Applied Energy 137 (2015) 545553

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Review of energy storage system for wind power integration support

Haoran Zhao a,b, Qiuwei Wu a,b,, Shuju Hu c, Honghua Xu c, Claus Nygaard Rasmussen a
a

Center for Electric Power and Energy, Department of Electrical Engineering, Technical University of Denmark, Kgs. Lyngby 2800, Denmark
Sino-Danish Center for Education and Research, Aarhus 8000, Denmark
c
Institute of Electrical Engineering, Chinese Academy of Science, Beijing 100190, China
b

h i g h l i g h t s
 The principle and characteristics of present EES technologies are reviewed.
 The potential applications of different roles played by EES are described.
 The recent researches about the EES planning problems, including type selection, optimal sizing and siting are summarized.
 The operation and control strategies are discussed.

a r t i c l e

i n f o

Article history:
Received 21 January 2014
Received in revised form 10 March 2014
Accepted 30 April 2014
Available online 23 May 2014
Keywords:
Energy Storage System (ESS)
Optimal siting
Optimal sizing
Planning and operation
Wind power

a b s t r a c t
With the rapid growth of wind energy development and increasing wind power penetration level, it will
be a big challenge to operate the power system with high wind power penetration securely and reliably
due to the inherent variability and uncertainty of wind power. With the exible chargingdischarging
characteristics, Energy Storage System (ESS) is considered as an effective tool to enhance the exibility
and controllability not only of a specic wind farm, but also of the entire grid. This paper reviews the
state of the art of the ESS technologies for wind power integration support from different aspects. Firstly,
the modern ESS technologies and their potential applications for wind power integration support are
introduced. Secondly, the planning problem in relation to the ESS application for wind power integration
is reviewed, including the selection of the ESS type, and the optimal sizing and siting of the ESS. Finally,
the proposed operation and control strategies of the ESS for different application purposes in relation to
the wind power integration support are summarized. The conclusion is drawn in the end.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
Nowadays, as the most popular renewable energy source (RES),
wind energy has achieved rapid development and growth. According to the estimation of International Energy Agency (IEA), the
annual wind-generated electricity of the world will reach
1282 TW h by 2020, nearly 371% increase from 2009. By 2030, that
gure will reach 2182 TW h almost doubling the year 2020 production [1].
Due to the intermittent nature of wind power, the wind power
integration into power systems brings inherent variability and
uncertainty. The impact of wind power integration on the system
stability and reliability is dependent on the penetration level [2].
From the reliability perspective, at a relative low penetration level,
Corresponding author at: Center for Electric Power and Energy, Department of
Electrical Engineering, Technical University of Denmark, Kgs. Lyngby 2800,
Denmark. Tel.: +45 4525 3529; fax: +45 4588 6111.
E-mail address: qw@elektro.dtu.dk (Q. Wu).
http://dx.doi.org/10.1016/j.apenergy.2014.04.103
0306-2619/ 2014 Elsevier Ltd. All rights reserved.

the net-load uctuations are comparable to existing load uctuations [3], and the Conventional Generators (CGs), such as thermal
or hydro units, have sufcient load tracking capability without
requiring additional operating reserve. As the wind penetration
level increases, the response time of committed CGs should be
short enough during sudden and large changes of wind power production and load due to random failures and wind gusts, and more
operating reserves will be required. From the stability perspective,
different from synchronous generators, Wind Turbine Generators
(WTGs) provide only small or even no contribution to frequency
stability [4]. The wind power variation can also degrade the grid
voltage stability due to the surplus or shortage of power [5]. An
Energy Storage System (ESS) has the ability of exible charging
and discharging. Recent development and advances in the ESS
and power electronic technologies have made the application of
energy storage technologies a viable solution for modern power
application [6]. The potential applications mainly cover the following aspects. Through time-shifting, the power generation can be

546

H. Zhao et al. / Applied Energy 137 (2015) 545553

regulated to match the loads. The ESS can also be used to balance
the entire grid through ancillary services, load following and load
leveling [7]. Moreover, it can meet the increasing requirement of
reserves to manage the uncertainty of wind generation [8] which
can increase the system operation efciency, enhance power
absorption, achieve fuel cost savings and reduce CO2 emissions.
Additionally, the ESS is a potential solution to smooth out the uctuations, and improve supply continuity and power quality [9].
For a specic application, the rst task of an ESS project is planning. It generally includes the type selection and size determination. Sometimes, the ESS siting also needs to be considered.
Several factors, such as technical features, economical cost and
local wind power characteristics, can inuence the ESS selection
[10]. Once a specic ESS type is chosen, the optimal sizing needs
to be done by balancing the benets and cost. If there are no geographical constraints, the ESS could be optimally installed to
achieve the maximum benet, mainly in the reduction of transmission system upgrade cost.
The operation and control strategies of the ESS are designed for
different application purposes. The recent studies mainly focus on
the coordinated control of wind farms and on-site ESSs. The shortterm (daily or hourly) dispatch scheme of an ESS and uctuation
smoothing by a wash-out lter are the two attractive areas. It is
also proposed to combine many dispersed ESSs as a virtual storage
unit and control centrally [10]. Since the ESS is an expensive solution, it is not economically viable for the ESS to work for a single
application service. It can also contribute to the system wide
control.
This paper is to review the state of the art of the ESS technologies and the applications for the wind power integration support
from different aspects. The paper is organized as follows: Section
2 introduces the principle and characteristics of present ESS technologies. The potential different roles played by the ESS for the
wind power integration support are described in Section 3. The
recent research of the ESS planning problem, including type selection, and optimal sizing and siting, are summarized in Section 4.
Finally, the operation and control strategies of the ESS for wind
power integration support are discussed in Section 5. The conclusion is drawn in Section 6.
2. Energy storage technologies
The electrical energy can be stored in different energy forms:
mechanical, electro-chemical, chemical, electromagnetic, thermal,
etc. [3,7]. The classication of energy storage technologies according to the stored energy form is illustrated in Fig. 1.

There are various characteristics of the ESS required to be taken

into consideration for different applications, including capital cost,
power and energy rating, power and energy density, ramp rate,
efciency, response time, self-discharge losses, and life and cycle
time [11,12]. The overview of the capital cost and the technical features of the ESS is listed in Tables 1 and 2, respectively.
The technical details of the ESS have been described in many literatures [10,11,13,14]. A short description of the principles and
potential capability of several commonly used ESSs for wind power
integration support is presented in this section.
2.1. Pumped Hydro Storage (PHS)
The PHS is the largest and most mature energy storage technology available [15]. It represents nearly 99% of the worldwide
installed electrical storage capacity with over 120 GW [10,16].
The conventional PHS consists of two water reservoirs. The water
body at the relatively high elevation represents the potential or
stored energy. During off-peak hours, it pumps water from the
lower reservoir to the upper one, considered as a charging process.
In the discharging process, water from the upper reservoir is
released and ows through hydro turbines which are connected
to generators, producing electrical energy [14].
As illustrated in Table 2, the PHS has the largest power and
energy rating, long lifetime, high efciency and very small discharge losses. The main applications of the PHS for wind power
integration are energy management via time-shifting, frequency
control and non-spinning reserve supply. Due to the slow response,
the PHS is not suitable for suppressing wind uctuations. The
installation of the PHS is dependent on geographical conditions
and has an impact on the nature environment. Therefore, the exibility of its application is low.
The economic benets of the PHS combined with Wind Farms
(WFs) are described and analyzed in [17,18] shows that the hybrid
PHS-WF system can meet the hourly energy demand.
2.2. Compressed Air Energy Storage (CAES)
The CAES is a technology known and used since the 19th century for different industrial applications [10]. Electrical compressors are used to compress air and store it in either an
underground structure (salt cavern, abandon mines, rock structures) or an above-ground system of vessels or pipes. When
needed, the compressed air is released and mixed with natural
gas, burned and expanded in a modied gas turbine. Current
research on the CAES is focused on the development of systems

Fig. 1. Energy storage classication.

547

In (cycles)
In (years)

0.11.4
24
325
20100
20100
15100
820
580
580
600020,000
220


Efciency (%)

with fabricated storage tanks which will remove the geological

dependency and the compressed air will be stored with higher
pressure. So far, there are only two CAES units in operation. They
are located in Huntorf, Germany and MacIntosh in Alabama, USA
[19]. There are several CAES units which are either planned or
under construction [20].
From Table 2, it is shown that the high power and energy capacity rating makes the CAES another choice for wind farms for the
energy management purposes, similarly to the PHS. The storage
period can be over a year due to very small self-charge losses
[11]. However, the CAES installation is also limited by topographical conditions.

Very small
Small
100
0.10.3
0.20.6
0.10.3
20
Small
Small
0
2040
1015

$ (kW h-per cycle)

5100
250
10005000
200400
8001500
6002500
300500
1501000
1501000

3002000
100010,000

Response time

7080
4175
8090
7590
6080
6575
7085
6075
6575
3444
8598
7580

$ (kW h)

6002000
4008000
250350
300600
5001500
12004000
10003000
6001500
7002500
10,000+
100300
200300

min
min
<s
<s
<s
<s
<s
s
s
smin
<s
<s

$ (kW)

Lifetime

PHS
CAES
FES
LA
NiCd
Li-on
NaS
VRB
ZnBr
FC
SC
SMES

Capital cost

Self- discharge per day (%)

System

>50
>25
1520
315
520
5100
1015
520
510
1030
412


Table 1
Capital cost of ESS.

>15,000
>10,000
104107
2501500
15003000
6001200
25004500
>10,000
10003650
103104
104105


H. Zhao et al. / Applied Energy 137 (2015) 545553

0.10.2
0.20.6
5000
90700
75700
130010,000
120160
0.52
125
0.220
40,000120,000
2600

Energy density (W h/l)

Power density (W/l)
Discharge time typical

124 h+
124 h+
sh
sh
sh
minh
sh
s10 h
s10 h
s24 h+
ms1 h
ms8 s
1005000
5300
00.25
020
040
00.1
0.058
0.033
0.052
050
00.3
0.110

Power rating (MW)

System

The BESS stores electricity in the form of chemical energy [22].

A conventional secondary battery consists of a set of low-voltage/
power battery cells connected in parallel and series to achieve a
desired electrical characteristic. Each cell is made up of a liquid,
paste or solid electrolyte together with anode and cathode [11].
A battery is charged by an internal chemical reaction under a
potential applied to both electrodes. The reaction is reversible
and let the battery deliver the absorbed energy for discharging.
So far, various types of second batteries have been developed for

Table 2
Technical features of ESS.

2.4. Battery Energy Storage System (BESS)

PHS
CAES
FES
LA
NiCd
Li-on
NaS
VRB
ZnBr
FC
SC
SMES

Density
Rating

The rst generation of the FES has been available since 1970s
which uses a large steel rotating body on mechanical bearings. In
the FES, the rotational energy is stored in an accelerated rotor, a
massive rotating cylinder [10]. The main components are a rotating
cylinder (comprised of a rim attached to a shaft) in a compartment,
bearings and a shaft. The whole structure is placed in a vacuum
enclosure to reduce windage losses. During the charging process,
the rotor is accelerated to a very high speed which can reach from
20,000 to over 50,000 rpm. The energy is stored in the ywheel by
keeping the rotating body at a constant speed. During the discharging process, the ywheel releases energy and drives the machine as
a generator.
The main advantages of ywheels are the excellent cycle stability, a long life of providing full chargedischarge cycles, little maintenance cost, high power density and high efciency. The FES is
mainly applied as a power quality device to suppress fast wind
power uctuation, provide ride-through of interruptions of several
seconds or bridge the shift between two sources [11]. Besides, it is
also designed to provide damping enhancement [21]. The main
drawbacks are the short operation duration and high self-discharge
losses. They are considered as a support for wind turbines in combination with other ESSs rather than standing alone [13].

0.22
26
2080
5080
1580
200400
15300
2070
65
600 (200 bar)
1020
6

2.3. Flywheel Energy Storage (FES)

548

H. Zhao et al. / Applied Energy 137 (2015) 545553

commercial use, including Lead Acid (LA) battery, Nickel Cadmium

(NiCd) battery, Nickel Metal Hybrid (NiMH) battery, Lithium Ion
(Li-ion) battery and Sodium Sulphur (NaS) battery.
As illustrated in Table 2, as a whole, secondary batteries have
very rapid response time (<s) which allows tracking load changes
for the system stability enhancement. The self-discharge loss is
small and the round-trip efciency is high. Due to the high power
and energy densities, the BESS construction is facilitated by the
short lead time, potential convenient sitting and technology modularity [11]. However, most batteries contain metal toxic materials
which lead to an ecological problem for disposal. For the intercomparison, the limited number of cycle lifetime restricts the
application of LA batteries as large-scale storage devices. The life
span of NiCd batteries is seriously dependent on the cycle depth.
A deep cycle can reduce the NiCd battery lifetime signicantly.
Besides, this technology suffers from the memory effect [15]. The
Li-ion battery has the highest power and energy densities. The
research related to the Li-ion battery focuses on its application of
electrical vehicles. The high capital cost, shown in Table 1, limits
the large-scale use of the Li-ion battery for wind power integration
support. The NaS battery is an economical solution for power quality improvement and peak-shaving applications. However, due to
the high operation temperature (300340 C), the stored energy
is partly used for heating which reduces the operation performance. The high cost is another barrier.
The ow battery is another type of battery. The energy is stored
in one or more electro-active species which are dissolved in liquid
electrolytes [10]. Additional electrolyte is stored externally, generally in tanks and is usually pumped through the cells of the reactor.
The energy rating is determined by the quantity of electrolyte
whereas the power rating is dependent on the active area of the
cell stack [11]. The typical ow batteries are Vanadium Redox Battery (VRB), Polysulphide Bromide (PSB) and Zinc Bromine (ZnBr).
Flow batteries have been built in MW class and can play a more
important role for future large scale application. In [5], the ZnBr
battery is used to dispatch the wind power based on the optimal
control method. In [23], a washout lter-based scheme is adopted
to smooth out short-term power uctuations of a wind farm with
VRBs.
2.5. Superconducting Magnetic Energy Storage (SMES)
The SMES consists of superconductive coil, power conditioning
system, refrigerator and vacuum [15]. The energy is stored in the
magnetic eld created by DC current circulating through a superconducting coil [10]. In order to avoid the losses caused by the current ow, the coil is kept in the superconducting state.
The SMES has very rapid response. The power requested is
available almost instantaneously [10]. The SMES is very promising
as a power storage system for load leveling or a power stabilizer
[24,25]. The SMES can be incorporated into a back to back DC link
[25]. In this case, a back-to-back system is used as a power conditioning system for the SMES coils. It is also utilized by the coordination with wind farms for power quality improvement [24,2628]
and dynamic stability enhancement [21,29]. However, the superconductive coil is very sensitive to temperature changes. The operational reliability is crucially dependent on the refrigeration
system. Up to now, only a few SMES with small capacity are available for commercial use.
2.6. Super-Capacitor (SC)
Great progress has been achieved in the capacitor storage technologies. Instead of the common arrangement of a solid dielectric
between the electrodes, an electrolyte solution is placed between
two solid conductors for the SC. Compared with conventional

capacitors, it has much larger capacitance and energy density, thus

enabling a compact design [10,11].
The SC has nearly unlimited cycle stability as well as extremely
high power density, and fast charging and discharging due to
extraordinarily low inner resistance. Other advantages are durability, high reliability, no maintenance, long lifetime, and operation
over a wide temperature range and in diverse environments. They
are environmentally friendly and easily recycled or neutralized.
The efciency is typically around 90% and the discharge time is
in the range of seconds to hours. The current research for wind
power integration support focuses on the power leveling of wind
farms [30], coordination with batteries for smoothing fast uctuations [31].
Other ESS technologies, including Fuel Cell (FC), MetalAir (MA)
battery, Solar Fuel, Cryogenic Energy Storage (CES), Synthetic Natural Gas (SNG) and Thermal Energy Storage (TES) are either still
under development or technically developed, but still not widely
used. The technical maturity of different types of ESSs is shown
in Fig. 2.
Different applications require different technical features of the
ESS. Among them, energy and power ratings are the two main factors. In [10], a comparison of several storage technologies based on
these factors is illustrated in a double-logarithmic chart (Fig. 3).
3. ESS applications for wind power integration support
The ESS applications related to wind power integration can be
summarized and categorized in terms of roles it plays for different
stakeholders: the wind farm owner, the grid operator and the
energy consumer.
3.1. Generation-side roles of ESS
The main challenges with wind power integration are power
intermittency, ramp rate and limiting wind farm output [32]. The
generation-side role of the ESS aims to improve the grid-friendliness of the wind farm to dispatch wind energy such that they
can be controlled like conventional power plants. Additionally, it
shall be controlled to effectively utilize limited transmission
capacity.
3.1.1. Time shifting
Due to the stochastic characteristic of wind, wind power production is considered as a non-dispatchable resource and sometimes it demonstrates an anti-peak feature, e.g. high wind power
during off-peak demand or low wind power during peak demand.
The time shifting is to store extra wind energy during periods of
low demand and stands ready to dispatch energy to the grid during
periods of high demand [7]. The benet of storing electricity is
expected to be larger with the large gap of demand between peak
and off-peak. To fulll the time shifting function, large quantities of
energy for signicant periods of time (from hours to days) are
required for the ESS facility. Besides, the storage efciency is
another key factor to be considered for the economical operation
of time shifting, as signicant losses occur for an inefcient storage
system.
3.1.2. Output smoothing
The inherently variable nature of wind power can cause uctuations in frequency and voltage [7]. The ESS can be used to smooth
out these uctuations to keep the system stable. Accordingly, the
output power of the ESS needs to be rapidly regulated for absorbing the excess energy during output spikes and releasing energy
during output drops. Therefore, the ramping capability is very
important for the smoothing function. The output smoothing at

549

H. Zhao et al. / Applied Energy 137 (2015) 545553

Table 3
Grid-side roles of ESS.
Application

Time scales

Example of
EES

Energy arbitrage, load leveling

Hours to days

Frequency regulation

Seconds to
minutes
<1 s

PHS, NaS,
CAES, VRB
Li-ion, NaS,
FES, VRB
LA, NAS, FES,
VRB
PHS, FES, BES
PHS

Inertia emulation, oscillation damping,

voltage support LVRT
Primary reserves
Secondary reserves
Efciency use of transmission network
Fig. 2. Technical maturity of ESSs [11].

the plant level reduces the need for power quality and ancillary
services at the system level [33].
3.1.3. Transmission utilization efciency
Rich wind resources are often located in rural areas far from
existing high capacity transmission lines [34]. Due to the transmission constraints, the energy produced may not be transferred to
the load. Additional ESS can mitigate transmission congestion,
defer or avoid transmission and distribution upgrades.
3.2. Grid-side roles of ESS
Currently, the ESS is required by the grid operator to provide
ancillary services to mitigate the variability and uncertainty of
the entire grid, rather than specic loads or wind farms. These
applications are listed in Table 3. Due to the geographical distribution of wind resources, the net variability and uncertainty are less.
Therefore, the need of the overall service is reduced [7].
3.2.1. Energy arbitrage/load leveling
In the electricity market, the electricity price varies from time to
time, normally hourly [14]. The ESS can be used to store low-cost
off-peak energy and releases when the price is higher. It can reduce
market risk exposure to volatile on-peak prices and manage high
cost energy imbalance charges [35].
3.2.2. Frequency regulation
Modern wind farms are required to provide frequency regulation by the grid operator. With high wind penetration level, providing frequency response from a wind farm is technically feasible by

Emergency power supply, black start

10 min
Minutes to
hours
Minutes to
hours
Minutes to
hours

Li-ion
LA

utilizing additional droop control. However, it may cause fatigue of

wind turbines and instability problem [36,37]. An effective solution is the use of the ESS. For the primary frequency control, a local
droop control loop can be added to the active power controller of
the ESS. The droop control aims to produce an active power output
change which is proportional to the frequency deviation [38]. For
the secondary frequency control, the active power command is
generated by the centralized Automatic Generation Control (AGC).
3.2.3. Inertia emulation
The grid inertia reduces frequency variability and makes the
grid less sensitive to sudden generation changes. The instantaneous inertial response determines the Rate of Change of Frequency (ROCOF) [4]. The addition of the ESS could signicantly
increase the apparent inertia of a grid. The supplementary loop
can be added to the active power control of the ESS.
3.2.4. Oscillation damping
In an interconnected system, sudden changes of power in tieline might cause oscillations with frequency range between 0.5
and 1 Hz [14]. It may further result in synchronism loss of several
machines. Application of a damping controller is an effective control scheme to simultaneously handle the inherent power uctuations and enhance system stability for a large wind farm [26]. The
SMES and FES are utilized in [21,29] to damp the system oscillation. The tie-line power deviation is used as a feedback signal to
generate the phase component of the converter control.
3.2.5. Voltage control support
The wind power variability can degrade the grid voltage stability [5]. The installed ESS can provide adequate reactive power to
maintain the local voltage level. This service can be obtained by
the full scale converter connected to the grid [13].
3.2.6. Low Voltage Ride Through (LVRT) support
WTGs should have LVRT capability to remain connected during
severe grid faults specied by grid codes [36]. Furthermore, some
gird codes require that WTGs supply up to the maximum reactive
current during such faults. The converter should draw real power
to compensate for the switching losses associated with provision
of the reactive power. During severe faults, no power can be drawn
from the grid. As a result, the DC voltage falls and the converter
switches are blocked. For such cases, the ESS can support the DC
voltage during the faults.

Fig. 3. Comparison of rated power energy content of different ESS technologies.

3.2.7. Reserve application

Due to the forecast error of wind power, additional reserves are
required for emergency support. Based on the response time, the

550

H. Zhao et al. / Applied Energy 137 (2015) 545553

reserves can be generally classied into primary, secondary and

tertiary reserves [14].
3.2.8. Emergency power supply/black start
The ESS may be used to restart from a shut-down condition
without the assistance from the electrical grid and energize the
power system in the event of a catastrophic failure [7,14].
3.2.9. Transmission utilization efciency
The ESS can help grid operators efciently use the transmission
system capacity, defer transmission system upgrades to reduce
transmission costs and mitigate local dependency challenges of
wind power.
3.3. Demand-side roles of ESS
Most existing ESS applications for energy consumers are more
related to meet the energy needs rather than solving particular
challenges related to the integration of large-scale RES [39]. Only
one application has a signicant support for the wind power integration support: Vehicle-To-Grid (V2G) [7,40].
Due to the aggregation effect of many Electrical Vehicles (EVs)
plugged into the grid, these EVs can be considered as a Virtual
Power Plant (VPP) with relatively large capacity. This EV VPP
(EVPP) has to fulll the requirements of both vehicle owners and
grid operators. Since all the EVs are controlled as a whole, individual vehicles will not be locked in the charging station and its owner
has the full convenience. The grid operator can treat this VPP as a
provider of ancillary services [41], such as time-shifting, operation
reserve, and frequency regulation.
So far, many efforts have been made to investigate the EVPP feasibility and possible architecture [42]. One nished project is the
EDISON project nanced by the Danish TSO Energinet, DK. The
Danish island, Bornholm, was used as a test site. It aims to coordinate charging and discharging of EVs in order to optimize the utilization of wind energy in the island grid.
Since the ESS is an expensive solution, it is not economically
viable for the ESS to work for single application service. In [43],
the installed ESS is mainly used for output smoothing. It can also
contribute to system wide control such as frequency regulation
or oscillation damping. In [36], the applied BESS has the ability
to provide both frequency response service and energy timeshifting.
4. Energy storage system planning
4.1. ESS type selection
Besides the aforementioned technical characteristics of the ESS,
the wind power uctuation density at different time scales is
another key factor for the ESS selection to satisfy specic application purposes. Low frequency uctuation, ranged from minutes to
hours, is related to the generation reserve and the energy dispatch
of the power system, while high frequency uctuation, ranged
from seconds to minutes, affects power system frequency control
[44]. Based on the local wind data, the corresponding Power Spectrum Density (PSD) can be derived [45]. Through the PSD analysis,
some technical requirements of the ESS, including chargingdischarging duration range, power and energy density classes, can
be determined. It is observed that the dominant power components concentrate over the low frequency band, especially in the
dc and day cycle [46].
Although this observation is obtained from the local data, it has
a general implication, and similar results are derived in [11]. The
high frequency uctuations (above 1 Hz) are insignicant which

can be absorbed by the turbine generator inertia [47]. In [48], the

PSD derived from wind speed measurements from Changi, Singapore, is analyzed. For maximizing the energy capture, the low frequency band is taken into account. The selected storage medium is
required to have the capability to fully chargedischarge in hours
and have high specic density. Combined with other technical
and economical consideration, the NaS is selected for the daily dispatch purpose.
The rated capacity of modern wind farms can reach to several
hundred MWs. For the energy management purpose, large-scale
storage medium should be applied, such as BESS, PHS and CAES
[49]. Since the PHS and CAES are limited by topographical constraints, the BESS is considered as a more competitive option for
large-scale ESS application due to high power and energy density,
scalability, fast response, simple maintenance requirement and
high cycle life for both technical and economical consideration
[48,50].
For the power quality improvement purpose, the high frequency band of the PSD should be studied. The response time
and ramp rate capability are the main concerns for this purpose.
The storage mediums with fast response and small energy capacity, such as FES, super-capacitor, SMES, are potential options.
Since no single storage technology can provide the benets of
both high power density and energy density, a hybrid ESS with
the combination of multilevel storage devices can be used to deliver desired power time series [51]. The high energy density storage
medium, normally BESS, is adopted for low frequency uctuation
mitigation [31,52,53], and the high power density storage mediums, which can be super-capacitor [31], SMES [53], FES [52], are
used for smoothing high-frequency uctuations.
4.2. ESS sizing
Once the ESS type is determined, the optimal sizing problem
can be solved by balancing the benets and cost [5,54]. The sizing
problem includes the determination of both the power rating and
the energy rating. The methodologies of the ESS sizing for wind
power integration support are summarized and categorized in this
section.
Several methodologies for the BESS sizing for short-term (daily)
wind power dispatch are developed. In [55], through the spectral
analysis of wind and solar resources combined with daily load proles, a mathematical model is developed for a stand-alone ESS
where the storage is calculated for different levels of mean load.
The ESS is designed according to the worst case scenario (worst
day of the year) to obtain the ESS size. A new method has been
developed to obtain a probabilistic density function of wind power
forecast error in [56]. Based on the probabilistic density function,
the ESS nominal power can be determined. As the follow-up work,
the ESS capacity estimation was completed in [57]. A probabilistic
methodology is presented which obtains the ESS size as a function
of the un-served energy. The ESS sizing aims to compensate forecast errors up to a certain level.
The ESS sizing problem can also be modeled as an optimization
problem. The factors related to the sizing of the BESS, including
wind power forecast accuracy (error in range between 10% and
30% [58]), application purpose, control strategy and economical
aspect, can be quantied and taken into the optimization formulation as constraints. In [59], the cost function is formulated as the
power deviation from the next hourly average. The forecast error
is considered to be within 10%. Based on the State Of Charge
(SOC), current constraints and chargingdischarging rules, the
power reference of next hour can be generated to minimize the
cost. The results show 1525% of the rated wind farm capacity is
a proper ESS size to have an effective hourly dispatch. The same
cost function is used in [8]. The hour-long wind power production

H. Zhao et al. / Applied Energy 137 (2015) 545553

scenarios are rstly generated by a probabilistic forecast of the

hourly average wind power and the inverse of the joint cumulative
distribution. From these scenarios, the ESS characteristics are
determined as a function of the desired risk parameter.
The cost function can be formed as the economic benet
obtained from the dispatched power of wind farms against the
ESS cost. In [60], the optimal dispatched power level from wind
farms is obtained by solving the optimization problem based on
the given wind power prole. The DC voltage is another constraint
to be considered. The corresponding BESS power and energy ratings can then be determined. With the same procedure [61], aims
to minimize the cost of the BESS. Additionally, the ESS output is
required to meet the output forecast of a wind farm within 4%,
90% of the time. Besides the economic benets from wind power
dispatch and the ESS cost, a new benet component related to
the voltage stability is introduced into the cost function in [5].
Using the exhaustive search of all possible solution candidates,
the method is capable of nding a unique optimum of the ESS sizing. A new planning tool for medium voltage (MV) distribution networks is proposed in [35] based on Genetic Algorithm (GA) and
Dynamic Programming (DP). It is able to decide the optimal rating
of energy storage plants that minimize the overall network cost.
The cost function can also be formed as the service life of the
energy storage system. In [46], the historical long-term wind speed
data is assumed to be known. A new dispatch strategy is proposed
to ensure the BESS goes through full chargingdischarging cycle
and thus maximizes the energy storage potential of the BESS.
For smoothing the wind power uctuation, 1 year operation
data are analyzed in [28], and it is concluded to choose 20% rated
power of a given wind farm as rated power and 4 h chargingdischarging time for the ESS. For other applications [21], mentions
that according to the accumulated experiences on damping controller design during the past several years, approximate 25% of
the system capacity is a general selection.
4.3. Energy storage siting
As discussed before, some large-scale ESS types, such as the PHS
and the CAES, are heavily dependent on topographical conditions.
Other types can be installed not only on-site with wind farms,
but also at different locations in the power system for different
purposes [62], like deferring or avoiding transmission capacity
upgrade, reducing transmission and distribution losses, and
improving system stability [33]. Currently, only a few publications
have addressed the optimal placement of the ESS in a power system with large-scale wind integration.
For the on-site installation of the ESS with wind farms, the ESS
can either be placed at Point of Common Coupling (PCC) or
equipped with WTGs. The former conguration is adopted by the
most hybrid wind farm-ESS. The latter conguration is introduced
in [30]. Each wind turbine is equipped with a super-capacitor connected to the DC link of back to back converter. This scheme
enhances the control exibility of each wind turbine. However,
due to the smoothing effect of power uctuations of distributed
wind farms over a large geographical area, the requirement of
power and energy rating is high and not cost-effective.
In [3], a staged procedure is introduced to seek the minimum
number of storage nodes and total network storage that can still
mitigate the effects of renewable energy uctuations on network
constraints. The interesting result shows that instead of wind
power injection nodes, the other nodes at the end or the middle
of crucial transmission lines have higher degree of control over
congestion. The optimal siting problem is also investigated in
[63,64]. The optimal storage location which allows highest wind
power penetration is selected based on the calculation for different
location scenarios. The simulation results demonstrate that the

551

optimal storage distribution can effectively utilize transmission

capacity and eliminate the need for transmission expansion.
5. Energy storage system operation and control
5.1. Operation of ESS for wind power dispatch
The role of the ESS in the wind power trading in the modern
power market is analyzed in [65]. A properly designed ESS is
proved to bring additional economic benets. If the wind power
output can be scheduled in a manner similar to that of a conventional power plant, the prospect of the wind power will be much
improved as the optimal economic dispatch can then be achieved
[46].
The operation modes of wind farms with the ESS participating
in modern power market are described in [46,66,48,66] Wind
power for next few hours are estimated based on the wind speed
forecast. The power dispatch schedule of wind farms has to be submitted in advance to the grid operator. The grid operator adjusts
the system operating state to achieve the economic dispatch and
power ow regulation. The schedule beyond the certain hours is
allowed to be adjusted to compensate the forecast errors. Once
the power output of the wind farm differs from the schedule submitted, the wind farm operator will be nancially punished. Therefore, a lot of research concentrates on the short-term (daily)
optimal scheduling scheme design for wind farms by taking advantage of the exible chargingdischarging ability of the ESS.
In [32], the expected power dispatch is estimated based on the
hourly average forecast wind power without consideration of the
economic benet. In [67,68], the generation scheduling is based
on the forecast of electricity prices, load and wind generation. It
aims to maximize the expected prot from participating in the
day-ahead power market. In [58], the wind farm operator is
assumed to participant in the Nord Pool Spot market. The wind
forecasting error is considered as a stochastic variable. The ESS will
be hourly adjusted based on stochastic programming, seeking the
maximum benet of the wind farm and the ESS. Ref. [67] introduces a dynamic algorithm applied for daily scheduling in a power
market. The objective of the online operation strategy is to follow a
given generation schedule as close as possible. A statistical
approach is proposed in [46] to determine the power output of a
hybrid wind-BESS system. The developed new dispatch strategy
ensures the BESS goes through full chargedischarge cycle to maximize the energy storage potential. Ref. [17] proposes an hourly
discretized optimization algorithm to identify the optimal daily
operation strategy to be followed by the wind turbines and the
PHS, assuming that the wind power is perfectly forecasted.
The aforementioned are the use of single-unit ESS. In [69], a
dual BESS (one in service and one in standby) is proposed which
allows more exibility in terms of the design of the dispatch pattern. The dispatched power from the WTG-BESS power station
can be treated as a rm commitment. It can be determined based
on the long-term historical wind power data, the likelihood of
the success of the power delivery and the BESS capacity. With a
similar dual-ESS conguration [48], develops an optimal switchover dispatch scheme for the dual-BESS, by considering the wind
speed forecast and the chargingdischarging characteristics of
the installed BESS. The primary operation objective of the WTGBES scheme is to harness the energy available in the wind as much
as possible.
5.2. Operation of the ESS for uctuation mitigation
Wind power ltering algorithms are widely applied by means of
fast and short-term ESSs to mitigate the output uctuation of wind

552

H. Zhao et al. / Applied Energy 137 (2015) 545553

farms. The block diagram of a simplied ltering control of a

hybrid WF-ESS is shown in Fig. 4. Through the Low-Pass Filter
(LPF) with time constant Tf, the high-frequency part Pexp
out of wind
power Pwind is ltered as charging or discharging command of
the ESS. Since the ESS has a control lag, losses and an output limiter, the actual output does not match the target output. However,
due to its simplicity and fast computation speed, it is easy to be
implemented and suitable for application in real-time operation.
As proposed in [23,44], the power output reference of the ESS
Pord
bess can be determined by,

Pord
bess

sT f
P wind
1 sT f

While the Remaining Energy Level (REL) of the ESS can be calculated as,

REL

P ord
Tf
bess

Pwind
s
1 sT f

Eq. (2) shows that the larger time constant Tf has higher
smoothing effect but also results in a larger ESS power and capacity. It is very important to select proper Tf for the controller design.
Normally, Tf is decided by the local wind prole and the ESS capacity, and is kept constant during operation [23,43,46,70,44] proposes a exible LPF with an optimal Tf. Tf is regulated by the
Particle Swarm Optimization (PSO) algorithm in real-time. The
maximum uctuation of the combined power in any 1 min and
30 min time window is kept within predened constraints.
With large output variations of wind farms, it is easy to reach
the capacity limits of the ESS by the mere smoothing method without consideration of the charging level. Therefore, the SOC is introduced as a feed-back signal to keep the charging level within its
proper range [23,44]. In [46], an additional coefcient k is introduced to protect the ESS from over-limiting, as shown in (3). k is
changed according to the predened curve based on SOC.

Pord
bess k

sT f
P wind
1 sT f

Due to the measurement delay, the phase of the measured

power signal is shifted. To compensate this lag, an additional
lead-lag compensator is introduced and used for LPF [43].
5.3. Operation of the hybrid ESS
The previous subsections describe the operation of a single type
ESS. Since the single type storage technology can hardly meet the
requirement of both fast response and large energy capacity [7],
the logical solution is a hybrid ESS system, which combines multiple storage devices into several levels that can be used to deliver
the required power. In [51], a knowledge-based ANN control with
a washout-lter is used for the two-level storage for wind power
dispatch.
For the grid with many installed ESS dispersed in a large area,
the integration of these ESSs will have much better capability compared with the individual ESS. By gathering them into a virtual
assembly and control centrally, they can be used for many utility
applications, such as frequency regulation, load leveling and control of transmission power ow. A concrete example of the aggre-

Fig. 4. Basic principle of using LPF algorithm for power uctuation mitigation.

gated energy storage system using the BESS is proposed in Japan,

called Battery SCADA [10] and the demonstration of this technology started in 2012 in Yokohama, Japan.
6. Conclusions
The ESS is considered as an effective solution to handle the reliability and stability challenges of future power systems with large
scale wind power integration. Various ESSs with different technical
features are available in the market.
The ESSs can be used for different applications required by specic wind farms, grid operators or consumers. For the generationside, it can aim to improve the grid-friendliness of wind farms to
dispatch wind energy such that they could be controlled like conventional power plants. For the grid-side roles of the ESS, it can
provide ancillary services to mitigate variability and uncertainty
of the entire grid. For the demand-side roles, the aggregated EVPP
can fulll the requirements of both vehicle owners and grid
operators.
For the ESS planning, it is important to properly select the ESS
type, and determine the size and site of the ESS. The size of the
ESS, including both power and energy capacity, can be determined
by several methodologies, including the method of using historical
wind proles, the probabilistic method based on wind power forecast error, etc. The sizing problem can be formulated as an optimization problem with different cost functions. The siting of the ESS
without topographical limitation can be installed either on-site or
other locations to achieve high controllability. It shows the nodes
at the end or the middle of crucial transmission lines have higher
impact on the congestion management.
The recent research of the ESS operation and control focuses on
the daily dispatch scheme of the ESS with wind farms and uctuation mitigation. Different factors, including wind power forecast
error, technical constraints, market rules, and energy price, are
taken into consideration to determine the optimal operation strategy for single or multiple ESS. For the output smoothing purpose,
the high frequency component is ltered and compensated by
the ESS. The time constant is the key factor to strike a balance
among the smoothing effect, the required ESS power and capacity
rating.
Acknowledgement
The authors would thank Sino-Danish Center for Education and
Research (SDC) for the nancial support to the PhD project Coordinate control of Wind Power Plant and Energy Storage System.
References
[1] IEA. In: World energy outlook 2011. Paris: IEA Publications; 2011.
[2] Wang P, Gao Z, Bertling L. Operational adequacy studies of power systems with
wind farms and energy storages. IEEE Trans Power Syst 2012;27(4):237784.
[3] Backhaus SN, Chertkov M, Dvijotham K. Operations-based planning for
placement and sizing of energy storage in a grid with a high penetration of
renewable. Report no. LA-UR-11-03619. Los Alamos National Laboratory;
2011.
[4] Sun YZ, Zhang ZS, Li GJ, Lin J. Review on frequency control of power systems
with wind power penetration. In: International conference on power system
technology; 2010.
[5] Le HT, Santoso S, Nguyen TQ. Augmenting wind power penetration and grid
voltage stability limits using ESS: application design, sizing, and a case study.
IEEE Trans Power Syst 2012;27(1):16171.
[6] Kondoh J, Ishii I, Yamaguchi H, et al. Electrical energy storage systems for
energy network. Energy Convers Manage 2000;41(17):8631874.
[7] IEC. Grid integration of large-capacity renewable energy sources and use of
large-capacity electrical energy storage, Tech report; 2012.
[8] Boutsika T, Santoso S. Sizing an energy storage system to minimize wind
power imbalances from the hourly average. In: IEEE PES general meeting, San
Diego, California, USA, July 2226, 2012.
[9] Billinton R, Karki R, Gao Y, et al. Adequacy assessment considerations in wind
integrated power systems. IEEE Trans Power Syst 2012;27(4):2297305.

H. Zhao et al. / Applied Energy 137 (2015) 545553

[10] IEC. Electrical energy storage white paper. Tech report; 2011.
[11] Chen H, Cong TN, Yang W, et al. Progress in electrical energy storage system: a
critical review. Progr Nat Sci 2009;19(3):291312.
[12] Wang D, Ren C, Sivasubramaniam A, et al. Energy storage in datacenters: what,
where, and how much? ACM SIGMETRICS Perform Eval Rev
2012;40(1):18798.
[13] Swierczynski M, Teodorescu R, Rasmussen CN, et al. Overview of the energy
storage systems for wind power integration enhancement. In: Proceedings of
IEEE international symposium on industrial electronics, Bari, Italy, July 47,
2010.
[14] Energy storage for grid connected wind generation applications. EPRI-DOE
handbook supplement; 2004.
[15] Connolly D. An Investigation into the energy storage technologies available for
the integration of alternative generation techniques. Tech report; 2007.
[16] Zarakis D, Chalvatzis KJ, Baiocchi G, et al. Modeling of nancial incentives for
investments in energy storage systems that promote the large-scale
integration of wind energy. Appl Energy 2013;105:13854.
[17] Castronuovo ED, Lopes JAP, Member S. On the optimization of the daily
operation of a wind-hydro power plant. IEEE Trans Power Syst
2004;19(3):1599606.
[18] Dagdougui H, Minciardi R, Ouammi A, et al. A dynamic decision model for the
real-time control of hybrid renewable energy production systems. IEEE Syst J
2010;4(3):32333.
[19] Madlener R, Latz J. Economics of centralized and decentralized compressed air
energy storage for enhanced grid integration of wind power. Appl Energy
2013;101:299309.
[20] Van der Linden S. The commercial world of energy storage: a review of
operating facilities. In: 1st Annual conference of the energy storage council,
Houston, USA; 2003.
[21] Wang L, Yu JY, Chen YT. Dynamic stability improvement of an integrated
offshore wind and marine-current farm using a ywheel energy-storage
system. IET Renew Power Gener 2011;5(5):38796.
[22] Ribeiro PF, Johnson BK, Crow ML, et al. Energy storage systems for advanced
power applications. Proc IEEE 2001;89(12):174456.
[23] Yoshimoto K, Nanahara T, Koshimizu G. New control method for regulating
state-of-charge of a battery in hybrid wind power/battery energy storage
system. In: Power systems conference and exposition; 2006. p. 124451.
[24] Nomura S, Ohata Y, Hagita T, et al. Wind farms linked by SMES systems. IEEE
Trans Appl Supercond 2005;15(2):19514.
[25] Tam KS, Kumar P. Application of superconductive magnetic energy storage in
an asynchronous link between power systems. IEEE Trans Energy Convers
1990;5(3):43644.
[26] Chen SS, Wang L, Lee WJ, et al. Power ow control and damping enhancement
of a large wind farm using a superconducting magnetic energy storage unit.
IET Renew Power Gener 2009;3(1):2338.
[27] Wang Z, Yuwen B, Cheng M. Improvement of operating performance for the
wind farm with a novel CSC type wind turbine-SMES hybrid system. In: IEEE
international symposium on industrial electronics, Hangzhou, China; 2012.
[28] Liang L, Jianlin L, Dong H. An optimal energy storage capacity calculation
method for 100 MW wind farm. In: International conference on power system
technology, Hangzhou, China; 2010.
[29] Wang L, Chen SS, Lee WJ, et al. Dynamic stability enhancement and power ow
control of a hybrid wind and marine-current farm using SMES. IEEE Trans
Energy Convers 2009;24(3):62639.
[30] Qu L, Qiao W. Constant power control of DFIG wind turbines with
supercapacitor energy storage. IEEE Trans Ind Appl 2011;47(1):35967.
[31] Li W, Jos G, Blanger J. Real-time simulation of a wind turbine generator
coupled with a battery supercapacitor energy storage system. IEEE Trans Ind
Electron 2010;57(4):113745.
[32] Teleke S, Baran ME, Huang AQ, et al. Control strategies for battery energy
storage for wind farm dispatching. IEEE Trans Energy Convers
2009;24(3):72532.
[33] Tewari S, Mohan N. Value of NAS energy storage toward integrating wind:
results from the wind to battery project. IEEE Trans Power Syst
2012;28(1):53241.
[34] Korpas M. Distributed energy systems with wind power and energy storage.
PhD thesis; 2004.
[35] Celli G, Mocci S, Pilo F, et al. Optimal integration of energy storage in
distribution networks. In: IEEE PowerTech, Bucharest; 2009.
[36] Banham-Hall DD, Taylor GA, Smith CA, et al. Flow batteries for enhancing wind
power integration. IEEE Trans Power Syst 2012;27(3):16907.
[37] Carrillo C, Feijoo AE, Cidras J, et al. Power uctuations in an isolated wind
plant. IEEE Trans Energy Convers 2004;19(1):21721.
[38] Morren J, de Haan SWH, Kling WL, et al. Wind turbines emulating inertia and
supporting primary frequency control. IEEE Trans Power Syst
2006;21(1):20056.
[39] Moura PS, De Almeida AT. The role of demand-side management in the grid
integration of wind power. Appl Energy 2010;87(8):25818.
[40] Liu W, Hu W, Lund H, et al. Electric vehicles and large-scale integration of wind
power the case of Inner Mongolia in China. Appl Energy 2013;104:44556.
[41] Quinn C, Zimmerle D, Bradley TH. The effect of communication architecture on
the availability, reliability, and economics of plug-in hybrid electric vehicle-togrid ancillary services. J Power Sources 2010;195(5):15009.

553

[42] Binding C, Gantenbein D, Jansen B, et al. Electric vehicle eet integration in the
Danish EDISON project a virtual power plant on the island of Bornholm. In:
IEEE PES general meeting, Minneapolis, Minnesota, USA; 2010.
[43] Lubosny Z, Bialek JW. Supervisory control of a wind farm. IEEE Trans Power
Syst 2007;22(3):98594.
[44] Jiang Q, Wang H. Two-time-scale coordination control for a battery energy
storage system to mitigate wind power uctuations. IEEE Trans Energy
Convers 2013;28(1):5261.
[45] Barton JP, Ineld DG. Energy storage and its use with intermittent renewable
energy. IEEE Trans Energy Convers 2004;9(2):4418.
[46] Li Q, Choi SS, Yuan Y, et al. On the determination of battery energy storage
capacity and short-term power dispatch of a wind farm. IEEE Trans Sustain
Energy 2011;2(2):14858.
[47] Li W, Joos G, Abbey C. Wind power impact on system frequency deviation and
an ESS based power ltering algorithm solution. In: Power systems conference
and exposition, vol. 2; 2006. p. 207784.
[48] Yao DL, Choi SS, Tseng KJ, et al. Determination of short-term power dispatch
schedule for a wind farm incorporated with dual-battery energy storage
scheme. IEEE Trans Sustain Energy 2012;3(1):7484.
[49] Salgi G, Lund H. System behaviour of compressed-air energy-storage in
Denmark with a high penetration of renewable energy sources. Appl Energy
2008;85(4):1829.
[50] Baalbergen F, Bauer P, Ferreira JA. Energy storage and power management for
typical 4Q-load. IEEE Trans Ind Electron 2009;56(5):148598.
[51] Abbey C, Strunz K, Jos G. A knowledge-based approach for control of twolevel energy storage for wind energy systems. IEEE Trans Energy Convers
2009;24(2):53947.
[52] Lee H, Shin BY, Han S, et al. Compensation for the power uctuation of the
large scale wind farm using hybrid energy storage applications. IEEE Trans
Appl Supercond 2012;22(3):5701904.
[53] Ise T, Kita M, Taguchi A. A hybrid energy storage with a SMES and secondary
battery. IEEE Trans Appl Supercond 2005;15(2):19158.
[54] Succar S, Denkenberger DC, Williams RH. Optimization of specic rating for
wind turbine arrays coupled to compressed air energy storage. Appl Energy
2012;96:22234.
[55] Barton JP, Ineld DG. A probabilistic method for calculating the usefulness of a
store with nite energy capacity for smoothing electricity generation from
wind and solar power. J Power Sources 2006;162(2):9438.
[56] Bludszuweit H, Domnguez-Navarro JA, Llombart A. Statistical analysis of wind
power forecast error. IEEE Trans Power Syst 2008;23(3):98391.
[57] Bludszuweit H, Domnguez-Navarro JA. A probabilistic method for energy
storage sizing based on wind power forecast uncertainty. IEEE Trans Power
Syst 2011;26(3):16518.
[58] Yuan Y, Li Q, Wang W. Optimal operation strategy of energy storage unit in
wind power integration based on stochastic programming. IET Renew Power
Gener 2011;5(2):194201.
[59] Teleke S, Baran ME, Bhattacharya S, et al. Rule-based control of battery energy
storage for dispatching intermittent renewable sources. IEEE Trans Sustain
Energy 2010;1(3):11724.
[60] Wang XY, Mahinda Vilathgamuwa D, Choi SS. Determination of battery storage
capacity in energy buffer for wind farm. IEEE Trans Energy Convers
2008;23(3):86878.
[61] Brekken TKA, Yokochi A, Von Jouanne A, et al. Optimal energy storage sizing
and control for wind power applications. IEEE Trans Sustain Energy
2011;2(1):6977.
[62] Zhang Y, Zhu S, Chowdhury AA. Reliability modeling and control schemes of
composite energy storage and wind generation system with adequate
transmission upgrades. IEEE Trans Sustain Energy 2011;2(4):5206.
[63] Bose S, Gayme DF, Topcu U, et al. Optimal placement of energy storage in the
grid. In: IEEE 51st annual conference on decision and control (CDC); 2012. p.
560512.
[64] Ghofrani M, Arabali A, Etezadi-Amoli M, et al. A framework for optimal
placement of energy storage units within a power system with high wind
penetration. IEEE Trans Sustain Energy 2012;4(2):43442.
[65] Voller S, Al-Awaad AR, Verstege JF. Benets of energy storages for wind power
trading. In: IEEE international conference on sustainable energy technologies;
2008. p. 7026.
[66] Hu P, Karki R, Billinton R. Reliability evaluation of generating systems
containing wind power and energy storage. IET Gener Transm Distrib
2009;3(8):78391.
[67] Korpaas M, Holen AT, Hildrum R. Operation and sizing of energy storage for
wind power plants in a market system. Int J Electr Power Energy Syst
2003;25(8):599606.
[68] Bathurst GN, Strbac G. Value of combining energy storage and wind in shortterm energy and balancing markets. Electr Power Syst Res 2003;67(1):18.
[69] Yao DL, Choi SS, Tseng KJ, et al. A statistical approach to the design of a
dispatchable wind power-battery energy storage system. IEEE Trans Energy
Convers 2009;24(4):91625.
[70] Daz-Gonzlez F, Sumper A, Gomis-Bellmunt O, et al. Energy management of
ywheel-based energy storage device for wind power smoothing. Appl Energy
2013;110:20719.