Impact of SCIG, DFIG Wind Power Plant on IEEE 14

Bus System with Small Signal Stability Assessment

D.Rakesh Chandra, M. Sailaja Kumari, M. Sydulu F. Grimaccia, M. Mussetta, S. Leva, M.Q. Duong
National Institute of Technology Politecnico di Milano – Dept. of Energy
Warangal, India Via La Masa 34, 20156 Milano, Italy
rksh.chndr@gmail.com francesco.grimaccia@polimi.it, marco.mussetta@polimi.it

Abstract— In power systems voltage stability is a key issue During last decade, the average annual growth rate of
which attracts worldwide attention. This research presents an wind turbine installation is around 30 %. Ones of the major
implementation of a modified IEEE 14 bus system model in problems connected to solar and wind systems is their
Power System Analysis Toolbox (PSAT) – free and open source dependence on the weather conditions [2]. Thus, compared to
software. A newly developed Squirrel Cage induction generator
conventional generation, the stability of already existing grid
(SCIG) and Doubly Fed Induction Generator (DFIG) wind
turbine model are modeled and connected to a modified IEEE 14 is becoming important. Therefore, wind plants should behave
bus system. This paper investigates the impact of Squirrel Cage responsibly. For example, the important point during last
induction generator (SCIG) and Doubly Fed Induction several years is the continued grid-connection of wind turbine
Generator (DFIG) on the power system stability. Here considered at definite grid-voltage disturbance levels, to avoid voltage
wind generators are SCIG which is fixed speed and DFIG is drops and sectional energy deficits when wind energy units are
variable speed. Small signal stability study has been conducted disconnected.
on a modified IEEE 14 bus system with SCIG, DFIG wind Wind power plants must provide the power quality
turbine systems and their simulation results have been analyzed required ensuring the reliability of the power system where it
in this paper.
is connected to and to fulfill the clients connected to the same
Keywords— Small Signal Stability; Wind turbine; SCIG; grid. It is very important to understand the sources of
DFIG; STATCOM; PSAT disturbances that affect the power quality [4-6]. Currently
fixed speed and variable speed induction generators are using
I. INTRODUCTION for wind power generation. Integration them into grid arise
issues like voltage stability, transient stability problem. Power
The development of the installed capacity of grid system stability is depending on parameters belongs to
connected renewable energy source is continuously growing turbines, generators, governors. They affect both small signal
as a result of the environmental concerns in order to minimize stability and transient stability. There are several studies
the impact of conventional electricity generation [1]. Wind carried out recently for identifying required network
power is the world's fastest growing renewable source as reinforcement, reserve requirements and the impact of wind
shown in Figure 1. power on power system stability [7]. These studies are dealing
with different problems related to wind plant, such as
fluctuating nature of wind energy, location of wind
generations, various generator technologies and control. The
results are generally representing a super position of various
wind power aspects and predict required grid reinforcements,
reserve requirements, the impact on grid stability.
Reference [8] investigates the modeling and the transient
stability analysis of the wind integrated IEEE 14 test bus
system. The aim of the investigation is the enhancement of
transient stability using central area controller in a wind
integrated power system with storage. In [4, 9] a comparison
is made among 3 main type of wind turbines such as constant
speed wind turbine (CSWT), Doubly Fed Induction Generator
(DFIG), Direct Drive Synchronous Generator (DDSG) and
their steady and transient characteristics were analyzed and
simulated, respectively. Nordic grid model implemented using
Power System Analysis Toolbox was also validated through
Fig. 1. Growing wind turbine capacity time domain simulation by applying small and large
disturbances [10].

978-1-4799-5141-3/14/$31.00 ©2014 IEEE

 This research investigates small signal stability of two ⎡ ∂g 1 ∂g1 ⎤ ⎡ ∂g 1 ∂g1 ⎤
different wind generator configurations when connected to ⎢ ∂x ... ⎥ ⎢ ...
modify IEEE 14 bus system. This paper also introduced ∂x n ∂u ∂u r ⎥
⎢ 1 ⎥ ⎢ 1 ⎥
mathematical modeling of these wind generators in section II. C = ⎢ ... ... ... ⎥ D = ⎢ ... ... ... ⎥
SCIG, DFIG based IEEE modified 14 bus systems are ⎢ ∂g m ... ∂g m ⎥ ⎢ ∂g m ... ∂g m ⎥
simulated using PSAT are shown in III. All these simulation ⎢ ∂x1 ∂x n ⎥ ⎢ ∂u1
⎣ ∂u r ⎥⎦
work has done in PSAT software version 2.1.8. After ⎣ ⎦
simulations all results are compared and analyzed in section Where A , B , C and D are state matrix, input matrix,
IV, concluded in V. output matrix and coefficient matrix respectively. In order to
II. SMALL SIGNAL STABILITY find linearized state equation in frequency domain these
matrices have to be transformed using Laplace transform.
Finally, Eigen values will obtain after below equation has
In this section small signal stability of the system has been satisfied [11].
analyzed with S domain analysis. It is defined as the ability of
a power system to maintain the synchronism after subjected to det(sI − A) = 0 (4)
a small disturbance. In other words the system’s response to a A. SCIG Wind turbine modeling
small disturbance in power system state variables.
PSAT based Eigen value analysis has been done to SCIG equations are developed with respect to real (r ) and
determine small signal stability of the system. Here correlation imaginary (m) axis in terms of network reference angle. In
between Eigen values and power system dynamics is synchronous rotating frame of reference, the relation between
discussed. For this, a liberalized model is developed to find the stator machine voltage and the network is
out the resemblance between Eigen values, steady state matrix
and time domain simulation. State space equation and output vr = V sin(− θ )
equation is given by v m = V cos(θ ) (5)
X (t ) = f ( x, u, t ) (1) And the absorbed power is given by
y (t ) = g ( x, u, t ) (2) P = v r i r + v m i qm
Where equation 1 has all state variables such as ‘ u ’ is the
input variables, ‘ t ’ is the time and ‘ y ’is the output function.
(
Q = v m ir − v r i m + bc v r2 + v m2
(6)
)
The linearization of above 2 equations will help to study the Where bc is the fixed capacitor conductance. The voltage
response to small variations. In order to obtain this, differential equations, which were behind the stator resistance
polynomial equations are developed using Taylor’s series rs is given by:
formula where higher order terms are neglected. Linear
combination of system is presented as er' − v r = rS i r − x ' im
X = Ax + Bu (3) em' − v m = rS im + x ' ir
(7)
y = Cx + Du The relation between currents, voltage and state variables
Here A, B, C and D are obtained from the Jacobean matrix is as follows:
er' = Ω b (1 − ω m )em − (er − (x0 − x )im ) / T0
which contains partial derivates of the functions in terms of ‘ ' 1 ' '
f ’ and ‘ g ’ respectively to the input variable ‘ u ’ and the
state variable X . em' = −Ω b (1 − ω m )er' − (e1m − (x0 − x ' )ir ) / T0' (8)
Here x0 , x ' and T0 are calculated from the generator
⎡ ∂f 1 ∂f 1 ⎤ ⎡ ∂f1 ∂f n ⎤ parameters.
⎢ ∂x ... ...
∂x n ⎥ ⎢ ∂u ∂u r ⎥ x0 = x s + x m
⎢ 1 ⎥ ⎢ 1 ⎥
A = ⎢ ... ... ... ⎥ B = ⎢ ... ... ... ⎥ x x
⎢ ∂f n ...
∂f n ⎥ ⎢ ∂f n ...
∂f n ⎥ x ' = xS + R m
⎢ ∂x1 ∂x n ⎥ ⎢ ∂u1 ∂u r ⎥
xR + xm (9)
⎣ ⎦ ⎣ ⎦
xR + xm
T0' =
Ω b rR
The differential equations in terms of mechanical
parameters such as rotor inertia H m , turbine inertia H wr and
shaft stiffness K s is given by
 ω wr = (Twr − K s γ ) / (2 H wr ) Rotor side powers are

ω m = (K s γ − Te ) / (2 H m ) (10) Pr = vdr idr + vqr iqr

γ = Ω b (ω wr − ω m ) Qr = vqr idr − vdr iqr (18)
Where Te (electrical torque) is given by
For loss less system final injected power is given as
Te = e r' i r + e m' i m (11) P = v ds ids + v qs iqs + v dr idr + v qr iqr (19)
The mechanical torque is:
Q = v qs i ds − vds iqs
Pw (20)
Twr = In generator, motion equation single shaft model is used
ω wr (12) and it is assumed that converter controls can be able to filter
Here Pw is the mechanical power extracted from wind. shaft dynamics. For this reason, tower shadow effect is not
Wind and the rotor speeds are related as considered. So
ρ ω m = (Tm − Te ) / 2 H m
Pw = c p (λ )Ar vw3 (13)
2 Te = ψ ds iqs − ψ qs i ds
(21)
ρ = air density
The relation between generator currents, stator flux is
Cp= Rotor power coefficient given as:
λ = Tip speed ratio
ψ ds = −((x S + x m )ids + x m idr )
Ar = Area of cross section of rotor
ψ qs = −(( x S + x m )i qs + x m i qr )
v w = Wind speed (22)
Thus the electrical torque Te is given by
B. DFIG Wind turbine modeling Te = xm (iqr ids − idr iqs ) (23)
Doubly fed induction generator steady state electrical
And the mechanical torque, extracted mechanical power
equations are assumed, the flux dynamics of stator and rotor
equations are same as mentioned in SCIG.
are fast in comparison with grid dynamics and the generator
Converter dynamics are simplified and they are fast as
decoupling from the grid can be done by the converter control
electromechanical transients are considered. Here converter is
mechanism. These assumptions lead to
v ds = −rS ids + (( x S + x m )iqs + x m iqr ) modeled as an ideal current source. iqr , idr are state variables
and they are used to control both rotor speed, voltage
v qs = − rS iqs − (( x S + x m )ids + x m i dr ) respectively. Converter currents differential equations are as
(14)
v dr = − rR idr + (1 − ω m )(( x R + x m )iqr + x m i qs )
follows:
⎛ x + xm * ⎞ 1
v qr = − rR i qr − (1 − ω m )(( x R + x m )idr + x m ids ) iqr = ⎜⎜ − s Pw (ω m ) / ω m − iqr ⎟⎟
⎝ x mV ⎠ T∈
Where the stator voltage depends on both grid voltage
magnitude and phase: idr = K V (V − Vref ) − V / x m − idr (24)
v ds = V sin (− θ )
Where P
*
(ω m ) is known as power – speed characteristic
v qs = V cos(θ )
w

(15) which approximately optimizes the wind energy capture and it

The injected active and reactive power in to the grid is calculated based on current rotor speed value [12].
depends on both stator current and grid side current of the
converter. Finally equations are as follows:
III. TEST SYSTEMS
P = vds ids + vqs iqs + vdc idc + vqc iqc
In this paper Weibull distribution (which is commonly
Q = v qs i ds − v ds iqs + v qc idc − v dc i qc used to represent wind profile) wind speed model with
(16) maximum wind speed as 15m/s is used. Here wind speed
This can be rewrite by considering converter power profile representation is in PU. Weibull distribution is
equations which are shown below. Grid side converter powers presented in this case as shown in Fig. 2.
are:
Pc = v dc idc + v qc i qc
Qc = v qc i dc − v dc iqc
(17)
 0.85

0.8 Wind speed

0.75
W in d S p e e d in P U

0.7

0.65

0.6

0.55 Fig. 4. DFIG connected to modified IEEE14 bus system

0.5
0 2 4 6 8 10 12 14 16 18 20
IV. SIMULATION RESULTS & DISCUSSIONS
time (s)
Fig. 2. Weibull distribution wind speed
The impact of wind turbines on power system leads to a
Based on above mentioned modeling analysis, simulation small signal stability problem. In SCIG based system all poles
of modified IEEE 14 bus system (test system) which is based lie on the left hand side except one pole, which is on the origin
on PSAT has been built as follows. Time domain simulation as shown in Fig. 5, this shows that system is marginally stable.
has been performed and simulation time is set as fifty seconds In DFIG based system all poles lie completely on the left hand
side as shown in Fig. 6, so system is completely stable as small
(t=50s). At bus number 3 SCIG and DFIG wind turbines are
signal stability is consider. For further clarification, complete
connected as shown in figures Fig. 3 and Fig. 4 respectively.
statistics of SCIG, DFIG based systems regarding small signal
stability are given below after graphical representation of
Bus 13
individual systems.

20
Bus 14
Bus 10
15
Bus 12

10
Bus 09
Bus 11
5
Imag

Bus 07
0

Bus 06 -5
Bus 08
Bus 04
-10

Bus 05
-15
Bus 01

-20
-50 -45 -40 -35 -30 -25 -20 -15 -10 -5 0
Real
Bus 02

Fig. 5. S Domain analysis graphical representation of modified IEEE14 bus

system with SCIG

Bus 03 STATISTICS
DYNAMIC ORDER 49
EIGS WITH Re(mu) < 0 48
EIGS WITH Re(mu) > 0 0
Fig. 3. SCIG connected to modified IEEE14 bus system REAL EIGS 25
COMPLEX PAIRS 12
ZERO EIGS 1
 1.2
15
1

10 0.8
δ δ δ δ
Syn 1 Syn 2 Syn 3 Syn 4

P o w e r a n g le in P U
0.6

5 0.4

0.2
Imag

0 0

-0.2
-5
-0.4

-0.6
0 5 10 15 20 25 30 35 40 45 50
-10
time (s)

Fig. 8.Power angle deviations of modified IEEE14 bus system with DFIG
-15
-120 -100 -80 -60 -40 -20 0
Real 1.12
VBus 01
VBus 02
Fig. 6. S Domain analysis graphical representation of modified IEEE14 bus 1.1
VBus 03
system with DFIG
VBus 04
1.08
STATISTICS VBus 05

V o lt a g e in P U
DYNAMIC ORDER 48 1.06
VBus 06
VBus 07
EIGS WITH Re(mu) < 0 48
VBus 08
1.04
EIGS WITH Re(mu) > 0 0 VBus 09
REAL EIGS 30 1.02 VBus 10
VBus 11
COMPLEX PAIRS 9
1 VBus 12
ZERO EIGS 0 VBus 13

Here two small signal stability problems have been 0.98

0 5 10 15 20 25 30 35 40
VBus 14
45 50
discussed and they are rotor angle stability, voltage stability time (s)
and their simulations results have shown above. In case of
SCIG power angles (rotor angles) are deviated as shown in Fig. Fig. 9. Voltage variations of modified IEEE14 bus system with SCIG
7, by which the system is said to be near to instability. Where 1.12 VBus 01
as in case of DFIG rotor angles are almost constant with time
VBus 02
(in Fig. 8.) by which we can conclude that system is
VBus 03
completely stable. Coming to voltages at individual buses, 1.1

SCIG based systems having fluctuations but DFIG based VBus 04

systems voltage almost remains constant as shown in figures 1.08 VBus 05

Fig. 9 and Fig. 10 respectively. From the above simulation

V o l t a g e in P U

VBus 06
results it is clear that DFIG based system is more suitable to VBus 07
1.06
integrate into power system for wind power generation when VBus 08
compared to SCIG based system as power system stability is VBus 09
consider. 1.04 VBus 10

10 VBus 11
1.02 VBus 12
8
VBus 13
VBus 14
6 1
0 5 10 15 20 25 30 35 40 45 50
P o w e r a n g le in P U

4
time (s)

2
Fig. 10. Voltage variations of modified IEEE14 bus system with DFIG
δSyn 1 δSyn 2 δSyn 3 δSyn 4
0 V. CONCLUSION

-2
This research presents the modified IEEE14 bus system
model of which novelty consists in its implementation in a
-4
0 5 10 15 20 25 30 35 40 45 50
free and open source software: Power System Analysis
time (s) Toolbox. The model takes into account detailed modeling of
the dynamics which play an important role in the assessment
Fig. 7. Power angle deviations of modified IEEE14 bus system with SCIG
of the system’s behavior. Of particular significance is the [5] M. Q. Duong, F. Grimaccia, S. Leva, M. Mussetta, E. Ogliari, “Pitch
angle control using hybrid controller for all operating regions of SCIG
implementation of the recently developed wind turbine and wind turbine system,” Renewable Energy, Elsevier, Vol. 70, pp. 197-
controller model in this tool with modified IEEE14 bus test 203, 2014.
system in which, most of grid’s power plants are thermal and [6] M. Q. Duong, F. Grimaccia, S. Leva, M. Mussetta, and R. Zich,
hydro generators. Small signal stability analyses of the “Improving LVRT characteristics in variable-speed wind power
considered grid utilizing eigenvalue analysis are used to generation by means of fuzzy logics,” in 2014 IEEE International
Conference on Fuzzy Systems (FUZZ-IEEE), pp. 332-337, July 2014.
demonstrate the importance of accurate modelling.
[7] H. Schwarz, K. Pfeiffer, and L. Roskoden, “Integration of renewable
Simulation results show that SCIG based system is eneries to the east German grid. Actual problems and possibe solutions,”
marginally stable where as DFIG based system is completely in Proc. Of the 6th WSEAS International Conference on Power Systems,
stable and rotor angle stability, voltage stability of DFIG is pp. 7-13, 2006.
more predominant than SCIG. SCIG consisting of a fixed [8] A. G. Pillai, P. Thomas, K. Sreerenjini, S. Baby, T. Joseph, and S.
capacitor which may not sufficient to meet reactive power Srecdharan, “Transient stability analysis of wind integrated power
systems with storage using central area controller,” in 2013 Annual
demand by the system. Where, DFIG consists of power International Conference on Emerging Research Areas and 2013
electronics based controlling system by which it can supply International Conference on Microelectronics, Communications and
reactive power to the system if necessary and thereby it can Renewable Energy, pp. 1-5, 2013.
enhance small signal stability of the system. [9] Yanhua Liu and Yongning Chi and Weisheng Wang and Huizhu Dai,
“Impact of large scale wind farm integration on power system transient
REFERENCES stability,” in 2011 4th International Conference on Electric Utility
Deregulation and Restructuring and Power Technolgies, pp. 1301-1305,
[1] F. Grimaccia, M. Mussetta, R.E. Zich, "Neuro-fuzzy predictive model 2011.
for PV energy production based on weather forecast," 2011 IEEE [10] Y. Chompoobutrgool, and Wei Li and L. Vanfretti, “Development and
International Conference on Fuzzy Systems (FUZZ-IEEE), pp.2454- implementation of a Nordic grid model for Power System small-signal
2457, 27-30 June 2011. and transient stability studies in a free and open source software,” in
[2] M.Q. Duong, E. Ogliari, F. Grimaccia, S. Leva, M. Mussetta, "Hybrid 2012 IEEE Power and Energy Society General Meeting , pp. 1-8, 2012.
model for hourly forecast of photovoltaic and wind power," 2013 IEEE [11] Y.U. Lopez, J.A.D. Navarro, “Small Signal Stability analysis of wind
International Conference on Fuzzy Systems (FUZZ-IEEE), 7-10 July turbines with squirrel cage induction generators,” IEEE International
2013. Conference on Transmission and Distribution Conference and
[3] G. N. Sava, S. Costinas, N. Golovanov, S. Leva, and M. Q. Duong, Exposition pp.1-10, 2008.
“Comparison of active crowbar protection schemes for DFIGs wind [12] F. Milano, “Power System Analysis Toolbox,” Version 1.3.4, Software
turbines,” in 2014 IEEE 16th International Conference on Harmonics and Documentation, July 14, 2005.
and Quality of Power, pp. 448-552, 2014.
[4] M. Q. Duong, K. H. Le, F. Grimaccia, S. Leva, M. Mussetta, and R.
Zich, “Comparison of power quality in different grid-integrated wind
turbines,” in 2014 IEEE 16th International Conference on Harmonics
and Quality of Power, pp. 669-673, 2014.