2012 IEEE Electrical Power and Energy Conference

Energy Storage Technologies for Grid-Connected

and Off-Grid Power System Applications
Faruk A. Bhuiyan, Member, IEEE, and Amirnaser Yazdani, Senior Member, IEEE

Ahstract-This paper presents the updated status of energy storage (CAES) systems can be employed to eliminate the
storage (ES) technologies, and their technical and economical need for peak power plants [11]. For medium-and low-scale
characteristics, so that, the best technology can be selected either
applications, battery energy storage (BES) systems prove to be
for grid-connected or off-grid power system applications. Consid­
ering the wide range of applications, effective ways of storing and
the most effective options due to their high power and energy
retrieving electrical energy remains a challenge. In addition to densities, shorter response times, smaller sizes and relatively
load variations, the ever-increasing penetration of intermittent lower costs [3]. Although BES systems have been used in
renewable energies into the power system has presented the portable electronic devices for many years, their widespread
system with difficulties. Frequency regulation, voltage support,
use in power system applications is relatively new. Thus, this
load leveling, peak shaving, economic dispatch, and production
leveling represent the main power system applications, where
paper reviews the present status of ES technologies and their
ES can play an important role. In off-grid applications, ES can power system applications. Amongst the BES technologies,
be used to balance the generation and consumption, to prevent lead-acid batteries have been employed in power system ap­
frequency and voltage deviations. Due to the widespread use of plications [12], [13] since the 1980. Reference [14] discusses
battery energy storage (BES), the paper further presents various
thermal energy storage (TES), for buildings, whereas reference
battery models, for power system economic analysis, reliability
evaluation, and dynamic studies.
[15] has placed the focus on flywheel energy storage (FES),
lead-acid batteries and nickel-cadmium batteries. Reference
Index Terms-Energy storage (ES), battery energy storage
[16] studies the economic of large-scale ES, e.g. CAES,
(BES), flywheel energy storage (FES), compressed air energy
storage (CAES), pump hydro energy storage (PHES), supercon­ for grids with significant wind power integration. Remark­
ductive magnetic energy storage (SMES), electric double-layer able ES technologies include FES [17], flow batteries [18],
capacitor (EDLC), grid-connected power system, and off-grid [19], superconductive magnetic energy storage (SMES) [20],
power system. sodium sulphur (NaS) battery, zebra battery [21], [22], lithium­
ion battery, metal-air [23] battery, CAES [24], and electric
I. INTRODUCTION
drive vehicle (EDV) [4]. Ignoring power system applications
and emphasizing less the BES technologies, reference [25]
ffective and efficient methods of storing and ret�ieving
E energy, in the context of power systems, are sttll un­
der intensive investigation [1]- [3]. Introduction of energy
highlights the need for ES to strengthen power networks
and for load leveling. Ignoring the PHES, CAES, and TES,
reference [26] discusses the characteristics of electrochemical
storage (ES) systems can be beneficial by allowing time­ BES and electrical ES technologies. Reference [9] focuses
shifting between electricity generation and consumption, in on BES technologies, including EDV, and discusses models
grid-connected and off-grid networks [4], [5]. Due to the inter­ for economic analysis and power system studies. Reference
mittent nature of renewable energy resources, the penetration [15] has compared various ES technologies. Reference [27]
of renewable energies renders the power system vulnerable describes ES technologies for wind power plants, and [8]
to instability, deteriorated reliability, and compromised power discusses ES technologies for transportation and power system
quality [6]. Appropriate ES can mitigate the impact of in­ applications. This paper attempts to review different ES tech­
termittency, assist the power system ride through peak power nologies, compare their characteristics, indicates their power
demands, and allow the transmission and distribution networks system applications, and discuss the models for selecting suit­
to fulfill their full capacities [7], [8]. The need for ES have able technologies for grid-connected and off-grid applications.
thus been felt more with the recent increasing penetration of The rest of the paper is organized as follows: Section
renewable energies [9]. Large-scale ES, e.g. pumped hydro II introduces the ES technologies. Section III compares the
energy storage (PHES) systems, have been in use for a few technical and economical characteristics and power system
decades; but they suffer from relatively large transitional re­ applications. Section IV introduces BES models for analysis.
sponse times and also need special geographical topographies. Section V discusses the present growth trend of BES systems,
In the meantime, various ES technologies have been invented and Section VI concludes the paper.
and are being tested [4], [10]; they have different character­
istics and costs. For large-scale grid-connected systems, if a II. ENERGY STORAGE (ES) TECHNOLOGI ES
suitable site is available, the PHES and compressed air energy Fig. 1 indicates that ES technologies, [4], [9], [10], [12],
[26], and [27] can be classified based on their mechanical,
The authors are with the Western University. London. ON, Canada,
and Ryerson University. Toronto. Canada (e-mail: tbhuiyan@uwo.ca; yaz­ electrical, electrochemical, thermal, and virtual natures. These
dani@ryerson.ca) are explained below:

978-1-4673-2080-1/12/$31.00 ©2012 IEEE 303

 electrical electrochemical/
potential/ pneumatic) (electromagnetic / static battery energy storage
energy storage field) energy storage lead acid (i) flooded 1) ceramic thermal
type (ii) valve regulated storage
1) pumped hydro 1) superconductive type
energy storage magnetic energy storage 2) thermal fluid
2) nickel-cadmium storage
2) flywheel energy 2) capacitor
3) lithium-ion
storage 3) hot water storage
3) supercapacitor 4) sodium-sulpher
3) compressed air 4) borehole
5) zebra battery community thermal
energy storage
6) metal air
virtual energy 7) nickel metal- hydride
storage 8) tl ow batteri es (i)
1) demand side management vanadium redox (ii) zinc­
bromide (iii) cerium-zink
2) demand response
) hydrogen and fuel cell

Fig. 1. types of energy storage

A. Mechanical (Kinetic, Pneumatic, Potential, or Hy­ B. Electrical (Electromagnetic or Static Field) Energy Storage
dropower) Energy Storage Technologies (EES)
Mechanical energy storage technologies store energy in the
The family of EES systems consist of capacitors, super­
form of potential or kinetic energy.
capacitors and electric double-layer capacitor (EDLC) and
1) Pumped Hydro Energy Storage (PHES): The excess
super-conductive magnetic energy storage (SMES); these are
electrical power during low-load periods is used to pump
described below.
water from a low-lying reservoir to an elevated one, and
energy is stored as potential energy. When the demand for 1) Super-Conductive Magnetic Energy Storage (SMES):
electrical power peaks, water is released to drive a turbine­ A SMES system stores energy in a magnetic field created
generator set and produces electricity. Thus, a PHES requires by a direct current in a superconductor coil. By utilizing
specific geographical topography and is considered the most liquid helium or liquid nitrogen, the conducting coil is cooled
cost-effective ES technology. Due to the inertia of the water cryogenically below its superconductive critical temperature,
column, PHES systems are not very rapid in mode changes. at which the coil exhibits zero resistance. Even despite the
Typically, PHES is used in grid-connected systems, its use for cooling system losses, a SMES system offers a high efficiency.
off-grid systems is limited, mainly due to the need for suitable Moreover, it offers high power densities, and short transitional
topography, insufficient economy of scale, and low transitional times. SMES systems can be utilized to improve power quality
speed. in both grid-connected and off-grid systems.
2) Flywheel Energy Storage (FES): Flywheel is a mechan­ 2) Capacitors: Capacitors store energy in the form of
ical device that stores kinetic energy in a rotating mass. An electric charges distributed on two metallic plates separated
electric motor is employed to flywheel during off-peak periods by thin layers of a dielectric material. The capacitance is
to store electric energy in the form of kinetic energy, and then proportional to the area of the plates and permittivity of
the kinetic energy is released during peak demand periods. the dielectric, and is inversely proportional to the distance
FES requires minimal maintenance, offers a fast transitional between the two plates. Capacitors have a very large life cycle
response, and has a long life cycle. The method is typically and offers short response times, but suffer from low power
suitable for high-power systems, but not for high-energy densities.
systems. FES systems have been used in off-grid applications 3) Super-capacitors/Ultra-capacitors/ Electric Double­
to regulate frequency and to mitigate power fluctuations. Layer Capacitor (EDLC): If the electrode surface area
3) Compressed Air Energy Storage (CAES): During off­ of the conventional capacitor is increased by a special
peak periods, CAES plants are used to store excess electrical technique, then the capacitor becomes supercapacitor/ultra­
energy by compressing air in the form of pneumatic energy capacitorJEDLC or boost-capacitor. These capacitors do not
to a specially designed bunker/cavern. When the power de­ use metal or plastic for anode or cathode, unlike conventional
mand increases, the air is released to drive a conventional capacitors. Rather, they use two layers of metal filaments or
gas turbine-turbine generator set for producing electricity. porous carbons made from foils to form anode and cathode
Presently, CAES is being used for grid-connected applications. that are perfused in dielectric material of electrochemical
However, if an appropriate sites are available, the technology liquid. Energy is stored in the electrochemical dielectric
can also be employed for energy management in off-grid material. They are efficient in short charging and discharging
systems. time with high power rating.

304
C. Electrochemical IBattery Energy Storage (BES) separated by solid beta-alumina electrolyte. As the battery
ElectrochemicallBES systems have been widely used in dif­ needs high temperatures to operate, the construction of the
ferent applications, from portable electronic devices to utility battery requires special arrangement to prevent heat loss. The
systems. For improved characteristics, a variety of materials battery properties exhibit high power and energy densities,
can be utilized in the construction of batteries. good efficiencies, and long life cycles. The battery has been
1) Lead Acid (PbA): A cell of lead acid battery contains used for power quality and renewable energy applications.
a positive electrode of lead dioxide (Pb02) and a negative 6) Metal Air: Metal-air battery is environmentally benign,
electrode of lead (Pb). These are submerged into an aqueous compact, and less costly than other batteries. The anode of
sulphuric acid electrolyte and separated by a micro-porous ma­ the battery is built by metal, e.g. zinc or aluminium, and the
terial. When discharged, both the positive and negative plates cathode is built from porous carbon or metal mesh covered in
become lead-sulphate (PbS04) and the electrolyte converts catalyst. The electrolyte is taken from hydroxide (OH ) ion-

from sulphuric acid (H2S04) to water (H20). The lead acid conductor, e.g. potassium hydroxide (KOH). Due to severe
battery, the most widely used battery can be frequently charged recharging difficulties and low life-cycles, the metal air battery
and discharged. Its relatively short life decreases sharply if the is not being used for power system applications.
battery is discharged deeply. Excessive maintenance, heavy 7) Zebra Battery (NaNiCI2): This battery was invented by
weight due to lead, toxicity, low energy density, and short life Zeolite Battery Research Africa Project (ZEBRA) group of
cycle hinder its use in energy management applications. Due Scientific and Industrial Research (CSIR), South Africa. The
to low costs, they have been used in power system applications negative electrode of the battery is made from liquid sodium.
for several years. The positive electrode converts to nickel and nickel-chloride,
2) Lithium-Ion (Li-ion): The positive electrode of Li-ion respectively, during discharging and charging. The battery
batteries is made from Iithiated metal dioxide, out of different utilizes sodium-aluminum-chloride (NaAICI4) as electrolyte.
lithium compounds, and the negative electrode is made from Self discharge of the battery is negligible. The battery is
graphite. The electrolyte consists of lithium salt dissolved currently used in vehicles and marine equipment.
in organic graphite. A separator is provided for electrical 8) Flow Batteries [28]: Flow batteries store energy in
isolation. Intercalation of lithium into graphite occurs when the liquids, and acts like a rechargeable fuel cell. The electrolytes
battery is charged and the reverse takes place when the battery of the flow batteries pass through cells and convert chemical
is discharged. They show satisfactory operating performances energy into electricity. The batteries need two storage tanks
in terms of high energy densities, high efficiencies and long for electrolytes, and the electrolytes are circulated by pumps
life cycles. Due to the special packaging needs, overcharge to the cathode and anode. The electrodes are separated by a
protection circuitry, and the material costs, Li-ion batteries membrane. The electrodes and the membrane compose a unit,
are comparatively more costly than other types of battery, and called power conversion system/reactor. The rated power and
their utility applications are relatively new; but promising. energy of the batteries depend on, respectively, the reactor
3) Nickel Cadmium (NiCd): The NiCd battery utilizes sizes and storage volumes. Long life cycles, high power and
nickel oxyhydroxide as the cathode, and metallic cadmium or energy capacities, high reliability, reasonable operating and
cadmium hydroxide as the negative electrode; these are kept maintenance costs, flexibilities, full discharge capabilities, and
apart by a separator and are put into an alkaline electrolyte. state of art technologies are the factors behind to select the
Due to its higher energy density than that of PbA, lower batteries for power system applications.
cost relative to few other batteries, and tolerance for deep 9) Hydrogen Storage and Fuel cells: The energy density
discharge, the NiCd battery is preferred for medium-term of compressed hydrogen gas is enormous. Thus, the storing of
energy management for a while. Its main disadvantages are electrical energy in the form of hydrogen has received a great
that cadmium is toxic, and also the battery exhibits negative deal of attention. In power systems, specially in renewable
temperature coefficient that leads to charging difficulties at power systems, the excess electrical energy is used for elec­
higher temperatures. trolysis of water, which produces oxygen and hydrogen gas.
4) Nickel Metal-Hydride (NiMH): Like NiCd, the cathode The hydrogen is compressed and then stored in a container.
of NiMH battery is made from nickel oxyhydroxide, and When electrical power is required, the hydrogen is released to
the anode material is made from an intermetallic compound; run through fuel cell.
these are merged in an alkaline electrolyte usually potassium
hydroxide. The high power density, long life cycle, and very D. Thermal Energy Storages (TES)
low maintenance lead the batteries to use in electric hybrid TES technology can store off-peak electricity as heat. The
vehicle. They also possess better specific energy density than heat is generated utilizing resistive heater. Water, synthetic oil,
that of NiCd batteries, and have replaced NiCd batteries in ceramic, molten salt, and sodium hydroxide [14] [25] can be
many practical applications. Due to high self discharge, their used to store the heat in a reservoir. The more the energy added
usages in power system applications are very limited. to the reservoir, the more the temperature and pressure rises.
5) Sodium Sulphur (NaS): Tokyo Electric Power Company In power system applications, TES can be used for diurnal
and NGK Insulators Ltd, Japan, have developed and improved load-leveling, peak shaving, and also for off-grid applications.
the NaS battery. The positive electrode of NaS battery is To prevent heat loss, TES needs a significant amount of
built from liquid/molten sulfur, and the negative electrode is insulation that increases infra-structural costs. The storage can
constructed from liquid/molten sodium. The electrodes are be employed for residential and industrial applications.

305
 TABLE II
E. Virtual Energy Storages (VES) TIME BASED ENERGY STORAGES
By controlling the demand of consumers intelligently, YES
technology is able to compensate the load demand between short time medium time long time
peak and off-peak periods [4] [29]. Until now, two types (seconds to minutes) (minutes to hours) (hours to days)

of YES systems [4], namely, (a) demand side management SMES Li-ion batteries PHES
and (b) demand response techniques are proposed. In demand supercapacitor flow battery flow battery
side management, the less sensitive loads are switched off for FES NiCd battery CAES
short periods. In demand response, more energy is purchased capacitor TES, EDV, VES H2 and fuel cell
than the actually required at light load periods. For example, PbA battery PbA , NaS battery NaS battery
temperature of cold-storage can be decreased further during Li-ion battery metal air, zebra TES (hot water)
off-peak periods. Alternatively, the plant can be kept switched­
H2 and fuel cell TES (borehole)
off during peak periods.
TABLE III
F Electrical Drive Vehicles (EDV) Batteries FIELDS OF ApPLICATION IN POWER SYSTEM
In this concept, intelligent way of charging the BES
systems, used in EDV, can be utilized in a power system power quality energy management
network to compensate the frequency deviations and load (seconds) (minutes) (hours )

leveling purposes. By the vehicle to grid concept, it is voltage support spinning reserve peak shaving

expected that the combustion engine of vehicles will be frequency unbalanced load increase of renewable
replaced by electric motors. The motors need power from regulation compensation power penetration

battery to run. This concept is based on the idea that the flicker compensa­ uninterruptible energy trading iarbitrage
tion power supply
electrical vehicles are typically idle for more than 20 hours
transient stability blackstart cap. source leveling
per day [4] and the fleet of electric vehicles could be
connected to the grid during the idle time, essentially to serve LVRT support grid reliability

as an energy storage. islanded operation

line upgrade deferral

III. ENERGY STORAGES CHARACT ERISTICS AND A REAS

O F ApP LI CATION B. Energy Storage Technologies Based on Time Scale
A. Technical And Economical Characteristics of ES Technolo­ Table II illustrates the ES technologies based on discharge
gies time. The SMES, FES, capacitor, EDLC and few batteries
The technical and economical characteristics of ES tech­ are capable to charge and discharge very quickly, whereas
nologies are essential to establish for comparison purposes and most of the BES systems along with TES, EVD, and YES
to facilitate the selection of appropriate technology for power can be categorized as medium-time based ES systems. Table II
system applications. Table I summarizes the characteristics of further demonstrates that few BES system, TES, flow batteries,
major ES technologies; the values in Table I are extracted PHES and CAES have the capability to discharge for along
mainly from [8], [9], [22], [23], [25], [26] and [30]. Table I period of time. Thus, few of the BES systems can be employed
shows that transitional response times for PHES and CAES both for power quality and energy management applications.
are high, but they can deliver huge energy (large discharge If large amount of power is needed to discharge quickly,
time). Contrarily, FES, SMES, SCIEDLC and capacitor have then electrical ES systems should be the best choice among
very fast transitional response times, long life cycles and very all the ES technologies. Only the BES systems have the
high efficiencies; but energy costs and self discharges are flexibility to provide service of discharging for short, medium,
very high, and cannot deliver power for a long time. The and even long time. Table II further illustrates that the NaS,
energy related costs (lkWh) for PHES and CAES are low, flow batteries, and hydrogen systems can be put to the long
but the power related costs (lkW) for them are relatively high. discharge-time category.
However, the energy densities of BES systems, especially the
NaS, Li-ion, zebra, flow batteries, and H2 are high along with
their medium level costs, reasonable life cycles, and good C. Areas of Power System That Need ES Support
efficiencies. Table I further illustrates that PbA is relatively less Table III demonstrates the areas of power system ap­
expensive among the BES systems, but the power and energy plications, categorized based on time. Two major areas of
densities for PbA are low. Although the energy densities for power system applications are power quality, and energy
metal air and H2 fuel are highest among the BES technologies, management. Power quality applications can be composed of
their overall efficiencies are very low. Presently, NaS, flow frequency regulation, voltage support, flicker compensation,
batteries, and Li-ion batteries have been widely considered transient stability, spinning reserve, and low voltage ride
for power system applications. The TES systems are mostly through, whereas load leveling, energy arbitrage, peak shaving
used for energy management purposes in small industrial and etc can be termed as energy management applications. The
household applications. application fields are briefly described below.

306
 TABLE I
COMPARISON OF TECHNICAL CHARACTERISTICS AND COSTS OF ES TECHNOLOGIES

energy power Life res- typical self cost operating efficiency

Type density density cycle ponse discharge discharge /kW /kWh temp. %

(Wh/kg) (W/kg) (cycles) time time ($) ($) °C

PHES 0.3-t.5' -
> 25 yrs min. 20-t00 hrs negligible >1000 15-150 -
70-87

CAES 10-30 -
> 25 yrs min. 5-50 hrs negligible 400-1000 10-100 -
40-60

FES 5-70 1000-5000 >20000 ms 10 s-40 min very high >600 800-2200 -
95

PbA 20-45 25-100 200-2000 sec 1-6 hrs low >200 <300 -10 to 40 70-80

Lil 100-200 360 500-4000 sec 1-3 hrs low >300 >1200 -30 to 60 70-95

NaS 120-240 120-220 3000-9000 sec 6-12 hrs medium >1000 400-800 270-350 85

NiCd 40-70 140-180 500-2000 sec 1-6 hrs low >200 400-600 -40 to 50 60-75

NiMH 60-80 220 <3000 sec 10 min- 2hrs medium >300 600-1600 -30 to 50 50-80

VRB 25 80-150 >16000 sec 1-15 hrs very low >400 >500 o t040 80

NaNiCI2 80-125 120-180 <3500 sec 2-10 hrs negligible 200-400 200-600 -40 to 70 >90

metal air 200-2000 -

100-400 sec 1-5 hrs low 100-250 <300 -20 to 50 50

H2 fuel 500-3000 >400 >1000 >sec 1-8 hrs low >1200 <100 ambo temp. 40

Capacitor 5-25 >1000 >20000 < ms Is-lOs very high -

>10000 -60 to 80 80-90

SC/EDLC <50 >6000 >50000 < ms Is-30s very high -

>25000 -60 to 80 95

SMES 1-11 >10000 20 yrs < ms Is-40s high 300 >20000 -270 to -250 > 95

T ES 80-200 80-300 5-12 yrs >sec 1-24 hr medium -

10-40 -
40-60

note: 'energy density (Wh/kg or Wh/L) for PHES depends on reservoir height.

1) Frequency Regulation And Transient Stability: In a slow by the converter system (e.g. STATCOM).
dynamic response system, transient instability or grid angular 5) Spinning Reserve: Spinning reserve can be defined as
instability results due to the variation of load demand. The the amount of generation capacity that can be used to produce
differential power between load and generator changes the active power over a given period of time and which has not yet
rotational speed of synchronous generators, and deviates cor­ been committed to the production of energy during this period
responding grid frequencies. The situation can be improved by [31J. Spinning reserve refers to standby unused generation.
the use ES system. When a large number of wind power plants ES systems are faster and can be used as a spinning reserve.
are integrated in a grid, the rapid variation of wind resources Generally, spinning reserve regulates system frequencies.
causes the deviation of system frequency continuously. ES 6) Unbalance Load Compensation: The unbalance load can
systems characterized by short response times able to mitigate be compensated using three phase converter system along
the grid-frequency deviations. with ES. The control of ES is made in such a way that the
2) Voltage Support: The supply of required reactive power injecting and absorbing of energy in ES are performed based
to a load can maintain network voltage at its nominal value. on independent phases.
Generators are responsible to provide both real and reactive 7) Uninterruptible Power Supply: ES systems can be
power. If there is a large demand for both real and reactive placed in a feeder to supply a local load. The fast response
power, an ES system can share the real power with the and appropriate size ES is able to supply load demand, at least
generators, so that, the generators can deliver needed reactive for a short period of time before restoring the main power. In
power for the load. Thus, ES technology can support a system some area of Alaska, ES systems are used to supply the local
to prevent from voltage sag. load power at least for fifteen minutes, so that, the standby
3) Flicker Compensation: Flicker is a voltage fluctuation diesel generators can be started up.
that varies the intensity of light, generally caused by a time 8) Black Start: If a power system network is down due
varying load, e.g. arc furnace. The fast-response energy storage to a fault, then the process of restoring the system is called
system, e.g. EDLC, with static synchronous compensator black start. Generally, diesel generator systems are being used
(STATCOM) is able to mitigate the flickers both for grid­ for starting the power stations. However, an ES system along
connected and off-grid network. with a converter is capable to provide power to the grid, and
4) Low Voltage Ride Through (LVRT): LVRT capability consequently restore the power stations by energizing the field­
keeps equipment and network operational by disconnecting coils of the synchronous generators.
temporarily and then reconnecting the load, or staying con­ 9) Load Leveling: When an ES system is placed in a load
nected with the support of reactive power during grid distur­ bus of a power system, the variation of load can be leveled
bance/fault. During LVRT, wind turbines are required to keep by storing energy during peak times and releasing the energy
running, which can be done easily with the addition of ES. during off-peak periods. In USA, few large PHES are used
The ES systems are charged during fault by the active power for diurnal load leveling, i.e. pumps run in the night time, and
produced from the wind farm, and thus protect the DC link generators provide power during daytime. Thus, the use of ES
capacitor against over voltage. The reactive power is supplied reduces the need for conventional peaking power plant.

307
 10) Source Leveling: It is obvious that wind and solar I power quality I grid support loff.grid applications I energy management I
resources are unpredictable and intermittent. Thus, by forecast­ I flow batteries
tHES 1-11
ing weather, and consequently predicting daily average output, j� metal air NaS )TES 1
ES systems can be used in a renewable power plant to provide
i
1
1
Zebra
J - .{ FAES
J
1
constant power into the network. � LI Ion
L _L _______I

11) Peak Shaving: This application permits to reduce !�r PbAcld

J
J
L ,
J
.I
power consumption during peak periods, and to increase the
power consumption during off-peak periods. Thus, ES systems
are used to charge during off-peak period and discharge during
peak periods. The difference between peak shaving and load
leveling lies on the applied time. The peak shaving times are
comparatively smaller than that of load leveling.
12) Power System Reliability: Power system reliability power rating --

can be evaluated based on system up-time, down-time, and

Fig. 2. power rating and discharge time of energy storage systems
delivery of required power that may result partial loss. ES
systems can be used in a distributed way, so that, they can
deliver the extra power to the load, and thus ensures reliable vehicle applications. The FES and SMES are presently used
power to the consumer end. for improving power qualities both in grid-connected and off­
13) Transmission Up-grade Deferral: Many transmission grid systems. On the other hand, PHES, CABS and TES can
systems were built few decades back. In the meantime, the deliver power from 10 MW to few GW for several hours. Their
demand increases that causes the transmission system over characteristics lead them to employ for load leveling/energy
burden, especially during peak periods. The upgrade of those management in grid-connected and off-grid applications. Even
transmission system is not only expensive but also time though PHES and CABS have enough qualities to use in off­
consuming. Thus, ES system can be used for the interim period grid energy management applications, they are not utilized
in order to make the grid self-sustainable. ES system can be for the same mainly due to economy of scales and lack of
placed in the load bus, and grid flow can be regulated based suitable geographical sites. However, most of the BES systems
on demand. able to deliver power in the range of few kW to 100 MW
14) Renewable Power Penetration: The intermittent nature for few minutes to several hours. Thus, the BES systems can
of renewable resources, e.g. wind and solar, causes voltage be used for improving power qualities, grid supports, electric
and frequency variations in the grid. Generally, integration vehicles, off-grid applications. Among the BES technologies,
of small renewable resources into grid requires some flexible flow batteries, NaS, Li-ion and zebra batteries have shown
and controllable balancing-power from conventional generator. hours to days discharge time at a power range from 5 kW to
The fast-growing and wide spread use of wind, solar and 100 MW. The power ratings and discharge times of Li-ion and
tidal power deteriorate the power quality in the grid. Thus, zebra batteries are lower than that of flow batteries and NaS.
an enhanced level of renewable power penetration must be The PbA, NiCd and NiMH batteries are relatively cheap, their
supported by higher capacity balancing power, which can be applications reside mainly in power quality improvement. The
done by an ES system. low energy-density capabilities along with disposal problem
15) 1sland Grid/ Remote Micro-Grid: Remote micro-grids due to toxicity have diminished their chance to use in large­
usually employ large amount of renewable resources. Other scale utility applications.
than short-time scale variations, renewable energy resources
are very unpredictable. An ES in islanded grid can mitigate
both short-term fluctuations and long-term energy deficiency. E. Examples of ES Technologies In Power System Applications
16) Arbitrage/Bulk Energy Trading: ES system can be used Table IV illustrates a few large ES technologies, which are
to buy energy from grid when the time-of-use pricing is lower, being used in power system applications. The pumps of PHES
and sell back the energy to the grid when the time-of-use price in Fresno, California, USA operates at night, and hydroelectric
is higher. Profit can be earned using the energy arbitrage. If generation delivers power in the day time; thus being used for
ES is placed in a feeder that supplies to a local load, then diurnal load leveling. CABS in McIntosh, Alabama, USA is
grid flow can be regulated based on load variations. In such dedicated for load management, peak-shaving and spinning
a way, cost savings of energy could be done by discharging reserve. Table IV demonstrates that there are very large
BES energy during peak periods. installations of PbA and NiCd batteries for load leveling, peak­
shaving, and power quality applications (Chino, CA, USA and
Fairbanks, Alaska, USA). Although PbA and NiCd batteries
D. Power Rating and Discharge Time of ES are cheap, they are made out of toxic materials, and the PbA
Fig. 2 illustrates graphically the ES technologies based on needs frequent maintenance. Thus, the use of NaS, flow and
power rating and discharge time. Fig. 2 also demonstrates that lithium ion batteries are becoming popular recently. Among
super-capacitorIEDLC, FES or SMES can deliver few kW to them NaS battery becomes highly popular in recent years
20 MW power for few seconds. Thus, they can be used for and large number of project are in practical field (Rokkasho,
frequency regulation, LVRT applications and even in electric Japan). Flow batteries at Kyushu Electric Power, Japan and

308
 TABLE IV
EXISTING A FEW LARGE BATTERY ENERGY STORAGES (BESS) IN POWER B. Equivalent Circuit/ Nonlinear Dynamic BES Model
SYSTEM
Various methods, e.g. chemical process, black box based,
and equivalent circuits, are used to model a BES. Utilizing
des- specification location comments
[35] an equivalent circuit BES model is presented in Fig. 3.
cription power energy The voltages at BES terminal can be calculated as,
(MW) (MWh)

PHES 1212 20GWh California, USA load labeling [321

CAES 110 2860 Alabama, USA spinning reserve [331

FES 20 5 New York, USA frequency regulation

PbA 10 40 Chino, CA, USA load leveling [9], [121

NiCd 27 14.6 Alaska, USA spining reserve [3], [71

NaS 34 245 Rokkasho, Japan peak shaving

Fig. 3. BES model-2nd order Randle model
Zebra -
40 Switzerland spinning reserve [271

Li-Ion 12 5 Chile grid stability

VRB 0.25 2 Utah, USA voltage support [271

V b(t)=Vocv(t)-VRo(t)-V Cl(t)-V C2(t) (5)
ZnBr 1 4 Japana trans. deferral
VRo(t)=Ra[b(t) (6)

/ ( )
PSB 15 120 UK peak shaving [9], [101

V Cl (t)
V Cl (t)= � lb(t)-
SMES -
2.5 MJ Korea power quality
dt (7)
C\ R\

Innogys Little Barford station, UK is used for load leveling C2

J (
V C2(t)= � lb(t)- C2
V (t)
R2
dt ) (8)
and energy management applications. Due to high efficiencies

{
and short response times SMES systems penetrate in small­ For the developed model, the BES SoC can be expressed as,
scale utility (Korea) applications. SOCg -l1b f� [b(t)dt for SoC unit Ah
SoCb(t)- tlb(t)
_

(9)
IV. BES MODELS FOR ECONO MI C EVALUATION AND SOCba -l1b fa (3Jdt for SoC unit in %
POWER SY S T E M ANALY SI S
where Socg is the initial state of charge, and [b(t) instanta­
BES systems are integral part of power systems. Thus, neous charging (negative)/discharging (positive) current. Here,
appropriate models for BES are essential for various kinds the SoC need to be estimated indirectly, through the measure­
of studies. This section illustrates few BES models. ment of currents and voltages. This model can be simulated
in power system analysis softwares (PSCAD, Simulink etc) to
A. BES Models for Optimization, Economic, and Reliability study the electromagnetic transients and dynamics.
Evaluation
For optimization, economic, and reliability evaluation of C. Linearized BES Model
power system application, the BES power and state of charge
Sometimes, small scale dynamic study is required for power
(SoC) in terms of kW and kWh, respectively, can be expressed
system analysis. Fig. 4 represents a simplified linearized BES
from (1) to (4). The equations are modified form, which are
model [36]. where pb* and pb are, respectively, the reference
extracted mainly from [34].

Charging mode,

[
o x-SOCb(t)}
pb(t)=-min pbraP max o, {s Cf,za
llt
( )] (1)

SoCb(t +llt)=SoCb(t) (1 - 8) -l1�pb(t)llt (2) Fig. 4. first order BES model

Discharging mode,
power and actual output power for the BES; Eb is the state of
[
pb(t)=min prbap max o , { (
SOCb(t)-SOCf"in}
llt
)] (3) energy; where P,�ax and P'�in are, respectively, the maximum
and minimum charging power; and T is the time constant for
b
pb(t) the BES. The model can be utilized to study the small signal
SoCb(t +llt)=SoCb(t) (1- 8) - --llt (4)
l1bd stability, e.g. frequency deviations in any hybrid power system.

where, Pfat is the rated charging and discharging power; V. T REND OF BESs GRO W T H
pb(t) is the actual charging and discharging power ; 8 is
b Due to low costs, large number of PbA and NiCd batteries
the self discharge co-efficient of BES, and � is the charger have been employed in power system network since 1980,
TJd
and converter round-trip efficiency. Socf,zax is the maximum although they suffer with low energy densities and less life
capacity of battery; and SoC�in = (1 - DOD)SoC�ax' is the cycles. Table I indicates that the NaS battery and Li-ion
minimum level of battery Soc. This model can be employed battery have enough potential characteristics. The materials
in numerical and steady state simulation. of NaS battery are cheaper than that of Li-ion battery; thus,

309
the use trend of NaS battery in renewable energy system is [13] P. E. Pascoe and A. H. Anbuky, "Standby power system VRLA battery
exponential. It is expected that large-scale production of Li-ion reserve life estimation scheme," IEEE Trans. Energy Convers., vol. 20,
no. 4, pp. 887-895, Dec. 2005.
batteries will decrease production cost, and will subsequently [I4] A. M. Khudhair and M. M. Farid, "A review on energy conservation
penetrate in power system applications. For example, the in building applications with thermal storage by latent heat using phase
Tehachapi Wind Energy Storage project is under test for change materials," Energy Conversion and Management, vol. 45, no. 2,
pp. 263-275, Jan. 2004.
evaluating the performance of an 8 MW, 4 hour (32 MWh) [I5] I. Hadjipaschalis, A. P oullikkas and V. Efthimiou, "Overview of current
Li-ion battery system. and future energy storage technologies for electric power applications,"
Renewable and Sustainable Energy Reviews, vol. 13, no. 6-7, pp. 1513-
1522, Sep. 2009.
VI. CONCLUSIONS [16] H. Lund and G. Salgi, 'The role of compressed air energy storage
Presently, various ES technologies are in use for grid­ (CAES) in future sustainable energy systems," Energy Conversion and
Management, vol. 50, no. 5, pp. 1172-1179, May 2009.
connected and off-grid power system applications. With the [17] S. Talebi Rafsanjan, "Advanced high-speed flywheel energy storage
advent of improved ES technologies, it can be expected that ES systems for pulsed power application," in Proc. IEEE Electric Ship
technologies would play role models to improve power qual­ Technologies Symposium, pp. 379-386.
[I8] L. Joerissen, 1. Garche, C. Fabjan and G. Tomazic, "P ossible use of
ities, to establish economic and reliable power systems, and vanadium redox-flow batteries for energy storage in small grids and stand­
to develop enhanced penetration renewable energy systems. alone photovoltaic systems," Journal of Power Sources, vol. 127, no. 1,
To face various challenges, ES technologies are required in pp. 98-104, Mar. 2004.
[19] D. P. Scamman, G. W. Reade and E. P. L. Roberts, "Numerical modelling
conventional power systems, hybrid renewable energy systems, of a bromide-polysulphide redox flow battery: P art 1: Modelling approach
and in smart grids. The mature ES technologies, e.g. PHES, and validation for a pilot-scale system," Journal of Power Sources, vol.
CAES and few BES (e.g. PbA, NiCd etc) systems have been 189, no. 2, pp. 1220-1230, Apr. 2009.
[20] K. C. Seong, H. J. Kim, S. W. Kim, J. W. Cho, Y. K. Kwon, K. S. Ryu,
utilized in power system applications for few decades; but in I. K. Yu and S. Y. Hahn, "Current status of SMES in Korea," Cryogenics,
many cases, these technologies cannot fulfill the application vol. 42, no. 6, pp. 351-355, Jun.-Jul. 2002.
requirements. Thus, the research continues, and subsequently [21] c. H. Dustmann, "Advances in ZEBRA batteries," Journal of Power
Sources, vol. 127, no. 1, pp. 85-92, Mar. 2004.
new ES technology arrives, bringing with numerous advan­ [22] J. Dixon, I. Nakashima, E. F. Arcos and M. Ortuzar, "Electric vehicle
tages and challenges. Moreover, no single ES technology can using a combination of ultracapacitors and ZEBRA battery," IEEE Trans.
provide high power densities, high energy densities, low costs, Ind. Electron., vol. 57, no. 3, pp. 943-949, Mar. 2010.
[23] J. S. Lee, S. Tai Kim, R. Cao, N. S. Choi, M. Liu, K. T. Lee and
long life cycles, high efficiencies, shortest transition times, J. Cho, "Metal-air batteries with high energy density: Li-air versus Zn­
compactness, and benign environmental disposal. Thus, the air," Advanced Energy Materials, vol. 1, no. 1, pp. 34-50, Jan. 2011.
selection and evaluation of ES technologies should be based [24] D. J. Swider, "Compressed air energy storage in an electricity system
with significant wind power generation," IEEE Trans. Energy Convers.,
on application requirements, costs, and prioritize technical vol. 22, no. 1, pp. 95-102, Mar. 2007.
characteristics; and so the paper is capable to assist. [25] H. Ibrahim, A. Ilinca and J. P erron, "Energy storage systemscharacteris­
tics and comparisons," Renewable and Sustainable Energy Reviews, vol.
12, no. 5, pp. 1221-1250, Jun. 2008.
REFERENCES
[26] P. J. Hall and E. 1. Bain, "Energy-storage technologies and electricity
[I] P. F. Ribeiro, B. K. Johnson, M. L. Crow, A. Arsoy and Y. Liu "Energy generation," Energy Policy, vol. 36, no. 12, pp. 4352-4355, Dec. 2008.
storage systems for advanced power applications," in Proc. IEEE, vol. [27] M. Swierczynski, R. Teodorescu, C. N. Rasmussen, P. Rodriguez and
89, no. 12, pp. 1744-1756, Dec. 2001. H. Vikelgaard, "Overview of the energy storage systems for wind power
[2] H. Chen, T. N. Cong, W. Yang, C. Tan, Y. Li and Y. Ding, "P rogress in integration enhancement," in Proc. 2010 IEEE Industrial Electronics
electrical energy storage system: a critical review," Progress in Natural Symposium, pp. 3749-3756.
Science, vol. 19, no. 3, pp. 291-312, Mar. 2009. [28] c. P once de Leon, A. Frias-Ferrer, J. GonzaJez-Garcfa, D. A. Szanto
[3] D. H. Doughty, P. C. Butler, A. A. Akhil, N. H. Clark and J. D. Boyes, and F. C. Walsh, "Redox flow cells for energy conversion," Journal of
"Batteries for large-scale stationary electrical energy storage," in Proc. Power Sources, vol. 160, no. 1, pp. 716-732, Sep. 2006.
2010 T he Electrochem. Society Interface, pp. 49-53. [29] P. Du, Y. Makarov, N. Zhou and P. Etingov, "Application of virtual
[4] C. Dotsch, "Energy storage," in Technology Guide, New York: Springier, energy storage to partially mitigate unscheduled interchange caused by
2009, pp. 362-367. wind power," in Proc. 2011 IEEE Power Systems ConI, and Expo., pp.
[5] R. Walawalkar and R. Mancini, "Economics of electric energy storage 1-5.
for energy arbitrage and regulation in New York," Energy Policy, vol. 35, [30] S. M. Schoenung and W. V. Hassenzah1, "Long-vs. short-term energy
no. 4, pp. 2558-2568, Apr. 2007. storage technologies analysis, a life-cycle cost study, a study for the
[6] C. Abbey and G. Joos, "Energy storage and management in wind turbine DOE energy storage systems program," Sandia National Laboratories,
generator systems," in Proc. 2006 Power Electronics and Motion Control New Mexico, Tech. Rep., SAND2003-2783, Aug. 2003.
Conj, pp. 2051-2056. [31] Y. Rebours and D. Kirschen, "What is spinning reserve?," The University
[7] J. McDowall, "Integrating energy storage with wind power in weak of Manchester, Tech. Rep., Sep. 2005.
electricity grids," Journal of Power Sources, vol. 162, no. 2, pp. 959- [32] F. P. Bevc and R. J. Meehan, "Generator-motors for pacific gas and
964, Nov. 2006. electric's helms pumped storage project," IEEE Trans. Power App. and
[8] S. Vazquez, S. M. Lukic, E. Galvan, L. G. Franquelo and J. M. Carrasco, Syst., vol. PAS-99, no. 6, pp. 2021-2030, Nov. 1980.
"Energy storage systems for transport and grid applications," IEEE Trans. [33] A. Cavallo, "Controllable and affordable utility-scale electricity from
Ind. Electron, vol. 57, no. 12, pp. 3881-3895, Dec. 2010. intermittent wind resources and compressed air energy storage (CAES),"
[9] K. C. Divya and J. 0stergaard, "Battery energy storage technology for Energy, vol. 32, no. 2, pp. 120-127, Feb. 2007.
power systems: an overview," Electric Power Systems Research, vol. 79, [34] R. Dufo-L6pez and J. L. Bernal-Agustin "Design and control strategies
no. 4, pp. 511-520, Apr. 2009. of PV-Diesel systems using genetic algorithms," Solar Energy, vol. 79,
[10] A. Joseph and M. Shahidehpour,"Battery storage systems in electric no. 1, pp. 33-46, Jul. 2005.
power systems," in Proc. 2006 IEEE Power Engineering Society General [35] R. Lasseter and M. Erickson, "Integration of battery-based energy
Meeting Conj, pp. 8. storage element in the CERTS microgrid," US Department of Energy,
[11] S. Van der Linden, "Bulk energy storage potential in the USA, current DE-FC02-06CHI1350, Oct. 2009.
developments and future prospects," Energy, vol. 31, no. 15, pp. 3446- [36] T. Senjyu, A. Uehara, A. Yona and T. Funabashi, "Frequency control
3457, Dec. 2006. by coordination control of wind turbine generator and battery using H
[12] M. D. Anderson and D. S. Carr, "Battery energy storage technologies," infinity control," in Proc. 2009 Trans. & Distri. Conj & Expo., Asia and
Proc. IEEE, vol. 81, no. 3, pp. 475-479, Mar. 1993. Pacific, pp. 1-4.

310