable than alternative applications—and enough so
to warrant the purchase. Another aspect of this problem is that many
of the applications of energy storage technologies are either not well-
defined, or are easily captured by the owner of the facility. For example,
in the wholesale power market, many of the benefits of a cache of stored
power are because of the enhanced security of the network. Although
some of this can be captured through the emerging ancillary service
market, other benefits remain spread out among all market participants.
For example, utilities looking to upgrade a power line only evaluate the
capability of the energy storage facility to defer the upgrade versus that of
capacity expansion—and they do not take into account the subsequent
benefits toward system stability to later operations the energy storage
facility brings.
Avoid being the “next big thing.” Finally, although energy storage
technologies hold out great potential for widespread capabilities
throughout a number of invaluable applications, energy storage
technologies could remain one of those perennial “next big things.” As
seen frequently in the past, new technologies like this are often heralded
as the next “paradigm-busting disruptive technology” that will totally
revolutionize the industry. Often, the most hard-core boosters for these
new technologies envision the industry changing to its very foundation in
only a matter of years. Unfortunately, the sheer size and complexity of the
electric power industry are well-known. Even those nominally in charge
of the system have scant control over how it truly operates—hence the
push for massive investments for an intelligent grid with distributed
reactive and proactive capabilities. Without such an understanding of its
intended market, many of these paradigm-changing technologies fail to
live up to their hype and become harder to site—even if it would be a
beneficial and profitable placement.
269
Energy storage technologies have a real opportunity to avoid this

trap and become integral components of the electric power value chain.
Many next-big-thing technologies rely on being the center of activity for
whatever system they become involved in. Energy storage technologies
are different because they provide multiple benefits to multiple
beneficiaries. They not only enable more than one activity to operate
more efficiently and reliably, but they also provide widespread enhanced
stability and security benefits. Finally, because these technologies
promote an evolutionary rather than a revolutionary change to how the
power industry operates, they are far more readily adoptable, and they
make sense now—and in the future.
Right now, companies need a way to save money and free up

capital for other desperately needed projects. Expanding the presence
of energy storage systems now will help alleviate existing physical market
constraints, improve overall operating economics, and provide relief to
customers faced with increasingly unpredictable power quality and prices.
In the future, these same firms will look for flexible tools to help them
undertake new and creative business models that will emerge through
competition. After the power market has transitioned to its future form
(or continued to change as it has in the past), energy storage assets
will still provide flexibility to market participants, freeing up resources
for other purposes. By increasing the level of optionality in the market,
storage will reinforce the market’s self-regulating ability to reach policy
makers’ goals of a more efficient, reliable, and secure power industry. In
some roles, large-scale energy storage technologies, acting as a shock
absorber, will optimize existing generation facilities and stabilize power
flows. In other uses, energy storage technologies prevent the catastrophic
loss of vital information or disruption of an entire assembly line.
Energy storage technologies may not be the ultimate disruptive

technologies with paradigm-busting results, but they are real technologies
providing real results for real users.
270
References
1. Collinson, A. 2000. Electrical energy storage technologies for utility
network optimization. IEA Implementing Agreement on Energy
Conservation through Energy Storage (Annex 9). Paris, France:
International Energy Agency.
2. Distributed generation and electrical energy storage. IEA Implementing
Agreement on Energy Conservation through Energy Storage
(Annex 15) (Proposed). Paris, France: International Energy Agency.
3. EPRI-DOE handbook of energy storage for transmission and distribution
applications. 2003. (EPRI Report # 1001834). Palo Alto, CA: Electric
Power Research Institute.
4. Bowe, J., and J. McLeod. Submitted March 13, 2003. Comments of
the Energy Storage Council on notice of proposed policy statement
concerning establishment of incentives to promote efficient operation
and expansion of the electric transmission grid (FERC Docket
No. PL03-1-000).
5. California Energy Commission. Title 24, Part 6, of the California Code
of Regulations: California’s energy efficiency standards for residential
and nonresidential buildings. http://www.energy.ca.gov/title24/
(accessed March 27, 2005).
About the Author
Richard Baxter is a Senior Technology Analyst with Ardour Capital
Investments, an investment bank specializing in energy technologies and
alternative energy markets. In this role, Richard specializes in evaluating energy
technologies and their related business models. He is the author of numerous
reports, industry journal articles, and is an accomplished public speaker. Before
joining Ardour, Richard was the Director of Member Affairs at the Energy
Storage Council where he was involved in raising the visibility of energy storage
technologies for federal and state government officials. Prior to this, Richard led
the electric power research at the Yankee Group, evaluating growth prospects
for new technology firms and providing due diligence services to the financial
industry. Richard also spent time at the Standard & Poor’s DRI Energy Group
where he was responsible for key environmental and electric power components
of the U.S. electric power forecasts for senior utility executives on emerging
technical and market changes. Richard has a M.S. degree in Energy Management
and Policy from the University of Pennsylvania, a B.S. in History from Tennessee
State University, and a B.S. in Materials Engineering from Virginia Tech.
Acronyms and
Abbreviations
AC Alternating current
ATC Available transfer capability
Btu British thermal unit
CAES Compressed air energy storage
CAO Control area operator
C&I Commercial and industrial
DC Direct current
DG Distributed generation
DOD Depth-of-discharge
DOE Department of Energy
DR Demand response
DSM Demand side management
EEI Edison Electric Institute
EIA Energy Information Agency
EPA Environmental Protection Agency
EPAct Energy Policy Act
EPC Engineering, procurement, and construction
EPRI Electric Power Research Institute
ERCOT Electric Reliability Council of Texas
ESA Energy Storage Association
ESC Energy Storage Council
ESS Energy Storage Systems
FACTS Flexible AC current transmission system
FERC Federal Energy Regulatory Commission
GW Gigawatt
GWh Gigawatt hour
HTS High-temperature superconductivity
HVDC High voltage direct current
IEA International Energy Agency
ISO Independent system operator
kVA Kilo-volt-ampere
kW Kilowatt
kWh Kilowatt-hour
xiv
LDC Local distribution company

LTS Low-temperature superconductivity
MISO Midwest independent system operator
MJ Mega joule (1MJ=0.28 kWh)
MRO Maintenance, repair, and operation
MVA Mega-volt-ampere
MVAR Megavars
MW Megawatt
MWh Megawatt hour
NAS Sodium sulfur
NEP National Energy Policy
NERC North American Electric Reliability Council
NETL National Energy Technology Laboratory
NGA Natural Gas Act
NiCd Nickel cadmium
NREL National Renewable Energy Laboratory
NOx Nitrous oxides
NYMEX New York Mercantile Exchange
O&M Operation and maintenance
OMB Office of Management and Budget
ORNL Oak Ridge National Laboratory
PCS Power conversion system
PF Power factor
PHS Pumped-hydro (electric) storage
PJM Pennsylvania–New Jersey–Maryland Interconnection
PUC Public Utility Commission
PURPA Public Utility Regulatory Policy Act
R&D Research & development
RTO Regional Transmission Organization
SMES Superconducting magnetic energy storage
SNL Sandia National Laboratory
SO2 Sulfur dioxide
T&D Transmission & distribution
Tcf Trillion cubic feet
TES Thermal energy storage
UPS Uninterruptible power supply
US United States
VAR Volt-ampere reactive
VRLA Valve-regulated lead acid
WTG Wind turbine generator
Bibliography
Akhil, A., S. Swaminathan, and R. Sen. Cost Analysis of Energy Storage for Electric
Utility Applications, Albuquerque, NM: Sandia National Laboratories,
Feb.1997, Report SAND97-0443.
Akhil, A., and S. Kraft. Battery Energy Storage Market Feasibility Study,
Albuquerque, NM: Sandia National Laboratories, Sept. 1997, Report
SAND97-1275/1.
Alt, J., M. Anderson, and R. Jungst. Assessment of Utility Side Cost Savings From
Battery Energy Storage, Albuquerque, NM: Sandia National Laboratories,
1995, Report SAND95-1545C.
Andersson, L., M. Nakhamkin, and R. Schainker. AEC CAES Plant Delivers on Its
Promises: Plant Performance and Lessons Learned, presented at PowerGen
International Conference, Dallas, Texas, Dec. 1997.
Annual Energy Outlook—2004, U.S. Department of Energy, Washington, DC, 2004.
Annual Energy Review—2002, U.S. Department of Energy, Washington, DC, 2004.
Baxter, R., and J. Makansi. Energy Storage: The Sixth Dimension of the Electricity
Value Chain, Pearl Street, Inc., St. Louis, Missouri, Aug. 2002.
Baxter, Richard. “Electric Storage,” Electricity and Natural Gas Markets: Understand
Them! edited by Bob Willett, John Wiley & Sons, New York, 2003.
—. “Energy Storage: The Missing Link of the Electricity Value Chain,”
Energy Storage Council White Paper, St. Louis, MO, July 2002.
274
—. “Energy Storage in Today’s Power Market,” World Power 2003,

Isherwood Production, London, UK, Mar. 2003.
—. “Energy Storage: Enabling Renewable Energy,” Renewable Energy World,
James & James, London, UK, July 2002.
Beckman, K., and P. Determeyer. “Natural Gas Storage: Historic Development
and Expected Evolution,” GasTIPS, Gas Research Institute (GRI),
Washington, DC, June 1997.
Beno, J., J. Herbst, and R. Hebner. “Composite Flywheels for Energy Storage—
Design Considerations,” presentation at EESAT 2002, San Francisco, CA,
Apr. 2002.
Blackaby, Nigel. “VRB Technology Comes to the Fore,” Power Engineering
International, PennWell Publishing, Tulsa, OK, May 2002.
Blankenship, Steve. “Technology Improving DG’s Potential,” Power Engineering,
Jan. 2002.
Boyes, John. Energy Storage Systems Program Report for FY99, Albuquerque,
NM: Sandia National Laboratories, 2000, Report SAND2000-1317.
Boyes, John. “Flywheel Energy Storage & Superconducting Magnetic Energy
Storage Systems,” presentation at the IEEE Power Engineering Society
Summer Meeting, Seattle, WA, July 19, 2000.
—.“Overview of Energy Storage Applications,” presentation at the IEEE
Power Engineering Society Summer Meeting, Seattle, WA, July 19, 2000.
—. Overview of Energy Storage Applications, Albuquerque, NM: Sandia
National Laboratories, June 2000, Report SAND2000-1040C.
—. Technologies for Energy Storage Flywheels and Superconducting
Magnetic Energy Storage, Albuquerque, NM: Sandia National Laboratories,
June 2000, Report SAND2000-1041C.
Bradshaw, D., and M. Ingram. “Pumped Hydroelectric Storage and Compressed
Air Energy Storage,” presentation at the IEEE Power Engineering Society
Bradshaw, D., M. Nakhamkin, R. Moritz, D. Hargreaves, R. Wolk, and M.
DeLallo. “12 MW Cascaded Humidified Advanced Turbine (CHAT)
Plant with 45–46% Efficiency Is Ready for Demonstration,” presented at
PowerGen International Conferencer, New Orleans, LA, December 1999.
B I B L I O G RA P H Y
275
Butler, P. Battery Energy Storage for Utility Applications: Phase I—Opportunities

Analysis, Albuquerque, NM: Sandia National Laboratories, Nov. 1995,
Report SAND95-2605.
Butler, P., J. Iannucci, and J. Eyer. Innovative Business Cases for Energy Storage In
a Restructured Electricity Marketplace: A Study for the DOE Energy Storage
Systems Program, Albuquerque, NM: Sandia National Laboratories, 2003,
Report SAND2003-0362.
Butler, P., J. Miller, and P. Taylor. Energy Storage Opportunities Analysis Phase
II Final Report: A Study for the DOE Energy Storage Systems Program,
Albuquerque, NM: Sandia National Laboratories, 2002, Report
SAND2002-1314.
Callahan, T., J. Degnan, and D. Miller. “Upgrading the Taum Sauk Pumped
Storage Project,” presentation at the Waterpower XIII meeting, Buffalo,
New York: HCI Publications, 2003.
Cavallo, Alfred J. “Energy Storage Technologies for Utility Scale Intermittent
Renewable Energy Systems,” Journal of Solar Energy Engineering, 123, no. 4,
Nov. 2001.
—. “High-capacity factor wind energy systems.” Journal of Solar Energy
Engineering, Transactions of the ASME, 117: 137–143, 1995.
Cavallo, A.J., R.H. Williams, and G. Terzian. “Baseload Wind Power from the
Great Plains for Major Electricity Demand Centres.” Center for Energy
and Environmental Studies Report, Princeton University, Princeton, NJ, 25
pp. 1994.
Cavallo, A. and M.B. Keck. “Cost-Effective Seasonal Storage of Wind Energy,”
Wind Energy, SED-vol. 16 [Musial, W.D., S.M. Hock, and D.E. Berg (eds.)].
Book No. H00926-1995, American Society of Mechanical Engineers, 1995.
Clarke, Steve. “Observations on Building a Flow Battery Company,” presentation
at the Electricity Storage Association Conference, 2004.
—. “Introducing Cerium Based High Energy Redox Batteries,” presentation
at EESAT 2003 Conference, San Francisco, CA, 2003.
Commercial Potential of Natural Gas Storage in Lined Rock Caverns (LRC),
Federal Energy Technology Center, U.S. Department of Energy,
Washington, DC, Nov. 1999.
276
Corey, G., L. Stoddard, and R. Kerschen. Boulder City Energy Storage Feasibility
Study, Albuquerque, NM: Sandia National Laboratory, 2002, Report
SAND2002-0751.
Cowart, R. “Efficient Reliability: The Critical Role of Demand-Side Demand
Resources in Power Systems and Markets,” The National Association of
Regulatory Utility Commissioners (NARUC), June 2001.
Crotogino, F., K. Mohmeyer, and R. Scharf. “Huntorf CAES: More than 20 Years
of Successful Operation,”presentation at ASME Spring 2001 Meeting,
Orlando, FL, Apr. 2001.
Daley, J., R. Loughlin, M. DeCorso, D. Moen, and L. Davis. “CAES—Reduced to
Practice,” presentation at the ASME Turbo EXPO 2001, June 2001.
DeAnda, M., and J. Miller. “Reliability of Valve-Regulated Lead Acid Batteries for
Stationary Applications,” presentation at EESAT 2002, San Francisco, CA,
Apr. 2002.
DeAnda, M., J. Miller, P. Moseley, and P. Butler. Reliability of Valve-Regulated
Lead Acid Batteries for Stationary Applications, Albuquerque, NM: Sandia
National Laboratories, 2004, Report SAND2004-0914.
DeAnda, M., John D. Boyes, and W. Torres. Lessons Learned from the Puerto
Rico Battery Energy Storage System, Albuquerque, NM: Sandia National
Laboratories, Sept. 1999, Report SAND99-2232.
Deb, R., L. Hsue, A. Ornatsky, and J. Christian. Surviving and Thriving in the
RTO Revolution, 2001. http://www.energyonline.com/reports/RTO_
Revolution.pdf
Deb, Rajat K. “Operating Hydroelectric and Pumped Storage Units in a
Competitive Environment,” The Electricity Journal, April 2000, p 24–32.
Deb, R., P. Wagle, and R. Macatangay. “Generation and Transmission Investment
in Restructured Electricity Markets,” The Environmental Monitor, 2002.
http://www.energyonline.com/reports/GTinvestments.pdf
Devries, Tim. “System Justification and Vendor Selection for the Golden Valley
BESS,” presentation at EESAT 2002, San Francisco, CA, Apr. 2002.
Dietert, J. A., and D. A. Pursell. Underground Natural Gas Storage, Simmons &
Company International, Houston, TX, June 28, 2000.
Distributed Generation, edited by Borbely, Anne-Marie, CRC Press, New York, 2001.
277
Drouilhet, Steve M. “Energy Storage for Hybrid Village Power Systems,”

presentation at the Village Power ‘98 Technical Workshop, 1998.
—. “Power Flow Management in a High Penetration Wind-Diesel
Hybrid Power System with Short-Term Energy Storage,” presentation at
Windpower ’99, Burlington, VT, June 20–23, 1999.
Eisenstat, L., and M. Perlis. Price Volatility in Wholesale Electricity Markets:
Lessons Learned from the Midwest Price Spikes, Electric Power Supply
Association (EPSA), Washington, DC, Sept. 1998.
Electric Power Annual 2002, Volume 1 & 2, U.S. Department of Energy,
Washington, DC, Nov. 2003.
Electricity Storage Association. Web site, http://www.energystorage.org.
Electricity Technology Roadmap: 2003 Summary and Synthesis, Electric Power
Research Institute (EPRI), Palo Alto, CA, 2003.
Energy Infrastructure—Electricity Transmission Lines, Edison Electric Institute,
Washington, DC, Feb. 2002.
Energy Infrastructure—Getting Electricity Where It Is Needed, Edison Electric
Institute, Washington, DC, June 2001.
EPRI-DOE Handbook of Energy Storage for Transmission and Distribution
Application, Electric Power Research Institute (EPRI) 1001834, Palo Alto,
CA, Dec. 2003.
Eynon, R., T. Leckey, and R. Douglas Hale. The Electric Transmission Network:
A Multi-Regional Analysis, U.S. Department of Energy, Washington, DC, 2001.
Fancher, R.B., Dynamic Operating Benefits of Energy Storage, Decision Focus, Inc.:
Oct. 1986. Report EPRI AP-4875.
Fuldner, Arthur H. “Upgrading Transmission Capacity for Wholesale Electric
Power Trade,” U.S. Department of Energy, Washington, DC, 1999.
Gale, R., J. Graves, and J. Clapp. The Future of Electric Transmission in the United
States, PA Consulting Group, Jan. 2001.
Gandy, Simon. “A Guide to the Range and Suitability of Electrical Energy
Systems for Various Applications, and an Assessment of Possible
Policy Effects,” Imperial College of Science, Technology, and Medicine
(University of London), Sept. 2000.
278
Gordon, S.P., and P. K. Falcone. The Emerging Roles of Energy Storage in a

Competitive Power Market: Summary of a DOE Workshop, Albuquerque,
NM: Sandia National Laboratories, June 1995, Report SAND95-8247.
“Grid 2030—A National Vision for Electricity’s Second 100 Years,”
U.S. Department of Energy, Washington, DC, July 2003.
Grimsrud, P., S. Lefton, and P. Besuner. “Energy Storage Systems (ESS) Provide
Significant Added Value by reducing the Cycling Costs of Conventional
Generation,” Aptech Engineering, Sunnyvale, CA, June 2004.
Guey-Lee, Louise. “Wind Energy Developments: Incentives in Selected
Countries,” Renewable Energy Annual 1998, U.S. Department of Energy,
Washington, DC, Sunnyvale, CA, 2000.
Gyuk, Imre. “Electrical Energy Storage,” presentation at the Energy Storage
Association Meeting, Aug. 2000.
Handbook of Batteries, 3rd Edition, edited by D. Linden and T. Reddy,
McGraw-Hill, NY, 2002.
Hassenzahl, W.V. Energy Storage in a Restructured Electric Utility Industry—
Report on EPRI Think Tanks I and II: Electric Power Research Institute,
Palo Alto, CA, Sept. 1997. Report EPRI TR-108894.
Hennessey, Timothy D.J. “The Multiple Benefits of Integrating the VRB-ESS
with Wind Energy Producers—A Case Study in MWH Applications,”
presentation at the AWEA Conference 2004.
Hirst, Eric. Bulk Power Ancillary Services for Industrial Customers, Electricity
Consumers Resource Council, June 1999.
—. Expanding Transmission Capacity, July 2000.
—. Interaction of Wind Farms with Bulk-Power Operations and Markets,
Project for Sustainable FERC Energy Policy, Sept. 2001.
Hirst, E., and B. Kirby. “Ancillary Services: The Forgotten Issue,” printed in Electric
Perspectives, EEI, Washington, DC, July–Aug. 1998.
—. Electric Power Ancillary Services, Oak Ridge National Laboratory,
Oak Ridge, TN, Feb.1996, Report ORNL/CON-426.
—. Bulk Power Basics, Mar. 12, 2000.
279
—. “Unbundling Generation and Transmission Services for Competitive

Electricity Markets,” Oak Ridge National Laboratory, Oak Ridge National
Laboratory, Oak Ridge, TN, 1998, Report ORNL/CON-454.
—. “The Functions, Metrics, Costs, and Prices for Three Ancillary Services,”
Edison Electric Institute, Washington, DC, Oct. 1998.
—. “Measuring Generator Performance, in Providing the Regulation and
Load-Following Ancillary Services,” Dec. 2000.
Hopper, John M. “Can Natural Gas Storage Capacity Affect the Price of Electric
Power? You’d Better Believe It!” Power & Gas Marketing, Oildom Publishing
Company of Texas: Houston, TX, Jan./Feb. 2002.
Hrehor, R. and D. Sytsma. “Gas-Power Infrastructure: The Missing Link?”
Public Utilities Fortnightly, Vienna, VA, Feb.15, 2001: 32–37.
Iannucci, J., B. Erdman, and J. Eyer. “Technical and Market Aspects of Innovative
Storage Opportunities,” presentation at EESAT 2002, San Francisco, CA,
Apr. 2002.
Iannucci, J., and S. Schoenung. Energy Storage Concepts for a Restructured
Electric Utility Industry, July 2000, Report SAND2000-1550.
Isherwood, W., R. Smith, S. Aceves, G. Berry, W. Clark, R. Johnson, D. Das,
D. Goering, and R. Seifert. Remote Power Systems with Advanced Storage
Technologies for Remote Alaskan Villages, National Renewable Energy
Laboratory, Golden, CO, Dec. 1997, Report UCRL-ID-129289.
Jess, Margaret. “Restructuring Energy Industries: Lessons from Natural Gas,”
Natural Gas Monthly, U.S. Department of Energy, Washington, DC,
May 1997, vii–xxi.
Kamibayashi, Makoto. “High Charge and Discharge Cycle Durability of the
Sodium Sulfur Battery,” presentation at EESAT 2002, San Francisco, CA,
Apr. 2002.
Kroon, H., and G. H. C. M. Thijssen. “Determination of Commercial Viability of
Flow Batteries,” presentation at EESAT 2002, San Francisco, CA, Apr. 2002.
Lazarewicz, Matt. “A Description of the Beacon Power High Energy Power
Composite Flywheel Energy Storage Systems,” presentation at EESAT 2002,
San Francisco, CA, Apr. 2002.
280
Lehr, Ronald C. The National Wind Coordinating Committee Status Report and
Regulatory Recommendations, National Association of Regulatory Utility
Commissioners, Aug. 2001.
Lotspeich, Chris. “A Comparative Assessment of Flow Battery Technologies,”
presentation at EESAT 2002 Conference, San Francisco, CA, Apr. 2002.
Maass, Peter. “Operation Experience with Huntorf, 290 MW World’s First
Air Storage System Energy Transfer (Asset) Plant,” presentation at the
American Power Conference, Apr. 21–23, 1980.
MacCracken, Mark. “Thermal Energy Storage Myths,” ASHRAE Journal 45, no 9,
Atlanta, GA, Sept., 2003: 36.
McDowall, Jim. “Battery Life Considerations in the Energy Storage Applications
and their Effect on Life Cycle Costing,” presentation at the IEEE Power
Engineering Society, Summer Meeting, Vancouver, BC, 2001.
Makansi, Jason. “Energy Storage for Grid Management and RTO Management,”
presentation at Transmission: Operating and Planning Conference,
Washington, DC, 2002.
Mears, Dan. Introduction to NGK’s Sodium Sulfur Batteries, Technology Insights,
July 2004.
Miller, R., and J. Malinowski. Power System Operation, 3rd edition, McGraw-Hill,
New York, 1994.
Mitlitsky, F., B. Meyers, and A. H. Weisberg. Regenerative Fuel Cell Systems R & D,
June 1998, Report UCRL-JC-131087.
Miyaki, Shinichi. “Vanadium Redox-Flow Batteries for a Variety of a
Applications,” presentation at the IEEE Power Engineering Society,
Summer Meeting, Vancouver, BC, 2001.
Nakhamkin, M., and R. Wolk. “Compressed Air Inflates Gas Turbine Output,”
Power Engineering, PennWell Publishing, Tulsa, OK, Mar. 1999.
Nakhamkin, M., and R. Pot. Hybrid Plant Technology for Distributed
Power Generation, presented at PowerGen International conference,
New Orleans, LA, Dec. 1999.
National Transmission Grid Study, U.S. Department of Energy, Washington, DC,
May 2002.
Natural Gas Annual 2002, U.S. Department of Energy, Washington, DC, Jan. 2004.
281
Natural Gas Infrastructure & Storage, Office of Fossil Energy, U.S. Department of
Energy, Web site: http://www.fe.doe.gov/oil_gas/gasstorage/.
Natural Gas Storage—End User Interaction Task 2, Topical Report, Office of
Fossil Energy, U.S. Department of Energy, Jan. 1996.
Nichols, David. “Electricity: Challenges and Solutions,” presentation at the IEEE
Power Engineering Society Winter Meeting, Columbus, OH, Jan. 29, 2001.
Nichols, D., B. Tamyurek, and H. Vollkommer. “Sodium Sulfur Battery (NAS)
Applications,” presentation at the IEEE Power Engineering Society
Meeting, 2003.
Nourai, Ali. “Bulk Energy Storage Applications,” presentation at the Electricity
Supply Association Mini-Meeting, Washington, DC, Nov. 14, 2001.
Overview of Wind Technologies, U.S. Department of Energy, Washington, DC, 1997.
Pansini, A., and K. Smalling. Basics of Electric Power Transmission, PennWell
Publishing Company, Tulsa, OK, 1998.
Parsons, B., M. Milligan, B. Zavadil, D. Brooks, B. Kirby, K. Dragoon, and
J. Caldwell. Grid Impacts of Wind Power: A Summary of Recent Studies in
the United States, National Renewable Energy Laboratory, Golden, CO,
June 2003, Report NREL/CP-500-34318.
Patton, A.D., C. Singh, and D. Robinson. The Impact of Restructuring Policy
Changes on Power Grid Reliability, Albuquerque, NM: Sandia National
Laboratories, 1998. Report SAND98-2178.
Petersik, Thomas W. “Modeling the Cost of U.S. Wind Supply, Issues in
Midterm Analysis for Modeling the Costs of U.S. Wind Supply,”
Washington, DC, 1999, Report EIA/DOE-0607(99).
Platt, C., and J. Hurwitch. “Energy Storage: It’s not Just Load Leveling Anymore,”
Public Utilities Fortnightly, Vienna, VA, 136: 15, Aug. 1998, 52–56.
Price, A., G. Thijssen, and P. Symons. “Electricity Storage, A Solution in Network
Operation?” presentation at DistribuTech Europe, Oct. 12, 2000.
Price, Anthony. “The Vulnerability of the Electricity Supply System,”
presentation at the Electricity Supply Association Mini-Meeting,
Washington, DC, Nov. 14, 2001.
—. “Flow Batteries,” presentation at the IEEE Power Engineering Society
282
Randazzo, Sam. “Grid Business: The Midwest Region,” presentation at the Grid
Business Conference, Mar. 2002.
Reliability Assessment 2002–2011, North American Electric Reliability Council,
Princeton, NJ, Oct. 2002.
“Renewable Electricity Purchases: History and Recent Developments,”
Renewable Energy Annual 1998 Issues and Trends, U.S. Department of
Energy, Washington, DC, 1998.
Renewable Energy Annual—2002, U.S. Department of Energy, Washington, DC,
Nov. 2003.
Roberts, Brad. “Evolving Energy Storage Technologies for Protecting Large
Scale Critical Loads,” presentation at the IEEE Power Engineering Society
—. “High Power, Short Duration Energy Storage Systems,” presentation at
the IEEE Power Engineering Society Summer Meeting, Vancouver, BC, 2001.
Scaini, V., P. Lex, T. Rhea, and N. Clark. “Battery Energy Storage for Grid Support
Applications,” presentation at EESAT 2002, San Francisco, CA, Apr. 2002.
Schimmoller, Brian, “Compressed Air Energy Storage Makes Use of Space in
Underground Caverns,” Power Engineering, PennWell Publishing, Tulsa, OK,
Aug. 1, 2001.
Schoenung, Susan. Characteristics and Technologies for Long- vs. Short-term
Energy Storage, Albuquerque, NM: Sandia National Laboratories, Mar.
2001, Report SAND2001-0765.
Schoenung, S., and W. Hassenzahl. Long- vs. Short-Term Energy Storage
Technologies Analysis: A Life Cycle Cost Study, Albuquerque, NM: Sandia
National Laboratories, 2003, Report SAND2003-2783.
Sedano, Richard P. Dimension of Reliability: A Paper on Electric System Reliability
for Elected Officials, National Council on Competition and the Electric
Industry, Oct. 2001.
Shepard, S., and S. Van der Linden. “Compressed Air Energy Storage Adapts
Proven Technology to Address Market Opportunities,” Power Engineering,
Apr. 2002, p. 34–37.
Shirazi, M., and S. Drouilhet. “An Analysis of the Performance Benefits of Short-
Term Energy Storage in Wind-Diesel Hybrid Power Systems,” presentation
at the 1997 ASME Wind Energy Symposium, 1997, NREL/CP-440-22108.
283
Silvetti, B., and M. MacCracken. “Thermal Storage and Deregulation,” ASHRAE

Journal, Atlanta, GA, Apr. 1998.
Speaks, Kelly. “The Emerging Need for Intermediate Peaking Power,” Power and
Gas Marketing, Oildom Publishing Company of Texas, Houston Texas,
Mar./Apr. 2002, 32–33.
Stack, James. “Development of a Performance Data Reduction, Analysis, and
Reporting System for a Wind-Diesel Hybrid Power System,” Office of
Science, Department of Energy ERULF Program, Bucknell University/
National Renewable Energy Laboratory, Golden, CO, Aug. 2000.
“Study of Electric Transmission in Conjunction with Energy Storage
Technology,” Lower Colorado River Authority, Austin, TX, Aug. 2003.
Swaminathan, Shiva. Review of Power Quality Applications of Energy Storage
Systems, Albuquerque, NM: Sandia National Laboratories, July 1998,
Report SAND98-1513.
Swaminathan, S., N. Miller, and R. Sen. Battery Energy Storage Systems Life
Cycle Costs Case Studies, Albuquerque, NM: Sandia National Laboratories,
Aug. 1998, Report SAND98-1905.
Swaminathan, S., W. Flynn, and R. Sen. Modeling of Battery Energy Storage in
the National Energy Modeling System, Albuquerque, NM: Sandia National
Laboratories, Dec. 1997, Report SAND97-2926.
Symons, Philip C. “Opportunities for Energy Storage in Stressed Electricity
Supply Systems,” presentation at the IEEE Power Engineering Society,
Summer Meeting, Vancouver, BC, 2001.
Taylor, P., L. Johnson, K. Reichart, P. DiPietro, J. Philip, and P. Butler. “A Summary
of the State of the Art of Superconducting Magnetic Energy Storage
Systems, Flywheel Energy Storage Systems, and Compressed Air
Energy Storage System,” July 1999, Albuquerque, NM: Sandia National
Laboratories, Report SAND99-1854.
Taylor, R., D. Bradshaw, and J. Hoagland. “Energy Storage for Ancillary Services,”
presentation at EESAT 2002, San Francisco, CA, Apr. 2002.
Taylor, R., and J. Hoagland. “Using Energy Storage with Energy for Arbitrage,”
presentation at EESAT 2002, San Francisco, CA, Apr. 2002.
Thermal Energy Storage—Economics and Benefits, E3 Energy Services, LLC.,
Arlington, VA, 2002.
284
Thijssen, Gerard. “Supply Quality, Reliability & Storage,” presentation at the

Commercial Viable Electric Storage, London, UK, Jan. 30, 2001.
Torre, W., and V. Eckroad. “Improving Power Delivery Through the Application of
Superconducting Magnetic Energy Storage,” presentation at the IEEE Power
Engineering Society Winter Meeting, Columbus, OH, Jan. 29, 2001.
“Transmission Pricing Issues for Electric Generation from Renewable Resources,”
Renewable Energy Annual 1998 Issues and Trends, U.S. Department of
Energy, Washington, DC, 1998.
Transmission Reliability Multi-Year Program Plan FY 2001–2005, Office of Power
Technologies, Energy Efficiency and Renewable Energy, U.S. Department of
Energy, Washington, DC, May 2001.
The Value of Underground Storage in Today’s Natural Gas Industry,
U.S. Department of Energy, Washington, DC, Mar. 1995.
Van der Linden, Septimus. “CAES for Today’s Market,” presentation at EESAT
2002, San Francisco, CA, Apr. 2002.
—. “Compressed Air Energy Storage Adapts Proven Technology to
Address Market Opportunities,” Power Engineering, PennWell Publishing,
Tulsa, OK, Apr. 2001.
Williams, Brad. “A Case Study: The Use of VRB Energy Storage System
(VRB-ESS) as a Utility Network Planning Alternative," 14th Annual
Electricity Storage Association Conference, 2004.
Wilson, Roger. “Market Driven CAES,” presentation at Electric Power 2002,
Mar. 2002.
Wind, Thomas. “Compressed Air Energy Storage in Iowa,” presentation for the
Iowa Association of Municipal Utilities, Wind Energy Consulting, 2002.
Wolsky, A.M. “Introduction to Progress and Promise of Superconductivity for
Energy Storage in the Electric Power Sector,” May 1998, Report ANL/ES/
CP–96555.
Workshop on Electric Transmission Reliability, Workshop Report, U.S.
Department of Energy Transmission Reliability Program and the
Consortium for Electric Reliability Technology Solutions, Sept. 1999.
World UPS Markets: Alternative Energy Storage Solutions, San Jose, CA,
Frost & Sullivan, 2003.
285
World Lead Acid Battery Markets, San Jose, CA, Frost & Sullivan, 2002.
Zink, J.C. “Who Says You Can’t Store Electricity?” Power Engineering, PennWell
Publishing, Tulsa, OK, Mar. 1997, 21–25.
Zucchet, Michael J. “Renewable Resource Electricity in the Changing Regulatory
Environment,” Renewable Energy Annual 1995, U.S. Department of Energy,
Washington, DC, 1996.
List of Figures
1. Storage and the Electric Power Industry

Figure 1–1. Coal units lag nuclear in utilization improvements .............. 7
Figure 1–2. Power generation cycling cost breakdown .............................. 9
Figure 1–3. Transmission load relief incidents on the rise ......................16
2. Storage in Other Energy Markets

Figure 2–1. High deliverability enhances the value of storage ..............49
3. Electricity Storage Technologies

Figure 3–1. Pumped-hydro storage................................................................61
Figure 3–2. Compressed air energy storage (CAES) .................................69
Figure 3–3. Thermal and compressed air storage .....................................79
Figure 3–4. Vanadium redox flow battery.....................................................81
Figure 3–5. Zinc bromine flow battery ..........................................................81
Figure 3–6. Polysulfide bromide flow battery ..............................................82
Figure 3–7. Cerium zinc flow battery ............................................................82
Figure 3–8. Sodium sulfur battery (NAS) .................................................. 100
Figure 3–9. Lead-acid battery (LA) ............................................................. 108
Figure 3–10. Nickel cadmium battery (NiCd) ............................................. 117
Figure 3–11. Flywheel ...................................................................................... 125
Figure 3–12. Electrochemical capacitor ...................................................... 135
Figure 3–13. Superconducting magnetic energy storage (SMES) ........ 142
Figure 3–14. Thermal energy storage (TES) ............................................... 151
xii
4. Applications
Figure 4–1. Energy storage applications ................................................... 166
Figure 4–2. Application capability sizing requirements ........................ 167
Figure 4–3. Energy storage market roles .................................................. 167
Figure 4–4. Cycling issues ............................................................................ 171
Figure 4–5. Ownership costs ........................................................................ 173
Figure 4–6. Commodity arbitrage ............................................................... 177
Figure 4–7. Contingency reserves............................................................... 182
Figure 4–8. Blackstart..................................................................................... 185
Figure 4–9. Frequency regulation ............................................................... 188
Figure 4–10. Voltage regulation..................................................................... 194
Figure 4–11. Transmission and distribution (T&D) asset deferral ......... 197
Figure 4–12. Power quality.............................................................................. 203
Figure 4–13. Peak shaving .............................................................................. 206
Figure 4–14. Regenerative energy ................................................................ 209
5. Renewable Energy and Storage

Figure 5–1. Renewable energy shortfall .................................................... 214
Figure 5–2. Remote power ............................................................................ 222
Figure 5–3. Island grid ................................................................................... 226
Figure 5–4. Dispatchable wind .................................................................... 235
Figure 5–5. Wind capacity firming .............................................................. 237
Figure 5–6. Baseload wind ........................................................................... 241
6. Our New Energy Future

Figure 6–1. U.S. large-scale energy storage penetration declines ..... 255
Figure 6–2. Energy storage reduces peak prices .................................... 259
Appendix
Vendor Web Sites

ABB www.abb.com
Active Power www.activepower.com
AFS Trinity www.afstrinity.com
Alcad Limited www.alcad.com
Alstom Power www.power.alstom.com
American Hydro www.ahydro.com
American Superconductor www.amsuper.com
Baltimore AirCoil Company www.baltimoreaircoil.com
Beacon Power www.beaconpower.com
C&D Technologies, Inc. www.cdtechno.com
CAES Development Company www.caes.net
Calmac Manufacturing Corp. www.calmac.com
Crown Battery Manufacturing Company www.crownbattery.com
Cryogel www.cryogel.com
Dresser-Rand www.dr.com
Dunham-Bush, Inc. www.dunham-bush.com
East Penn Manufacturing Company www.dekabatteries.com
Evapco www.evapco.com
Exide www.exide.com
272
FAFCO, Inc. www.fafco.com

Gill Batteries www.gillbatteries.com
GNB Industrial Power www.gnb.com
Hitec Power Protection www.hitecups.com
Hoppecke Batterien GmbH www.hoppecke.com
MWH www.mwhglobal.com
NGK Insulators, Ltd. www.ngk.co.jp
Paul Mueller Company www.muel.com
Pentadyne Power Corporation www.pentadyne.com
RWE Piller GmbH www.piller.com
Plurion Systems, Inc. www.plurionsystems.com
Premium Power Corporation www.premiumpower.com
Saft www.saft.com
SatCon Technologies Corporation www.satcon.com
Sumitomo Electric Industries www.sei.co.jp
Trojan Battery Company www.trojanbattery.com
Tudor Batteries www.tudorbatteries.com
Urenco Power Technologies www.uptenergy.com
US Flywheel Systems www.us-flywheel.com
VRB Power Systems www.vrbpower.com
Vycon www.calnetix.com
ZBB Energy www.zbbenergy.com
INDEX
Index Terms Links
Active Power 79 130 134

web site 271
Air conditioning power usage 22 23
energy management with 208
thermal energy storage with 151 155
Alabama Electric Cooperative 75 178
Alternating current–direct current
(AC–DC) conversion 57
Aluminum industry 22
American Electric Power (AEP) 101 102 255
American Hydro 67 271
American Superconductor 143 147 148 150
151 195
web site 271
American Wind Energy Association
(AWEA) 10 264
Ancillary services
electric power industry 17
spending on 18
transmission charges unbundled from 18
Applications of energy storage 165
asset deferral with 187 195
blackstart capability for 176 183
This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links
Applications of energy storage (Cont.)

bridging power with 168
capability sizing requirement for 167
commodity arbitrage with 176
contingency reserves with 176 180
cost-minimizing strategies with 200
design issues for 168
distribution with 185
energy management with 168 209
FACTS with 186
flywheel systems for 209
frequency regulation with 186
load leveling with 168
peak shaving with 206
power quality with 168 204
regenerative energy with 209
retail power 199
roles of 167
summary of 165
transmission with 185
voltage regulation with 187 192
wholesale power 174
APTECH Engineering Services 8 10
Aquifers, natural gas market
storage in 42
Arizona Public Service 205
Asset deferral
transmission/distribution 187 195
vanadium-redox flow battery for 198

Index Terms Links
Australia’s Renewable Energy

Commercialization Program 232
Balance of plant (BOP) 59

CAES 70
Base gas 43
Baseload wind 240
CAES for 241
defined 240
Iowa Stored Energy Project with 242
normal operation of 241
transmission for 242
Batteries
blackstart capability with 184
cerium zinc 80 82 84 87
96
cycle life for 103 112 120
depth of discharge for 112 120
flow 80
lead-acid 108
nickel-cadmium 117
polysulfide bromine 80 82 84 87
90 91 95
sodium sulfur 99
valve regulated lead-acid 108 109
vanadium redox 80 86 89 92
231

Index Terms Links
Batteries (Cont.)
zinc bromine 80 83 86 89
91 94 99 195
Beacon Power Smart Energy Matrix
(SEM) 191
web site 271
Blackout
2003 Canada 2 131
blackstart capability 176 183
Blackstart capability 176 183
CAES for 184
defined 176
pumped-hydro facility for 184
sodium bromide flow battery for 184
BP Solar Malaysia 223
Broken Hill, New South Wales 95
CAES Development Company 78 271

California Energy Commission (CEC) 154 251 263
Canada blackout of 2003 2 131
Capacity firming 236
McCamey, TX, example of 238
storage requirements determined for 237
Capacity-firming capability 11
Cavallo, Alfred 240
Central Research Institute of Electric
Power Industry (CRIEPI) 254

Index Terms Links
Cerium zinc batteries 80 82

cost with 91 92
design of 87
developers of 99
efficiency of 90
historical origins of 84
installations with 96
operation of 87 90
Clean power 203
See also Power quality
CleanSource XR 79
Coal market
CAES benefiting 71
dead-storage for 34
economics of 34
electric power industry v. 26
facility types with 33
long-term supply contracts for 32
market uses of storage for 34
operations for 33
railroad delivery of 33
regulatory impact on storage for 33
rules governing storage for 34
storage needs of 32
technological change impacting 49
transportation impact on storage for 33
wholesale 32
Coal-fired units, facility utilization
with 6
Cold starts 8

Index Terms Links
Colello, Gary 97
Colorado River Authority 239
Commercial and industrial firms (C&I),
natural gas storage by 41
Commodity arbitrage 176
CAES for 178
defined 176
endurance capacity with 177
flow batteries for 178
profitability with 177
pumped-hydro storage for 178
Compressed air energy storage (CAES), 8 57 68
air compressors for 70 72 73
balance of plant for 70
challenges of 76
coal power with 71
contingency reserves with 182
controls for 70 73
cost issues with 74
design of 70
efficiency of 70 74
examples of 75 77
expander turbine train for 70 71
flexibility of 72 73
former Soviet Union facility with 75

Index Terms Links
Compressed air energy storage (CAES), (Cont.)

Huntorf, Germany, facility with 75 180 184
installations 74
Iowa Stored Energy Project as 77
Israel facility for 75
Japan facility with 75
load-following support with 69
Luxembourg facility with 75
major developers of 77
McIntosh, AL, facility with 75 178 181
Norton Energy Storage as 78 182
nuclear power with 71
operations with 70
peaking-power support with 69
power train motor/generator for 70
prospects for 76
recuperator for 70
related technologies to 78
Sesta, Italy, facility with 75
summary of 68
Texas facility with 75
utilities use of 257
variable inlet guide vanes for 73
wholesale power market with 175
wind power with 71 73
Congestion, transmission 3 15 238
Contingency reserves 17 176 180
CAES for 182

Index Terms Links
Contingency reserves (Cont.)

defined 176
pumped-hydro facility for 181
spinning v. nonspinning 180
technical requirements for 181
Cooling load
retail electric power 22 23
Cooling systems, thermal energy
storage with 151 153 155
Council for Scientific and Industrial
Research (CSIR) 228
Cycle life
energy storage design 170 173
flywheel 170
lead-acid battery 112
nickel-cadmium battery 120
sodium sulfur batteries 103
superconducting magnetic energy
storage 146
Cycling damage 8
operation and maintenance (O&M)
costs 9 188
Deliverability
electric power storage 48
natural gas 43
Deluxe Films 131

Index Terms Links
Demand charges 154

retail electric power 22 199
Demand side management (DSM) 152
Department of Energy (DOE)
oil holding costs from 31
renewable energy sources from 214
Department of Minerals and Energy
(DME) 228
Depleted reservoirs, natural gas
market storage in 42
Depth of discharge (DOD) 103
energy storage design 171
lead-acid battery 112
nickel-cadmium battery 120
Deregulation, natural gas market 25 46
Detroit Edison 94 195 255
Didcot Power Station 84
Diesel-reciprocating engine
Hluleka Nature Reserve example of 227
island grid with 224
King Island, Australia, example of 231
Metlakatla, AK, example of 228
operating efficiency for 224
wind power combined with 225
Dinorwig, Wales, UK PHS 65 181

Index Terms Links
Dispatchable wind 234

Alta Mesa, CA, example of 236
contingency reserves from 235
pumped-hydro storage of 236
restoration services from 235
Distribution 12
asset deferral for 187 195
energy storage applications for 185
infrastructure underinvestment with 12
just-in-time delivery with 13
voltage regulation for 187 192
Dogo Island, Japan 230
Donalek, Peter 66
Duke Power 201
Eastern Cape Provincial Government 228

Edison Electric Institute (EEI) 13 254
Efficiency 5
Electric power industry 1
ancillary services with 17
challenges of 1
change in 2
coal-fired units for 6
coal market v. 26
commercial/industrial spending on 19 22
deliverability impacting storage for 48
distribution for 12

Index Terms Links
Electric power industry (Cont.)

efficiency for 5
facility utilization for 6
lessons from other markets for 26 45
location impacting storage for 48
low utilization with 3
market structure influencing, 1
natural gas market v. 26 51
nuclear v. coal for 6
off-peak production cost with 7
operating costs for 7
paradigm shift for 4 51
petroleum market v. 26
policy goals for 5
power outage in 5
production/demand linkage with 3 4
ramping stress with 4
regulatory change impacting storage
for 51
reliability for 5
renewable energy for 10
retail 19
retail energy cost in 21
retail power quality in 20
scale impacting storage for 48
security for 6
shock absorber for 3
speed of market impacting storage
for 47

Index Terms Links
Electric power industry (Cont.)

supply/demand balance with 4
technological change impacting
storage for 49
transmission congestion with 3 15 238
transmission for 12
value of storage for 47
wholesale power market of 6
Electric Power Research Institute
(EPRI) 20 253
Handbook of Energy Storage 74 90 103 113
121 129 139 147
power quality study of 202
Electricity Storage Association (ESA) 264
Electricity storage technologies 55
CAES 68
cerium zinc batteries as 80 82 84 87
90 96 99
defined 55
electrochemical capacitors as 135
energy/procurement/construction
costs for 60
energy reservoir of 55
flow batteries as 80
flywheels as 125
lead-acid battery as 108
medium of 56

Index Terms Links
Electricity storage technologies (Cont.)

nickel-cadmium battery as 117
polysulfide bromine batteries as 80 82 84 87
90 95
power conversion system of 57
pumped-hydroelectric storage 60
sodium sulfur batteries as 99
storage as 142
system components of 56
thermal energy storage as 151
valve regulated lead-acid battery as 108 109
vanadium redox batteries as 80 86 89 90
92 98
zinc bromine batteries as 80 83 86 89
91 94 99
Electrochemical capacitors 135
challenges for 141
cost issues with 139
design of 137
efficiency of 138
example of 140
operations of 137
prospects for 141
Siemens’ SITRAS SES System with 140
summary of 135

Index Terms Links
Electrotechnical Laboratory (ETL) 83

Energy management
cooling loads in 208
demand charges with 206
peak shaving in 206
price volatility prevented with 205
regenerative energy in 209
repetitive motion processes with 209
retail electric power 205
TES for 207
Energy rating, power rating v. 169
Energy Research and Development
Administration (ERDA) 249
Energy Research Corporation 83
Energy reservoir 56
Energy service companies (ESCO)
storage technology goals of 258
Energy storage applications 165
asset deferral with 187 195
bridging power with 168
capability sizing requirement for 167
contingency reserves with 176 180
cost-minimizing strategies with 200
design issues for 168
distribution with 185
energy management with 168 209

Index Terms Links
Energy storage applications (Cont.)

FACTS with 186
future development of 247
peak shaving with 206
power quality with 168 204
retail power 199
roles of 167
summary of 165
transmission with 185
voltage regulation with 187 192
wholesale power 174
Energy storage design 168
cycle life in 170 173
cycling issues in 170
DOD in 171
energy v. power in 169
float losses in 174
ownership costs in 172
round-trip storage efficiency in 172 173
Energy Storage Systems (ESS) Program 249
Energy, procurement and construction
(EPC) 60
Entergy 255
Eskom 123 255

Index Terms Links
Expander turbine train, CAES 70 71
Facility utilization 6
nuclear v. coal 6
off-peak production cost with 7
operating costs with 7
peaking capacity with 8
Federal Energy Regulatory Commission
(FERC) 18 263
natural gas with 36
Order 436 39 40
Order 500 39 40
Order 636 40 41 51
Order 888 52
Order 889 52
Order 2000 52
Flexible AC transmission systems
(FACTS) 15 17
energy storage applications with 186
SMES v. 142 147
voltage stability with 17
Float losses 174
Flow batteries 80
asset deferral with 198
Broken Hill, New South Wales, using 95
challenges of 97

Index Terms Links
Flow batteries (Cont.)

cost issues with 90
design of 85
Detroit Edison using 94 195
developers of 98
four types of 80
Little Barford Power Station, using 95 184
Melbourne, Victoria, Australia, using 95
operation of 85
PacifiCorp using 93 198
Polysulfide bromine batteries examples
of 95
prospects for 97
summary of 80
Tennessee Valley Authority using 96
Totorri Sanyo Electric Company using 92
University of Stellenbosch using 93
vanadium redox batteries examples 92
voltage regulation with 195
zinc bromine batteries examples of 94
Flywheels 23 125
applications of 128
challenges for 134
components of 127
cycle life of 170

Index Terms Links
Flywheels (Cont.)
Deluxe Films, Canada with 131
design of 127
Dogo Island, Japan, example of 230
efficiency of 129
examples of 130
Lyon France Metro with 132
New York City Transit with 210
operations of 127
power quality with 204
prospects for 133
San Francisco MUNI with 210
STMicroelectronics, France, with 130 205
summary of 125
Trackside Energy Management System
with 210
Urenco Power Technologies with 132 210
Usibelli Coal Mine with 132 211
Ford Motor Company 100
Frequency regulation 17
CAES facilities providing 189
flywheels providing 191
lead-acid battery storage providing 190
load following with 189

Index Terms Links
Frequency regulation (Cont.)

NiCd battery storage providing 190
operation/maintenance costs with 188
prompt-response capability for 191
pumped-hydro facilities providing 189
transmission/distribution with 186
Fuji Electric 230
Future energy storage development 247
beneficiaries of 267
deployment in 260
economics for 267
federal support for 261
industry programs for 253
industry support for 264
international programs for 252
obstacles with 266
paradigm, changes of 268
state energy programs for 250
state support for 263
technology in 260
three core issues of 260
user goals for 256
U.S. Federal government in 248
value in 260
visibility for 266
General Electric (GE) 100

Index Terms Links
Golden Valley Electricity Association

(GVEA) 120 122 190 211
Gould, Inc. 83
Green Building Rating System 264
Grid connected power 232
baseload wind with 240
capacity firming with 236
dispatchable wind with 234
transmission congestion with 238
Grossmont Hospital 156
Handbook of Energy Storage 74

electrochemical capacitors in 139
flow batteries in 90
flywheels in 129
lead acid battery in 113
nickel-cadmium battery in 120
sodium sulfur battery in 103
storage in 147
Hirst, Eric 18
Hluleka Nature Reserve 227
Hot starts 9
Huntorf, Germany, CAES facility 75 180 184
Hydro Tasmania 231

Index Terms Links
Hydroelectric power, Metlakatla, AK,

example of 228
Independent system operator (ISO) 15

Industry programs 253
American Electric Power 101 102 255
CRIEPI 254
Detroit Edison 94 195 255
Edison Electric Institute 13 254
Electric Power Research Institute 20 74 90 103
113 121 129 139
147 202 253
Entergy 255
Eskom 123 255
Tokyo Electric Power Company 100 102 107 108
255
Infrastructure, underinvestment in 12
International Energy Agency (IEA) 29 252
International programs 252
governmental programs 253
International Energy Agency 252
Iowa Association of Municipal Utilities
(IAMU) 242
Iowa Stored Energy Project (ISEP) 77 242
CAES for 242
Island grid
defined 224
diesel-reciprocating engine for 224

Index Terms Links
Island grid (Cont.)

load following services with 227
load profile for 224
peak demand for 224
wind/diesel combined for 225
wind/diesel/storage hybrid system for 226
Kansai Electric Power Company 83

Kinetic Energy Storage System (KESS) 230
King Island, Australia 231
Kirby, Brendan 18
Lawrence Livermore National

Laboratory (LLNL) 225
Lazarewicz, Matt 191
Lead-acid (LA) battery 108
challenges for 116
cycle life of 112

Index Terms Links
Lead-acid (LA) battery (Cont.)

design of 110
DOD of 112
efficiency of 112
electrodes in 111
examples of 114
Malaysia PV example with 223
Metlakatla Power & Light using 115 229
operations of 110
photovoltaic power with 222
prospects for 115
Puerto Rico Electric Power Authority
using 114 190
Southern California Edison using 114
summary of 108
two types of 108 110
Vernon, CA, facility using 114
Leadership in Energy and
Environmental Design (LEED) 264
Lefton, Steven 8
Liquefied natural gas (LNG) 37
Little Barford Power Station 84 95
Load following 9
ancillary services with 19

Index Terms Links
Load following (Cont.)

Load leveling 168
Load pockets 48
Local distribution companies (LDC) 38
natural gas storage by 41
Loss prevention 20
Lyon France Metro 132
Malaysia Ministry of Rural

Development 223
Matsushita Electric Industrial
Company 136
McCamey, TX 238
McIntosh, AL, CAES facility 75 178 181
Melbourne, Victoria, Australia 95
Metlakatla Power & Light using 115 228
island grid example 228
MWH Global 61 66 67 272
National Electricity Regulator (NER) 228

National Renewable Energy Laboratory
(NREL) 213 225
National Wind Technology Center 213
Natural gas market
aquifers as storage for 42

Index Terms Links
Natural gas market (Cont.)

base gas in 43
deliverability impacting storage for 48
deliverability in 43
depleted reservoirs as storage for 42
deregulation in 25 46
downstream storage for 37
economics of 43
end user storage for 40
facility types for 42
geological formations determining 43
growth of 36
hedging/speculation in 37
LNG in 37
market uses of storage for 43
operations of 42
pipeline utilization for 39 44
power generation with 36
regulatory impact on storage for 38
salt caverns as storage for 42 44 48 50
seasonal variation in 37 44
state-level regulation of 41
storage needs for 36
technological change impacting 49
trading in 37
wholesale 36
working gas in 43

Index Terms Links
Natural Gas Policy Act of 1978 38

NEC 136
New York City Transit (NYCT) 210
New York State Energy Research
and Development Authority
(NYSERDA) 251 263
Nickel-cadmium (NiCd) battery 117
applications for 120
challenges for 123
cycle life for 120
design of 119
DOD for 120
Eskom, South Africa with 123
examples of 122
Golden Valley Electricity Association
with 120 122 190
operations of 119
prospects for 123
summary of 117
North American Electric Reliability
Council (NERC) 14
Norton Energy Storage 78 182
Nuclear power
CAES benefiting 71

Index Terms Links
Nuclear power (Cont.)

facility utilization with 6
Off-peak cooling 152. See also

Termal energy storage (TES)
Off-peak production cost, facility
utilization with 7
Office of Electricity and Energy
Assurance 249
Oglethorpe Power 178
Okinawa, Japan, PHS 65
Operating costs, facility utilization
with 7
Operation and maintenance (O&M)
costs 9 188
PacifiCorp 15 93 97 198
Paradigm shift, electric power
industry 4 51 268
Peak shaving, energy management
and 206
Peaking capacity 8
transmission congestion during 15 17
Petroleum market
economics of storage for 30

Index Terms Links
Petroleum market (Cont.)

facility types with 29
history of storage for 28
inventories and production for 30
operations for 29
primary storage for 29
regulation of storage for 29
secondary/tertiary storage for 30
spot price movement in 30
storage costs for 31
storage needs of 27
wholesale 27
Photovoltaic power (PV) 222
lead-acid batteries for 222
Malaysia example of 223
remote power with 222
Pinkerton, Joe 80
Plante, Gaston 109
Plurion Systems 84 96 99 272
Polysulfide bromine batteries 80 82
cost with 91
design of 87
developers of 99
efficiency of 90
examples of 95
Little Barford Power Station using 95
operation of 87 90

Index Terms Links
Polysulfide bromine batteries (Cont.)

prospects for 97
Tennessee Valley Authority using 96
PowerCell Corporation 84
Power conversion system (PCS) 57
AC–DC conversion in 57
commodity arbitrage basis for 58
costs for 58
power quality basis for 58
Power outage 5
Powerplant Award 65
Power quality 20
clean 203
cost of 20
Duke Power estimate of 201
EPRI study on 202
fault-intolerant information technology
with 201 202
flywheels for 204
interruptions in 201
lost work product and 201
retail electric power 200
voltage sags in 201 202
Power rating, energy rating v. 169
Power train motor/generator, CAES 70
Premium Power Corporation 97 99 272
Price curve 26

Index Terms Links
Price volatility 26
Production Tax Credit (PTC) 213
Projass Enecorp 223
Projass Engineering 223
Public utility commissions (PUC) 41
Puerto Rico Electric Power Authority 114 190
Pumped-hydroelectric storage (PHS) 8 57 60
adjustable speed machines for 66
capacity of 64
challenges for 66
contingency reserves with 181
cost issues with 64
cycle time requirements of 63
design of 61 62
Dinorwig, Wales, UK, facility for 65 181
efficiencies of 63 67 68
examples of 65
expansion of 62
fast reaction of 63
installations 64
Oglethorpe Power with 178
Okinawa, Japan, facility for 65
operations of 62

Index Terms Links
Pumped-hydroelectric storage (PHS) (Cont.)

prospects for 66
Rocky mountain, GA, facility for 65 178
summary of 60
Taum Sauk facility for 67
underground cavern reservoir version of 67
wholesale power market with 175
PureWave UPS system 205
Ramping stress, electric power

industry 4
Regenerative energy
Regenesys Technologies 84 95 96 97
98 99
Regulatory change 25 46 51
federal support with 261
international programs for 252
natural gas market 25 38 46 51
state energy programs 154 157 250
state support with 263
storage technology goals of 259
strategic petroleum reserves in 51
U.S. Federal government in 248
wholesale electric power from 51

Index Terms Links
Reliability 5
ancillary services for 17
Remote power
cost for 221
environmental impact for 221
photo voltaic 222
wind power for 222
Renewable energy 10 213
advantages of storage with 225
CAES for 239 241
capacity firming with 236
competitiveness of 233
cost for 221
diesel-reciprocating engine combined
with 225
DOE on 214
environmental impact for 221
flywheel with 230
grid connected 232
grid instability with 220

Index Terms Links
Renewable energy (Cont.)

lead acid batteries for 108 111
Malaysia example of 223
noncoincident peak with 218
nondispatchable quality of 218
percent of energy production 220
photovoltaic 222
resource utilization challenges with 216
shortfall in 214
solar 222
system stability with 219
wind power as 215
Renewable Energy Commercialization
Program 232
Renewable Portfolio Standards (RPS) 213
Retail electric power 19
cooling loads in 22 23 208
cost-minimizing strategies for 200
demand charges in 22 199 206
energy cost in 21

Index Terms Links
Retail electric power (Cont.)

multistage tariffs with 199
peak shaving in 206
power quality in 20 200
TES for 207
UPS with 200 203 205
voltage sags in 201 202
Ridge Energy Storage 69
Rocky mountain, GA PHS 65 178
Round-trip storage efficiency 172 173
S&C Electric Company 205

Salt-dome storage, natural gas in 42 44 48 50
San Francisco MUNI rail system 210
San Jacinto College District 157
Sandia National Laboratories 250
Sappi paper mill 149
SatCon Power Technologies 131 134 272
Security 6
Simmons & Company International 50
SITRAS SES System 140
SITRAS Static Energy Storage (SES)
System 140

Index Terms Links
Sodium sulfur batteries (NAS) 99

advantages of 106
Akiruno, Japan, facility with 104
boat racing facility with 105
challenges of 107
cost issues for 103
cycle life of 103
design of 101
DOD of 103
examples of 104
Gahanna, Ohio, facility with 105
operations of 101
prospects for 106
summary of 99
Solar energy 222
lead-acid batteries for 222
nickel-cadmium battery for 120
Southern California Edison 114 236
Speed-to-value index 49
Standard Oil of Ohio 136
Stankiewicz, Charles 150
State Energy Conservation Office
(SECO) 157 239

Index Terms Links
State energy programs 250

California 154 250
New York 251
Texas SECO 157 239
STMicroelectronics 130 205
Storage medium 56
costs of 56
density of 57
Straight-fixed-variable (SFV) design 41
Strategic Petroleum Reserve (SPR) 29
Sumitomo Electric Industries (SEI) 83 98 272
Superconducting magnetic energy
storage (SMES) 142
applications for 145
challenges for 150
cycle life of 146
design of 144
efficiency of 146
examples of 148
major developer of 151
operations of 144
prospects for 149
Sappi paper mill with 149
Stanger, South Africa, facility with 149
summary of 142

Index Terms Links
Superconducting magnetic energy (Cont.)

Wisconsin Public Service with 146 147 148 195
Supply/demand balance, electric
power industry 4
System components of electricity
storage 56
energy reservoir of 55
energy/procurement/construction
costs for 60
medium of 56
power conversion system of 57
Takayama, Ted 106

Taum Sauk facility PHS 67
TenderLand Power Company 236
Tennessee Valley Authority (TVA) 96
Texas State Energy Conservation Office
(SECO) 157 239
Termal and compressed air storage
(TACAS) 78
CleanSource XR as 79
design of 79
UPS v. 79 80
Termal energy storage (TES) 151
challenges for 159

Index Terms Links
Termal energy storage (TES) (Cont.)

design of 153
efficiency of 154
examples of 156
Grossmont Hospital with 156
operations of 153
prospects for 158
San Jacinto College District with 157
summary of 151
Villa Julie College with 157
Termal storage unit (TSU) 78
Tokyo Electric Power Company (TEPCO) 100 102 107 108
255
Totorri Sanyo Electric Company 92
Trackside Energy Management System 210
Transmission 12
asset deferral for 187 195
congestion 3 15 238
infrastructure underinvestment with 12
just-in-time delivery with 13
load relief incidents with 16

Index Terms Links
Transmission (Cont.)
voltage regulation for 187 192
U.S. Department of Defense 249

U.S. Department of Energy (DOE) 13
National Energy Plan of 248
SO2 cap-and-trade system of 248
storage program 58 248
U.S. Department of Transportation 249
U.S. Federal government 248
Department of Defense 249
Department of Energy 13 58 248
Department of Transportation 249
Energy Storage Systems Program 249
Federal Energy Regulatory Commission 18 36 51 52
263
Office of Electricity and Energy
Assurance 249
Uninterruptible Power Supply (UPS) 4 20
flywheels for 125 128 129
lead-acid battery for 108 109 110
PureWave 205
retail electric power with 200 203 205
TACAS v., 79 80
Unit-trains 33
United Mine Workers’ strike 35
University of Stellenbosch 93
Urenco Power Technologies (UPT) 132 210 230

Index Terms Links
User goals for storage technologies 256

commercial/industrial 256
energy service companies 258
regulators 259
utilities 257
Usibelli Coal Mine 132 211
Utilities, storage technology goals of 257
Valve regulated lead-acid battery

(VRLA) 108 109
applications of 110
Vanadium redox batteries 80
asset deferral with 198
cost with 90 91
design of 86
developers of 98
efficiency of 89
examples of 92
operation of 86 89
PacifiCorp using 93 198
Totorri Sanyo Electric Company using 92
University of Stellenbosch using 93
Variable inlet guide vanes, CAES 73
Villa Julie College 157

Index Terms Links
Voltage control 17
Voltage regulation
reactive power controlling 192
transmission/distribution with 187 192
zinc-bromide flow battery for 195
Voltage surges 20
VRB Power Systems 93 98 99 198
web site 272
Warm starts 8
Wholesale power market 6
commodity arbitrage for 176
compressed air energy storage for 176
contingency reserves for 176 180
cycling damage with 8
facility utilization with 6
infrastructure underinvestment in 12
nuclear v. coal in 6
off-peak production cost in 7
operating costs in 7
operation and maintenance (O&M)
costs for 9
peaking capacity in 8
pumped-hydro facilities for 175
renewable energy with 10
storage facilities importance to 26

Index Terms Links
Wholesale power market (Cont.)

wind power in 10
Williams, Brad 199
Wind power 10 215
advantages of storage with 225
ancillary services for 236
CAES benefiting 71 73 239 241
capacity firming with 11 236
capacity payments for 236
competitiveness of 233
decoupling production/demand for 11
diesel-reciprocating engine combined
with 225
flywheel with 230
grid connected 232
grid instability with 220
market penetration of 217
noncoincident peak with 218

Index Terms Links
Wind power (Cont.)

nondispatchable quality of 218
percent of energy production 220
potential for 215
power forecasts for 11
reserve support requirements for 236
resource utilization challenges with 216
retarded growth of 215
storage technologies with 215
system stability with 219
turbine technology for 10
weather forecasts for 219
wind/diesel/storage hybrid system
with 226
Wisconsin Public Service (WPS) 146 147 148 195
Wisconsin Public Utilities 14
Working gas 43
Wyoming’s Powder River Basin 32
ZBB Energy Corporation 94 99 195 272

Zinc bromine batteries 80
Broken Hill, New South Wales, using 95
cost with 91
design of 86
Detroit Edison using 94 195
developers of 99

Index Terms Links
Zinc bromine batteries (Cont.)

efficiency of 89
examples of 94
Melbourne, Victoria, Australia, using 95
operation of 86 89
Zito , Ralph 84

Energy Storage A Nontechnical Guide - Compress

Uploaded by

Copyright:

Available Formats

Energy Storage A Nontechnical Guide - Compress

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Energy Storage A Nontechnical Guide - Compress

Uploaded by

Copyright:

Available Formats

Energy Storage

Director: Mary McGee

Library of Congress Cataloging-in-Publication Data

Azar, Jamal J., 1937-

Printed in the United States of America

3 4 5 6 7 12 11 10 09 08

Dr. Hermann Scheer

1. Storage and the Electric Power Industry.......... 1

Transmission and Distribution .........................................................12

2. Storage in Other Energy Markets .............................25

Coal Market ................................................................................................31

Natural Gas Market ................................................................................36

Lessons Learned ........................................................................................45

3. Electricity Storage Technologies ..............................55

Pumped-Hydroelectric Storage (PHS) ..........................................60

Compressed Air Energy Storage (CAES) .....................................68

Flow Batteries ............................................................................................80

Sodium Sulfur Battery ...........................................................................99

Lead-Acid Battery ................................................................................. 108

Nickel Cadmium Battery .................................................................. 117

Flywheels ................................................................................................... 125

Electrochemical Capacitors ............................................................. 135

Superconducting Magnetic Energy Storage ........................... 142

Thermal Energy Storage .................................................................... 151

References ................................................................................................. 160

4. Applications ................................................................................... 166

Wholesale Power................................................................................... 175

Transmission and Distribution ...................................................... 186

Retail Market ........................................................................................... 200

References ................................................................................................. 212

5. Renewable Energy and Storage ..................... 214

Remote Power ........................................................................................ 222

References ................................................................................................. 245

6. Our New Energy Future...................................................... 248

User’s Goals for Storage Technologies ....................................... 257

A Road Map for the Future ............................................................. 261

References ................................................................................................. 271

Appendix—Vendor Web Sites ............................................ 272

The electric power industry has some immense challenges before

Although the industry may prefer slow, evolutionary change, it

For these and other reasons, energy storage technologies provide

• Improving low utilization of power facilities

By absorbing or providing power—even a small amount—at

commodity value. This is a simplistic, one-dimensional view of storage

Although treated as a new technology, the electric power industry

Therefore, energy storage technologies are not simply a solution in

1. Efﬁciency. Storage technologies can be used to improve the

3. Security. Storage technologies can provide a means to blunt

Large-scale energy storage facilities can help increase the utilization of

Even combined-cycle units—having improved signiﬁcantly from their

Fig. 1–2. Power generation cycling cost breakdown (Courtesy of Aptech

Large-scale storage facilities can help these mid-merit power facilities

Advancements in wind turbine technology have created many

prices. Unpredictable power means that not only is the capacity

and use of supporting components such as the transmission facilities.

As investments in the infrastructure continue to fall behind the

Although many hold out for a regulatory-led movement to

voltage stability equipment or capacity expansion alternatives. At the

(FACTS) devices are being promoted, particularly in susceptible