Direct power control strategy based variable speed

pumped storage system for the reduction of the wind

power fluctuation impact on the grid stability

Bitew Girmaw Teshager, IEEE Student Member, Han Minxiao, Simiyu Patrobers, IEEE Student Member,Zmarrak Wali Khan,
Luu Kim Tuan, Faisal Mehmood Shah
State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources
North China Electric Power University
Beijing, China 102206
[email protected]

Abstract— Harnessing wind energy is the most rapidly growing system collapse at large. Study in [5] shows that the VSPS can
amongst the renewable energy sources. However, because of its improve the steady state operations and dynamic stability of the
intermittency nature, wind power results in unfavorable power system. Therefore, the pumped energy storage plant can
influences on power system operations and stability. The voltage play a key role in stabilizing the deviations of the grid frequency
flicker and grid frequency fluctuation are significant in this and voltage flickers caused by wind power fluctuation.
regard. To minimize the effect of wind power fluctuations on the
grid frequency and AC bus voltage, this paper presents direct Many works have been focused on the converter topology
power control strategy-based variable speed pumped storage development [6],[7],[8],[9],[10],[11] which brought an
system. Modeling of the system is undertaken based on phasor improvement of the VSPS system performance and on the
model technique. The case study is made by considering development of renewable energy resource penetration
conventional synchronous machines for hydropower and thermal level[12]. Additionally the control strategy have been
power plans in the grid and induction machines for the VSPS and intensively studied in DFIM based VSPS applications. The
wind turbines. The performance is analyzed using works have been engaged mainly on the vector control category
MATLAB/Simulink platform. The results show that the proposed including field oriented control (FOC) [9],[13],[14] and direct
VSPS control strategy achieves well-regulated and stable grid torque control (DTC) [8],[11],[13],[14],[5]. From this,
frequency and AC voltage in the power system. significant achievements of system performance improvement
have been made.
Keywords—Variable speed pumped storage system; wind power
fluctuation; direct power control; frequency control; AC voltage From the control strategy perspective, in reference [8] for
control; modeling and simulation example, the frequency regulator based on direct power control
(DPC) implemented in a full-size converter fed synchronous
I. INTRODUCTION machine was conducted. The DPC was also used in[11] for
Due to increasing environmental concerns, wind power application of DFIM based VSPS system with emphasis on
generation has undergone rapid growth. In 2013, the world wind converter topology study. Grid-integrated VSPS system is
power production capacity was 318GW and it is forecasted to presented in [5] to study the effect of faults on the grid frequency
reach 712 GW, 1480GW, 2089 GW and 2672 GW by 2020, stability and active power flow of the VSPS. A stator voltage
2030, 2040, and 2050 respectively in moderate scenario [1],[2]. FOC strategy has been presented in [9] for suppressing the effect
of a wind farm power fluctuations, but implemented in an H-
Wind energy production has many advantages; it is eco-
bridge cascaded multilevel converter for VSPS the application.
friendly, sustainable, space efficient, incredible domestic
On the other hand, power filtering algorithm solution control
potential, has low operational cost, revitalizing to rural
approach was proposed in [3] to regulate the deviations of the
economies, etc. However, the major drawback of wind power is
grid frequency caused by wind power fluctuations. However, a
its intermittency and stochastic in nature over time. As a result,
DPC strategy based VSPS system for the reduction of wind
significant wind power fluctuations can be observed.
power fluctuation impact on grid frequency and AC voltage
For a power system even with moderate wind power stability based on phasor technique hasn’t been well addressed
penetration, the fluctuations should be mitigated otherwise this in aforementioned papers and hence presented in this paper The
may lead to substantial deviations in the grid frequency [3], converter employed in the paper is the three-level back-to-back
voltage flicker at the grid buses[4], equipment damage and voltage sourced converter neutral point clamped (VSC NPC)

This work was supported by the State Key Laboratory of Smart Grid
Protection and Control, “111” Project (B08013) of China.
converter. The performance is validated and analyzed in smoothing out the variations of the grid frequency and voltage
MATLAB/Simulink platform. flicker.
II. THE WIND POWER AND ITS FLUCTUATION
The wind speed is stochastic in nature over time. As a result,
significant wind power fluctuations can be observed.
For a given power coefficient Cp, the mechanical power (Pm)
and torque (Tm) produced by the wind turbine generator rotor
can be defined as
Pm = 0.5 ρ AC p ( λ , β ) vw3 (1)

Pm (2)
Tm =
ωr
where ρ, vw, A, and ωr are respectively the air density, wind
speed, swept area and angular speed of the generator rotor.
Maximum mechanical power extraction is possibly made at
maximum value Cp and is a function of the pitch angle (β) and
tip-speed ratio (λ). The power coefficient expression Cp has been Fig. 1. Turbine power versus speed characteristics
shown to have a unique maximum, Cp-max , for λopt 8.1 and βopt=
0 [15]. The maximum coefficient of Cp i.e the Betz limit is 16/27.
But maximum value of Cp is determined in this paper. Based on
the modeling turbine characteristics of [16], a generic equation
(3) is used to model Cp(λ,β).
C p ( λ , β ) = C1 (C 2 / λi − C3 β − C4 ) e − C5 / λi + C6 λ
(3)
1 1 0.035
= −
λi λ + 0.08 β β3 +1
Particular optimal value of λ is defined (λopt=8.1). Thus, for
β = 00 and the coefficients of C1 = 0.5176, C2 = 116, C3 =
0.4, C4 = 5, C5 = 21 and C6 = 0.0068, the maximum value
of Cp which is Cpmax = 0.48 achieved.
Normalizing equation (1) in the pu system, we get
Pm − pu = C p − pu ( λ , β ) vw3 − pu (4)
Thus, implementing the above equations, the turbine power Fig. 2. Typical wind turbine characteristics; a) wind speed in
is determined and the result is shown in Fig.1 as power versus m/s, b) corresponding generated active power in MW
speed of the wind turbine. The electromagnetic torque (Tem) is
also expressed as III. MODELING OF THE SYSTEM
Tem = p ( Ψ ds iqs − Ψ qs ids ) (5) A. The proposed system description
The output active (Pg) and reactive power (Qg) of a wind
turbine supplied to the grid are computed by
Pg = 1.5( − vds ids + v qs iqs − v qr iqr − vdr idr )
(6)
Qg = 1.5( vqs ids + v ds iqs − v dr iqr − vqr idr )
The distributions of wind speed and direction change from
region to region and from season to season. According to the
statistics, the typical frequency of wind power fluctuations
record from minutes to days [9].
For the slow fluctuations which is not a noticeable problem
for example from day to day, the power grid has adequate ways
of controlling. However, it is difficult to regulate fast
fluctuations varying from minute to minute only by thermal
power and other traditional means. Hence, to solve such
problems, the VSPS is a possible choice for stabilizing and Fig. 3. Single-line diagram of the VSPS-wind farm-grid
integrated system
 The structure of the proposed system is illustrated with a d ωm 1
= (Tem − F ωm − Tm )
single-line diagram in Fig.3. The conventional synchronous dt 2H
generator are represented by SG1, SG2 and SG3 respectively. dθm (11)
= ωm
The VSPS based on DFIM system is incorporated. As shown in dt
Fig.4, the induction machine of the VSPS is configured in such where, Ψdqs Ψdqr, Vdqs idqs, Lls Llr , Vdqr idqr, Lm, Rs Rr, ω ωr, Tm
a way that the rotor is connected with the rotor side voltage Tem , H, F, s and p are respectively dq-axis stator and rotor fluxes,
sourced converter (VSC) NPCr and the stator is connected with dq-axis stator voltage and current, stator and rotor leakage
the grid. Whereas, the grid side VSC NPCg is connected with
reactance, dq-axis rotor voltage and current, magnetizing
the grid through the coupling inductor. Details are explained in
reactance, stator and rotor resistance, synchronous and rotor
[17]. IG represents an induction generator based wind farm.
angular speeds, turbine shaft and electromagnetic torque,
combined inertia constant, combined viscous friction
coefficient, slip ratio of the induction machine, and number of
B. The VSPS machine modeling
pole pairs.
Fig. 5 illustrates the block diagram of the DFIM fed VSPS
unit based on the VSC NPC converter topology. C. The VSC-NPC converter modeling for deploying VSPS
The VSC, a self-commutated converter, is fast and
controllable for AC/DC interface applications. One type of the
VSC converters is a three-level NPC power converter. As shown
in Fig.6, the three-level block contains three arms each having
their own four switching components with their antiparallel
diodes and two NPCs.

Fig. 4. Block diagram of the DFIM-fed VSPS based on VSC

NPC converter topology

Fig. 6. The structure of the three level VSC NPC converter

Fig. 5. Circuit diagram of the wound-rotor machine of Fig.4 in Since this paper is engaged to validate the performance of
dq-axis reference frame the proposed system in accordance with the phasor model
technique. The phasor approach is preferable for this study
Based on the circuit diagram of Fig.5, the dynamics model because the proposed system is large for detailed model
of the three phase induction machine is expressed in equations technique. Therefore, the following assumptions are taken: The
(7) to (11) including electrical and mechanical systems. Further power losses in the converters are neglected. The switching
details are explained in [17]. dynamics are also neglected because the pulse width modulation
frequency in three-level VSC NPC is much greater than the
d Ψ ds system frequency [18].
= Vds − Rs ids − ω s Ψ qs
dt
d Ψ qs Thus, the model of the VSC NPC is developed as in (12).
= Vqs − Rs iqs + ω s Ψ ds (7)
dt udqr = mdqr Vdc / 2
d Ψ dr udqg = mdqg Vdc / 2 (12)
= Vdr − Rr idr + (ωs − ωr ) Ψ qr
dt where udqr, udqg and mdqr, mdqg are converter injected voltages
d Ψ qr
= Vqr − Rr iqr − (ωs − ωr ) Ψ dr (8) and modulated signals respectively in the rotor and grid side
dt converters.
Tem = 1.5 p ( Ψ ds iqs − Ψ qs ids ) (9)
Ψ ds = Ls ids + Lm idr
The converter dynamics can be given by
Ψ qs = Ls iqs + Lm iqr dVdc 1 1
Ψ dr = Lr idr + Lm ids
= ic = (Pg − Pr )
dt C CVdc
Ψ qr = Lr iqr + Lm iqs
= (1 / CVdc )(u dg idg + uqg iqg − udr idr − uqr iqr ) (13)
Ls = Lls + Lm
Lr = Llr + Lm (10) Or in the Laplace transform,
 1 1 (14) I dqr ( s ) 1 (18)
Vdc ( s ) = Ic ( s ) = [Pg ( s ) − Pr (s)] = Gi ( s ) = *
sC sCVdc ( s ) I dqr − ref ( s ) L
s +1
Rr
where C and Vdc are capacitance and DC voltage respectively
Thus, the current control (inner loops) structure of Fig.8 is
and ic is the DC link capacitor charging current. Pg and Pr are
supposed to be developed. The PI control of Ki(s) is defined by
active power flow in the grid and rotor side of the converters
respectively. Hence, equation (14) can be illustrated in phasor Pi s + Ii
Ki ( s ) = (19)
model by Fig.7 [17]. s
where Pi and Ii are proportional and integral gains. Pi and Ii are
determined and tuned based on the control stability theory.
1
Vdc
1
C s

Fig. 7. DC-link dynamic phasor model

D. Control strategies for the VSC NPC system Fig. 8. The current control structure of the VSC NPC1
A direct power control (DPC) strategy is employed to have
fast response and lets the control system compensate the wind From equation (19), we obtain
power fluctuations directly and effectively. Even if this is a local I qr ( s ) = Gi ( s ) I qr − ref ( s ) (20)
control scheme that is placed in the VSPS unit, the significant
output that is an adjustable and nearly constant power flow can Multiplying both sides of (20) by 1.5VsLm/Ls, we get
be seen from the network of the power system as far as the VSPS
1.5Vs Lm / Ls I qr ( s ) = Gi ( s )1.5Vs Lm / Ls I qr − ref ( s )
operates within the specified limits. The power command Pop-vsps (21)
of the VSPS is determined from the capacity of the VSPS and Therefore, based on equation (18), we can deduce (23) as
its efficiency. follows
An external set point, Pset which is the input to the power Ps ( s ) = Gi ( s ) Ps − ref ( s ) (22)
control system, in this regard, should be assumed for the VSPS.
The set point depends on the VSPS capacity and the situation of That is the dynamic model for active power control of rotor
the power grid. The value of this set point is dedicated to side converter VSC NPCr is the same as the current control
optimize the system energy balance and update relatively with dynamic model. The PI control of Kp(s) of Fig.9 is defined in
slow rate. The main concern is however to control the wind (23)
power fluctuations and regulate the VSPS power for mitigating
Pp s + I p
the impact on the grid. K p (s) = (23)
s
Equations of active power Ps and reactive power Qs provide
the basis for control strategy modelling, and given as
Ps = 1.5(vqs iqs + vds ids )
Qs = 1.5(vqs ids − vds iqs ) . (15)
Based on the detail works in [17], we have
Ps = 1.5Vs ( Lm / Ls )iqr
Qs = 1.5Vs [( Ψ s / Ls ) − ( Lm / Ls )idr ] (16) Fig. 9. The active power control structure of the VSC NPC1
which implies the independent control of the real power and Considering the power command ΔPwind to the VSPS power
reactive power in the application of DFIM in VSPS. control unit, it is computed by
Starting with the voltage equations of (7) to (10) and taking Δ Pwind = Pwind − Pset (24)
the assumptions of a voltage drop across stator resistor being where Pwind is the wind farm power measured, and Pset an
very small compared to the grid voltage and a magnitude of assumed forecasted power output. This Pwind is added to the
stator flux fairly constant, we have the mathematical equations power command Pop_vsps to feed the fluctuated wind power to the
of (17). control loop for regulating. Thus, the above strategies are
diqr combined for direct power control scheme in the VSPS unit in
i + L*r
Vqr = Rr qr + (ωs − ωr )( Lr idr + Lm ids )
dt proposed system as shown in Fig. 10.
* didr
Vdr = Rr idr + Lr − (ωs − ωr )( Lr iqr + Lm iqs ) The overall control structure for the VSPS unit is illustrated
dt in Fig.11. The models and details of the reactive power control
* 2
where Lr = Lr − ( Lm / Ls )
(17) in both rotor side and grid side converters and DC voltage
From (17), the rotor dynamics model is determined and control in the grid side converter are presented in [17]. The grid
defined as icon of Fig.11 represents combinedly the conventional
synchronous machines SG1, SG2, SG3 and fixed loads L1, L2 of overshoot is possibly perceived during the instruction signal
Fig.1. switches from lower to higher value. The settling time is about
one second. Fig.13 also shows the response of the inner loop
signals when the power fluctuation instructions are imposed on
the outer loop system. The response shows perfect tracking of
the respective reference signals. Hence, the results imply that the
DPC strategy based VSPS system can adjust the grid frequency
and AC voltage fluctuations caused by wind energy in a power
grid integrated system quickly and flexibly.

Fig. 10. Direct power control method for VSPS system

Fig. 12. Tracking of power instructions on outer loop

structures in the DPC based VSPS system
0.2
d-axis current

Fig. 11. Overview of a DPC system for the VSC converter 0

grid-side

(pu)

topology of the VSPS system -0.2

-0.4

IV. SIMULATION MODEL -0.6

0 10 20 30 40 50 60 70 80 90 100
Time (seconds)
In this paper a case-study is undertaken involving a 300 MW 0.2
d-axis current

DFIM-based VSPS and a power grid system integrated with a 0

rotor-side

(pu)

wind-farm comprising of 7 identical 15 MW wind turbine -0.2

induction generators. To investigate the impact of wind -0.4

fluctuations on the grid frequency, a medium power system -0.6

0 10 20 30 40 50 60 70 80 90 100
Time (seconds)
model consisting of two 200MVA hydropower plants and one 0.6

15 MVA diesel power unit with conventional synchronous

q-axis current

0.4
grid-side

generators is established. The phasor model technique in

(pu)

0.2

MATLAB/Simulink is applied. For the conventional 0

synchronous machines, the primary voltage regulation method -0.2

0 10 20 30 40 50 60 70 80 90 100

in automatic voltage regulators based on the standard IEEE type Time (seconds)

I and the primary frequency regulation in turbine governors are

1.5
q-axis current

used.
rotor-side

1
(pu)

0.5

V. RESULTS and Discussion

0
0 10 20 30 40 50 60 70 80 90 100

A. VSPS grid-integrated network simulation and implications Time (seconds)

for wind power fluctuation impact reduction

Fig. 13. Tracking of reference signals on inner loop control
The control of power in the DFIM based VSPS system with structures in the DPC based VSPS system
wind farm is essentially the power control of the VSC fed VSPS
system. Fig. 12 shows the evolution of the proposed network B. Grid frequency and AC voltage response
depicted in Fig.3 but excluding the wind farm. As shown in Fig.
12, the proposed control system tracks the power fluctuation In this section, the proposed control strategy for Fig.3 is
instructions very well. A small distortion is observed while the tested and presented. The evolution of the network is shown in
instruction signals vary from one value to the other. A 20% Fig.14. A grid frequency response is observed with the
deviations of ±0.04Hz while the active power is regulated to IEEE PES Power Systems Conference and Exposition, PSCE 2006 -
track the wind power fluctuation. On the other hand, when the Proceedings, vol. 2, pp. 2077–2084, 2006.
reference reactive power Q_ref=0pu, the controller tracks its [4] M. A. Chowdhury, N. Hosseinzadeh, and W. X. Shen, “Smoothing wind
reference value, i.e. the VSPS turbine couldn’t generate reactive power fluctuations by fuzzy logic pitch angle controller,” Renewable
power. As a result, some AC voltage deviation is found, i.e. Energy, vol. 38, no. 1, pp. 224–233, 2012.
17.6 kV at steady state. To maintain the AC voltage at its [5] S. Xiaofang, B. G. Teshager, H. Minxiao, M. Zhaojun, X. Jianbing, and
nominal value of 18 kV, amount of reactive power is W. Xinzhu, “Stability and Control of a Grid Integrated DFIM Based
determined to be generated using the above equations. Thus, 22 Variable Speed Pumped Storage System,” in 2017 International
MVar is required to be generated and equivalently Electrical and Energy Conference (CIEEC 2017), 2017, pp. 175–181.
Q_ref=0.067pu is suppled at time t=25sec. Generally, as shown [6] A. Bocquel and J. Janning, “Analysis of a 300 MW variable speed drive
in Fig.14, the grid frequency and AC voltage dynamic for pump-storage plant applications,” 2005 European Conference on
responses are quick, quite stable and the deviations are in Power Electronics and Applications, pp. 1–10, 2005.
acceptable ranges. [7] S. Furuya, T. Taguchi, K. Kusunoki, T. Yanagisawa, T. Kageyama, and
T. Kanai, “Successful Achievement in a Variable Speed Pumped
Storage,” in Conference Record of the Power Conversion Conference,
1993, pp. 603–608.
[8] C. Xin, H. Minxiao, and Z. Chao, “Power Control Analysis for Variable
Speed Pumped Storage with Full-Size Converter,” in IECON2015-
Yokohama, 2015, pp. 1327–1332.
[9] O. H. Abdalla, M. Han, and C. Liu, “Multi-level Converter Based
Variable Speed Pump Storage for Wind Power Compensation,” in 2014
International Conference on Information Science, Electronics and
Electrical Engineering, 2014, no. 3, January, pp. 1497–1501.
[10] P. K. Steimer, O. Senturk, S. Aubert, and S. Linder, “Converter-fed
synchronous machine for pumped hydro storage plants,” in 2014 IEEE
Energy Conversion Congress and Exposition, ECCE 2014, 2014, pp.
4561–4567.
[11] M. V Pronin, O. B. Shonin, A. G. Vorontsov, and G. A. Gogolev,
“Features of a Drive System for Pump-Storage Plant Applications based
on the use of Double-Fed Induction Machine with a Multistage-
Multilevel Frequency Converter,” in 15th International Power
Electronics and Motion Control Conferenc, 2012, pp. 1–8.
[12] S. Rehman, L. M. Al-Hadhrami, and M. M. Alam, “Pumped hydro
energy storage system: A technological review,” Renewable and
Fig. 14. Evolution of the network; a) grid frequency response Sustainable Energy Reviews, vol. 44, pp. 586–598, 2015.
associated with control of the VSPS active power, b) AC [13] Y. Djeriri, A. Meroufel, A. Massoum, and Z. Boudjema, “A comparative
voltage response associated with control of the VSPS reactive study between field oriented control strategy and direct power control
power strategy for DFIG,” Journal of Electrical Engineering, vol. 14, pp. 169–
178, 2014.
VI. CONCLUSION [14] M. Merzoug and F. Naceri, “Comparison of field-oriented control and
direct torque control for permanent magnet synchronous motor
This paper presents a study on a DPC based VSPS system (pmsm),” International Journal of Electrical, Computer, Energetic,
for the reduction of wind power fluctuation impact on the grid Electronic and Communication Engineering, vol. 2, no. 9, pp. 299–304,
stability. Mathematical models based on phasor model technique 2008.
is presented. Case study is conducted. The performance is
[15] M. K. Bourdoulis, S. Member, and A. T. Alexandridis, “Direct Power
validated in MATLAB/Simulink platform. The results show that Control of DFIG Wind Systems Based on Nonlinear Modeling and
the proposed control strategy for the VSPS system achieves Analysis,” vol. 2, no. 4, pp. 764–775, 2014.
well-regulated grid frequency and AC voltage with small errors
[16] S. Heier, Grid Integration of Wind Energy: Onshore and Offshore
of acceptable ranges in the power grid system in spite of the
Conversion Systems, 3rd ed. John Wiley & Sons Ltd, 2014.
wind power fluctuations.
[17] B. Girmaw Teshager, M. Han, Z. Meng, X. Song, J. Xu, and P. Simiyu,
REFERENCES “Phasor Model Simulation of a Grid Integrated Variable Speed Pumped
Storage System,” IET The Journal of Engineering, no. November, 2017.
[1] Global Wind Energy Council, “Global Wind Energy Outlook 2014,”
2014. [18] T. Demiray, G. Andersson, and L. Busarello, “Evaluation study for the
Simulation of Power System Transients using Dynamic Phasor Models,”
[2] OECD/International Energy Agency and Energy Research Institute,
in 2008 IEEE/PES Transmission and Distribution Conference and
“China Wind Energy Development Roadmap 2050,” 2011.
Exposition, 2008, no. 2, pp. 1–6.
[3] W. Li, G. Joós, and C. Abbey, “Wind power impact on system frequency
deviation and an ESS based power filtering algorithm solution,” 2006