APPEEC2019 1570581102












Virtual Inertia Control Strategy for Battery Energy






Storage System in Wind Farm




Ruixu Liu, Zhen Wang* Huadong Xing


College of Electrical Engineering, Zhejiang University Inner Mongolia Power Research Institute




Hangzhou, China Hohhot, China
*

Email: [email protected]





Abstract—A large number of wind turbines connected to the wind power curtailment. Besides, considering the different
grid by power electronic devices have caused the reduction of operating conditions of several wind turbines, there will be

power system inertia, resulting in serious problems such as fre- mutual influence [11]. Recently, energy storage technology

quency stability degradation. This paper presents a virtual iner- has been vigorously grown. Some scholars develop coordinat-

tia control strategy for battery energy storage system (BESS) in ed control of wind farm and energy storage to emulate the

wind farm to emulate inertia behavior similar to synchronous outer characteristics of the SG [12]-[14]. Energy storage has

generator. The proposed control strategy adjusts the BESS out- the characteristics of dynamically absorbing and releasing

put power reference value by adding an additional virtual iner- energy, which can not only provide the virtual inertia for wind
tia control loop into the BESS basic control strategy, so that the

BESS can provide the virtual inertia for the grid while smooth-
farms, improving the ability of supporting grid frequency sta-

ing power fluctuations. In addition, dynamic allocation coeffi-
bility, but also make up for intermittent and volatility of wind

cients are introduced in order to readjust the proportion of en-
power, increasing the controllability of the wind farm output

ergy storage in terms of responding to fluctuations of the grid power. Thus, energy storage technology has a good applica-
tion prospect of contributing to the inertia support in wind

frequency and wind power. The effectiveness of the proposed
farms.

control strategy is verified by simulation results.

Index Terms—Energy storage, frequency response, power fluc-
For the purpose of providing the frequency support and

smoothing wind power fluctuations, this paper proposes a
tuations, virtual inertia, wind farm.

virtual inertia control strategy for the BESS in wind farm.
Firstly, this paper presents a BESS basic control strategy

I. INTRODUCTION smoothing wind power fluctuations. Secondly, by adding a

virtual inertia control loop to basic control strategy, the active
In order to promote the sustainable development of re-

sources and environment, renewable energy sources such as power reference is adjusted to provide the virtual inertia for

wind power have developed rapidly over the past decades. the wind power-battery energy storage system. Then, in order

Most wind turbines are connected to the grid through low- to dynamically readjust the proportion of energy storage in

inertia power electronic interfaces. Moreover, in the maximum terms of responding to frequency and power changes, this

power point tracking (MPPT) control mode, the wind power paper introduces allocation coefficients to the virtual inertia

systems is decoupled from the grid and loses inertia response control strategy. And the design of allocation coefficients is

behavior of the traditional synchronous generator (SG), which discussed. Finally, the effectiveness of the proposed control

causes it hardly responds to changes of the grid frequency and strategy are verified by simulation results.

poses a threat to the safe and stable operation of the grid [1]-
[3]. To this end, there is an urgent need for wind power sys- II. A WIND POWER-BATTERY ENERGY STORAGE SYSTEM


tems to actively participate in grid frequency regulation.

A. Overview of Wind Power-Battery Energy Storage System

In recent years, inspired by the operation of traditional Fig. 1 shows the structure of a wind power-battery energy

power systems, some scholars proposed virtual synchronous storage system. The system is mainly composed of the wind

generator (VSG) technology [4]. Through the reasonable con- power system, the battery energy storage system (BESS), the

trol of power electronic equipment such as the converter, the power conversion system (PCS) and the power grid. The
wind farm can have the virtual inertia which helps it equiva- BESS adopts a centralized configuration scheme, being in-


lent to a SG [5]. Utilizing the wind turbine’s own rotor to pro- stalled at the point of common coupling (PCC) of the wind

vide the virtual inertia is a method of emulating the SG [6]- farm, so that it does not affect the control of wind turbines.

[10]. However, the way of responding to the grid frequency by And it is controlled to charge/discharge through the PCS.

the wind turbine rotor's momentum is usually at the expense of





This work is jointly supported by the National Key R&D program of

China (No. 2017YFB0902000) and the Science and Technology Project of

Inner Mongolia Power Group (No.2014-42).








1June 07,2020 at 03:19:13 UTC from IEEE Xplore. Restrictions apply.

Authorized licensed use limited to: University of Exeter. Downloaded on
 eq are the current and voltage, respectively, which are ob-
Wind
BESS Farm tained by Park Transformation of the three phase current iabc
Pwind
and voltage eabc shown in Fig. 1.
+ +
PCS Grid ed
+ + Pbat udc eabc Pbat
DC DC ref idref K udref
C Pref K PI i KCPd  CId
ibat DC AC iabc  K PP 
bat
s
+ s
+ Z Liq
(a)
id ZL
ibat
PWM PWM eabc iabc ref
ibat K CI uref
uref ref
uabc  KCP 
abc-dq s
iq
DC-DC
(b)
ZL
Current Inner dq-abc
ed eq id iq Z Lid
Loop Control ref
u dc iq K CIq udref
ibat _ ref uq _ ref ud _ ref KCPq 
DC-AC udcref K idref s
DC-DC DC-AC Voltage Outer  KVP  VI
Power Outer Current Inner id _ ref Loop Control udc _ ref s eq
Loop Control Loop Control Qref (c) (d)
iq _ ref
Pref 1.5ed
Basic Control
Figure 2. Basic control diagram: (a) DC-DC power control; (b) DC-DC
Virtual Inertia current control; (c) DC-AC voltage control; (d) DC-AC澳current control.
Control

f0 Pref 0 Virtual Inertia Control The above control strategy can achieve the goal of smooth-

 ing wind power fluctuations. However, this control strategy
Pwind
Pcmd has little response to frequency disturbances in the grid and
the low inertia of the power electronic interface will have a
negative impact on the steady operation of the power system.
Figure 1. Overview of wind power-battery energy storage system.
Thus, in order to improve the stability of the system, the con-
trol strategy needs to be further improved.
B. Basic Control Strategy of Power Conversion System
The basic control block diagram of PCS is shown in Fig.澳2. III. VIRTUAL INERTIA CONTROL STRATEGY
PCS is the interface between the BESS and the grid, which is
used to realize bidirectional transmission of energy. In this A. Analysis of Virtual Inertia Control
paper, the PCS is composed of DC-DC and DC-AC converters The inertia of the traditional SG derives from the SG rotor.
to control the charging and discharging of BESS. Introducing the SG rotor motion equation into the PCS control
can emulate the inertia of SG and provide inertial support for
The DC-DC converter employs a non-isolated half bridge the system to improve stability.
DC-DC bidirectional converter, and is mainly responsible for
controlling the charging/discharging power of the BESS. And The motion equation of the SG rotor is shown as equation
it adopts the double loop control strategy of power outer loop (2).
and current inner loop in which K PP / K PI and KCP / K CI are
respectively the proportion coefficient and integral coefficient. dZ
The BESS output active power reference for smoothing wind  J Tm  Te  
dt
power fluctuations called BESS smoothing control reference
is the deviation of the given combined active power and the where J is the inertia moment of SG, Tm is the input me-
wind power actual output active power, shown as equation (1). chanical torque, Te is the output electromagnetic torque.
The mechanical torque Tm , the electromagnetic torque Te ,
 Pref 0 Pcmd  Pwind   the mechanical power Pm , and the electromagnetic power Pe
have the relationship shown as the following equation (3). For
where Pref 0 is the BESS smoothing control reference, Pcmd is ease of analysis and calculation, the system's nominal angular
the time-series power scheduling submitted to system operator, velocity Z0 is equivalent to the actual angular velocity Z .
Pwind is the wind power output active power based on the
short-term prediction and in this paper the wind power system
output without BESS is set as the above value. Originally, the ­ Pm Pm
°Tm |
BESS charge/discharge power reference Pref is equal to the ° Z Z0
BESS smoothing control reference Pref 0 in the basic control.  ®  
°T Pe Pe
|
The grid-side converter employs a PWM converter and °̄ e Z Z0
adopts a voltage-oriented control strategy with a decoupling
controller of current inner loop and voltage outer loop in Conventional synchronous generators use a frequency
which K CPd / K CPq , K CId / K CIq and KVP / KVI are respectively modulator to control the prime mover torque, which can re-
the proportion coefficient and integral coefficient. U dc is the spond to system frequency deviations by adjusting the output
DC voltage between the DC-DC converter and the DC-AC power. Referring to the principle of the frequency modulator,
converter. U dc _ ref is the DC voltage reference. id , iq , ed and the frequency control section is added to the PCS control, and

2June 07,2020 at 03:19:13 UTC from IEEE Xplore. Restrictions apply.

Authorized licensed use limited to: University of Exeter. Downloaded on
the droop control of the active power-frequency is adopted, as ­ x, |'f | >0.05 and x  [0,1]
shown in equation (4).  D ®  
¯1 , |'f | <0.05
 Pm  Pref 0 Dp ( f0  f )  
­1, |'f | >0.05
where Pm is the virtual mechanical power of the BESS, Pref 0  E ®  
¯ y , |'f | <0.05 and y  [0,1]
is the BESS smoothing control reference, Dp is the droop
coefficient, f is the real-time grid frequency; f 0 is the rated B. Virtual Inertia Control Strategy of BESS
grid frequency, f 0 50 Hz .
Based on the above analysis, the P-f virtual inertia control
Combining equation (2), (3) with (4) and considering the strategy is shown in Fig. 3. The virtual inertia control is main-
coupling between the conventional SG’s angular velocity and ly composed of the given active power calculation part, filter
the system frequency Z 2S f can yield the P-f virtual inertia ref
and limiter. T1 is the time constant in the filter. Pmax ref
and Pmin
control equation of the PCS, as shown in equation (5). are the upper and lower limit, respectively.

df f0
Pe Pref 0  Dp ( f 0  f )  (2S )2 Jf 0
dt
   f 1 
df  KD
=Pref 0  K D ( f 0  f )  K H 1  T1 s
dt Filter
where K H is equivalent to the inertia time constant of the SG, d 
K D is equivalent to the droop coefficient, Pe is the virtual KH 
dt
electromagnetic power of BESS濁
The structure of equations (5) can be used to emulate the E
output behavior of a SG, so that the BESS can respond to the ref
 Pmax
frequency change and provide virtual inertia for the grid while Pref 0 D 
ref
Pmin
Pref
smoothing power fluctuations.
Limiter
It can be seen that the first term of the equation (5) relates
Figure 3. P-f virtual inertia control block diagram.
to the effect of smoothing wind power fluctuations and the
latter two are closely related to the frequency response. On the
basis of not changing the equation (5) form, the proportional C. Performance Evaluation
coefficients D and E are added to it, called allocation coeffi- The choice of allocation coefficients is related to the effect
cients, which are used to allocate the proportion of the energy of suppressing power and frequency fluctuations. In order to
storage system in smoothing the wind power fluctuation and comprehensively quantify the effect of control strategies under
providing the frequency support to achieve equilibrium and different allocation coefficients, this paper proposes an evalua-
maximize the use of storage. The adjusted BESS output active tion index. Firstly, calculate the standard deviation of power
power reference Pref can be shown in equation (6). and frequency, shown as equation (10). Then, weight and sum
up the above standard deviations to obtain the comprehensive
evaluation index S , shown as equation (11).
df
 Pref D ˜ Pref 0  E ˜ [ K D ( f 0  f )  K H ] 
dt ­ t2
f (t )  f 0 2
Under the per-unit system setting the wind farm rated out- ° S1 ¦( )
° t t1 f0
put power (kW) and rated grid frequency as the base value, the  ®   
equation (6) can be described as equation (7). ° t2
P (t )  PPCC 2
° S2 ¦ ( PCC
PPCC
)
df* ¯ t t1

 Pref* D ˜ Pref* 0  E ˜ [ K D* f 0 (1  f* )  K H* f 0 ]  where f (t ) and PPCC (t ) are respectively the frequency and
dt
active power at PCC in real time, PPCC is the average active
where Pref* , Pref* 0 , K D* , f* and K H* are the corresponding per- power of PPCC within a certain time range.
unit values. The base value of active power (kW) is the rated
output power of wind power system.  S k ˜ S1  S2   
It is worth mentioning that the allocation coefficients are where k is the weight coefficient.
dynamically adjusted by the actual operation to reduce the
interaction between the frequency response and the power Considering that the main task of the virtual inertia control
response. According to the range of frequency deviation, this strategy is to contribute to the inertia support, reducing fre-
paper sets the values of D and E according to the following quency fluctuation, and secondly to help stabilize the power
fluctuations, the weight of the frequency response effect
rules, shown as equation (8) and (9).

3June 07,2020 at 03:19:13 UTC from IEEE Xplore. Restrictions apply.

Authorized licensed use limited to: University of Exeter. Downloaded on
should be relatively great for the overall evaluation of the con-
trol strategy.

IV. CASE STUDY

A. A Wind Power-Battery Energy Storage System
In order to verify the effectiveness of the proposed control
strategy, this paper uses the structure shown in Fig. 4 to build
a simulation model in MATLAB/Simulink platform. As
shown in Fig. 4, a wind farm connects the distribution grid
which is equivalent to a SG. The BESS is installed at the PCC
of the wind farm by PCS. Rated output power of the wind Figure 5. Wind speed signal.
turbine is 9MW. The capacity of the BESS is 1MW灄h. The
detailed parameters of the wind turbine and line can be found Based on the range of equation (7) and (8) , the trials take
in [15]. And other main simulation parameters are shown in different values with 0.1 as the interval for D and E respec-
Table I. tively to simulate the system and calculate the evaluation in-
dex S . Considering the weight and magnitude of S1 and S 2 ,
Wind Farm G the weight coefficient k in evaluation index is taken 100. The
575V / 25kV S result of S under different allocation coefficients is shown in
10.5MW Line
30km 25kV /110kV 110kV Fig. 6. It can be seen that at (0.1, 0.5, 9.4504), the value of S
BESS DC C
DC
47MW
is the smallest. So x 0.1 and y 0.5 are designed as the
DC AC
380V / 25kV
Load value of allocation coefficients.
10.5MW
PCC

Figure 4. Structure of simulated wind power-battery energy storage system.

TABLE I. MAIN SIMULATION PARAMETERS

Parameter Value
*
K D 6.9

Virtual K H* 21.9
Inertia Controller ref
( Pmin ref
, Pmax ) (MW灄h) (-1, 1)
T1 0.02
Figure 6. Evaluation index S under different allocation coefficients濁
DC-DC Power Controller K PP / K PI 0.01/0.05
DC-DC Current Controller K CP / K CI 0.5/0.001 C. Performance Verification under Wind Speed Fluctuation
DC-AC Voltage Controller KVP / KVI 220/200 and Load-step Change
K CPd / K CPq 500/800 The situation of wind speed fluctuations and load-step
DC-AC Current Controller change is consistent with part B. The power fluctuations main-
K CId / K CIq 450/750
ly come from the wind speed fluctuations, and the frequency
Nominal Voltage (V) 250 fluctuations mainly come from load-step change.
BESS
Initial SOC (%) 50
Fig. 7 and 8 show the frequency response and power re-
Pcmd (MW) 3.35 sponse at PCC bus respectively under different control strate-
Wind Farm
Qref (Var) 0 gies, including three cases: general control without BESS
(fnoBESS, PnoBESS), BESS virtual inertia control without alloca-
B. Design of Allocation Coefficients under Wind Speed tion coefficients (fVI, PVI) and BESS virtual inertia control
Fluctuations and Load-step Change with allocation coefficients (fVI(α,β), PVI(α,β)). The simulation
The design of allocation coefficients is discussed under result shows that the BESS adopting virtual inertia control
wind speed fluctuations and load-step change. At the initial strategy can effectively respond to disturbances caused by
moment, the system operates without load in PCC. At 5s, the load-step change and wind speed fluctuations. What’s more,
10MW load is suddenly connected to the system at PCC. At the virtual inertia control strategy with the allocation coeffi-
the same time, the wind speed to the wind farm has fluctua- cients has better performance in not only improving the fre-
tions shown as Fig. 5. The frequency and power at PCC is quency stability of the system, but also reducing wind power
monitored in real time. Considering that the system is dis- fluctuations.
turbed by wind speed or load after 4s, the performance of the
system after 4s is analyzed.

4June 07,2020 at 03:19:13 UTC from IEEE Xplore. Restrictions apply.

Authorized licensed use limited to: University of Exeter. Downloaded on
 wind farm. Combing the VSG technology with BESS basic
control strategy smoothing power fluctuations, the active
power reference of PCS control is adjusted to make it actively
participate in the grid frequency modulation. In addition, dy-
namic allocation coefficients are introduced to the virtual iner-
tia control strategy to achieve the dual effect of suppressing
wind power fluctuations and improving frequency stability.
And the design of allocation coefficients is discussed. The
effectiveness of the proposed control strategy is verified by
simulation under wind speed fluctuations and load-step
change. And the simulation results prove that by adopting the
proposed control, the wind power-battery energy storage sys-
Figure 7. Simulation results of wind speed fluctuation and load-step change: tem can respond better to power and frequency changes than
Frequency of PCC.
adopting general control strategy without BESS or BESS vir-
tual inertia control strategy without allocation coefficients.

REFERENCES
[1] J. Ekanayake and N. Jenkins, "Comparison of the response of doubly
fed and fixed-speed induction generator wind turbines to changes in
network frequency," IEEE Trans. Energy Convers., vol. 19, no. 4, pp.
800-802, 2004.
[2] A. B. T. Attya and T. Hartkopf, "Control and quantification of kinetic
energy released by wind farms during power system frequency drops,"
IET Renew. Power Gener., vol. 7, no. 3, pp. 210-224, 2013.
[3] R. Cardenas, R. Pena, S. Alepuz, and G. Asher, "Overview of Control
Systems for the Operation of DFIGs in Wind Energy Applications,"
IEEE Trans. Ind. Electron., vol. 60, no. 7, pp. 2776-2798, 2013.
Figure 8. Simulation results of wind speed fluctuation and load-step change: [4] H.-P. Beck and R. Hesse, "Virtual synchronous machine," in 2007 9th
Active power of PCC. International Conference on Electrical Power Quality and Utilisation,
2007, pp. 1-6.
[5] J. Liu, Y. Miura, and T. Ise, "Comparison of dynamic characteristics
D. Performance Verification under Load-step Change between virtual synchronous generator and droop control in inverter-
The case of less frequency fluctuations is discussed in this based distributed generators," IEEE Trans. Power Electron., vol. 31, no.
5, pp. 3600-3611, 2015.
part. The wind speed is constant at 13 m/s. At 5s, the 5MW
[6] J. Morren, S. W. De Haan, W. L. Kling, and J. Ferreira, "Wind turbines
load is suddenly connected to the system at PCC. emulating inertia and supporting primary frequency control," IEEE
Trans. Power Syst., vol. 21, no. 1, pp. 433-434, 2006.
Fig. 9 shows the frequency response at PCC bus under the [7] G. Lalor, A. Mullane, and M. O'Malley, "Frequency control and wind
three control strategies. The simulation results show that even turbine technologies," IEEE Trans. Power Syst., vol. 20, no. 4, pp.
with small frequency fluctuations, the virtual inertia control 1905-1913, 2005.
strategy added the allocation coefficients has better perfor- [8] X. Zhu, Y. Wang, L. Xu, X. Zhang, and H. Li, "Virtual inertia control of
mance in contribution to frequency response. DFIG-based wind turbines for dynamic grid frequency support," in IET
Conference on Renewable Power Generation (RPG 2011), 2011, pp. 1-
6.
[9] Q. Shi, G. Wang, L. Fu, L. Yuan, and H. Huang, "State-space averaging
model of wind turbine with PMSG and its virtual inertia control," in
Conference of the IEEE Industrial Electronics Society, 2014.
[10] S. Wang, J. Hu, and X. Yuan, "DFIG-based wind turbines with virtual
synchronous control: Inertia support in weak grid," in Power & Energy
Society General Meeting, 2015.
[11] L.-R. Chang-Chien and Y.-C. Yin, "Strategies for operating wind power
in a similar manner of conventional power plant," IEEE Trans. Energy
Convers., vol. 24, no. 4, pp. 926-934, 2009.
[12] K. Shi, H. Ye, W. Song, and G. Zhou, "Virtual inertia control strategy in
microgrid based on virtual synchronous generator technology," IEEE
Access, vol. 6, pp. 27949-27957, 2018.
[13] Y. Ma, W. Cao, L. Yang, F. F. Wang, and L. M. Tolbert, "Virtual
synchronous generator control of full converter wind turbines with
Figure 9. Simulation results of load-step change: Active power of PCC. short-term energy storage," IEEE Trans. Industrial Electron., vol. 64,
no. 11, pp. 8821-8831, 2017.
[14] T. Yuan et al., "Virtual inertia adaptive control of a doubly fed
V. CONCLUSION induction generator (DFIG) wind power system with hydrogen energy
Large-scale wind power system-grid connection reduces storage," Energies, vol. 11, no. 4, p. 904, 2018.
the regulation capacity and the inertia of the grid, to a certain [15] The MathWorks. Wind Farm - DFIG Average Model. Available:
https://www.mathworks.com/help/physmod/sps/examples/wind-farm-
extent, and poses a threat to the safe and stable operation of dfig-average-model.html
the power system. In this paper, a virtual inertia control strate-
gy with allocation coefficients is proposed for BESS in the

5June 07,2020 at 03:19:13 UTC from IEEE Xplore. Restrictions apply.

Authorized licensed use limited to: University of Exeter. Downloaded on