256

Int. J. Renewable Energy Technology, Vol. 1, No. 3, 2010

Long-term stabilisation of grid connected wind farms with FACTS controllers N. Senthil Kumar* and M. Abdullah Khan
Department of Electrical Engineering, B.S.A. Crescent Engineering College, Vandalur, Chennai-48, India E-mail: nsksai@rediffmail.com *Corresponding author
Abstract: The increasing power demand has led to the growth of new technologies that play an integral role in shaping the future energy market. Keeping in view of the environmental constraints, grid connected wind turbines are promising in increasing system reliability. This paper presents the impact of FACTS controllers on the long-term dynamic stability of power systems connected with wind farms based on doubly fed induction generator systems. The stability assessment is made first for a three-phase short circuit without the FACTS controllers in the power network and then with the FACTS controllers. The long-term dynamic simulation results yield information on: 1 the impact of faults and slow wind speed changes on the performance of wind driven induction generators 2 the change in controllable parameters of the FACTS controllers following the faults and wind speed changes 3 transient reactive power ratings of FACTS controllers for enhancement of rotor speed stability of induction generators and angle stability of synchronous generators. Keywords: wind energy conversion systems; flexible AC transmission systems; rotor angle stability; rotor speed stability; long-term dynamic simulation. Reference to this paper should be made as follows: Kumar, N.S. and Khan, M.A. (2010) Long-term stabilisation of grid connected wind farms with FACTS controllers, Int. J. Renewable Energy Technology, Vol. 1, No. 3, pp.256278. Biographical notes: N. Senthil Kumar is presently working as an Assistant Professor (Senior Grade) in the Department of Electrical and Electronics Engineering at BSA Crescent Engineering College. He has 12 years of teaching and six years of research experience. He has undergone the EUROSTAG software training programme at Tractable Engineering, Brussels, Belgium. His area of research includes modelling of FACTS devices for power system studies and modelling of wind energy conversion systems for power system stability analysis. He has three international journals to his credit. M. Abdullah Khan has 40 years of teaching and research experience. His area of research includes power system analysis and optimisation and power system dynamics. He has published several papers in reputed international journals.

Copyright 2010 Inderscience Enterprises Ltd.

Long-term stabilisation of grid connected wind farms with FACTS controllers 257

Introduction

As power systems became interconnected, areas of generation were found to be prone to electromechanical oscillations. These oscillations have been observed in many power systems worldwide. As the level of power transmission rose, largely through existing interconnections, which were becoming relatively weak and inadequate, load characteristics added to the problem causing spontaneous oscillations. The oscillations may be local to a single generator or generator plant (local oscillations, 1.02 Hz) or they may involve a number of generators widely separated geographically (interarea oscillations, 0.20.8 Hz). If not controlled, these oscillations may lead to total or partial power interruption (Kundur, 1994). Wind energy development is consumer and environment friendly, it requires shorter construction time and is cost competitive. It becomes one of the most competitive sources of renewable energy. However, wind power has some disadvantages. For example, wind power is considered as an intermittent power supply because wind does not blow 100% of the time. Besides, the superior wind sites are usually located in remote areas; therefore, it may require substantial infrastructure improvement to deliver the wind-generated power to the load centre. There are four major types of wind generators, which are used very widely: 1 2 3 4 squirrel cage induction generators doubly fed induction generators direct driven synchronous generator permanent magnet synchronous generator.

For the present work, we have used doubly fed induction generator model for executing dynamic simulations. The dynamic stability of a single wind turbine generator supplying an infinite bus through a transmission line was studied by developing the linearised model of the power system under different loading conditions (Abdel-Magid and El-Amin, 1987). The effect of wind turbines on the transient fault behaviour of the Nordic power system was investigated for different faults in Jauch et al. (2007). A novel error driven dynamic controller for the static synchronous compensator (STATCOM) FACTS device was designed to stabilise both wind energy conversion stand-alone system as well as hybrid scheme of wind turbine with hydro generators (Elmoursi and Sharaf, 2006). A new definition on rotor speed stability of asynchronous generators is proposed in Samuelsson and Lindahl (2005). A control structure for DFIG-based turbines under unbalanced conditions is proposed in Xiu and Wang (2007). The application of voltage source converter (VSC) based transmission controllers for wind energy conversion systems is discussed in Varma and Sidhu (2006). Wind turbine model for power system stability analysis was derived and discussed in Coughlan et al. (2007). The dynamic behaviour of the power system is analysed with high wind power penetration in Akhmatov (2003). The impact of FACTS controllers on the rotor speed/rotor angle stability of power systems connected with wind farms is discussed in Kumar and Khan (2008). The objective of the present work is to study the impact of FACTS controllers on the long-term dynamic stability of power systems with wind farms based on doubly

258

N.S. Kumar and M.A. Khan

fed induction generators. The dynamic stability of the system is studied by running time domain simulations without and with FACTS controllers. The following FACTS controllers are taken up for the analysis: 1 2 3 4 static var compensator (SVC) static synchronous compensator (STATCOM) thyristor controlled series capacitor (TCSC) unified power flow controller (UPFC).

The induction generator absorbs reactive power during its normal operating condition. This may create low voltage and dynamic instability in the system (Chompoo-Inwai et al., 2005). This paper is organised as follows. Section 2 presents the modelling of power system and wind turbine along with FACTS controllers. Section 3 presents the dynamic simulation results obtained for a three-phase transient fault on the system with and without FACTS controllers. Section 4 presents the results and discussion on the simulation results obtained. Section 5 presents the conclusions.

Modelling of the power system and induction generator

In general, for the purpose of dynamic analysis, power systems are modelled by a set of differential and algebraic equations (DAE), that is:
x = f (x, y, , p) 0 = g(x, y, , p)

(1)

where x R n is a vector of state variables associated with the dynamic states of generators, loads and other system controllers; y R m is a vector of algebraic variables associated with steady state variables resulting from neglecting fast dynamics (e.g., most load voltage phasor magnitudes and angles); R l is a set of uncontrollable parameters, such as variations in active and reactive power of loads; and p R k is a set of controllable parameters such as transformer tap, AVR settings and FACTS controller reference voltages (Mithulananthan et al., 2003).

2.1 Modelling of wind energy conversion system

Normally, a wind turbine creates mechanical torque on a rotating shaft, while an electrical generator on the same rotating shaft is controlled to produce an opposing electromagnetic torque. The power and torque equations for the wind turbine are as follows. The doubly fed induction generator is the most commonly used machine for wind power generation. The rotor terminals are fed with a symmetrical three-phase voltage of variable frequency and amplitude. This voltage is supplied by a VSC usually equipped with IGBT-based power electronics circuitry. The basic structure of the DFIGbased wind energy conversion scheme is shown in Figure 1.

Long-term stabilisation of grid connected wind farms with FACTS controllers 259
P= 1 Cp AV3 2 (2)

T= where P T A V

(3)

output power of the turbine (W) mechanical torque (Nm) rotor speed of wind turbine (rad/s) density of air (= 1.22 kg/m3) swept area of the blade (m2) wind speed (m/s).
Wind energy conversion scheme and its associated controls

Cp performance coefficient

Figure 1

The wind farm is represented as an aggregated model of ten wind turbines of each 2 MW (EUROSTAG Release 4.3 Package Documentation Parts 1, 2 and 3, 2004). The main controller manages the overall control functions, whereas pitch and power controller are subordinate units. The wind energy conversion scheme used for

260

N.S. Kumar and M.A. Khan

simulation consists of a doubly fed induction generator (rotor circuit connected to the grid through power electronic converter). The converter adjusts the frequency of the rotor feeding current to enable variable speed operation. The power electronic converter consists of two-VSCs connected through capacitor. If shaft, turbine and generator damping are neglected, the two-mass model is described by the following equations:
Tt = J t Te = J g d t + K s s dt dr K s s dt

(4) (5)
(6)

d = t r dt

where Tt is the mechanical torque referred to the generator side (Nm), Te is the electromagnetic torque (Nm), Jt is the equivalent turbine-blade inertia referred to the generator side (kg m2), t is the turbines rotational speed (rad/s), r is the generators rotational speed (rad/s), Ks is the shaft stiffness (N.m/rad) and s is the angular displacement between the ends of the shaft (rad). Figure 2 gives the two mass representation of the wind turbine (Gaztanaga et al., 2007).
Figure 2 Two-mass representation of the wind turbine

2.2 Doubly fed induction generator model

Equations (7) to (11) represent the complete set of mathematical relationships that describe the dynamic behaviour of the machine. The per unit system is adopted as a unit of measurement for all quantities and the sign convention is chosen in such a way that consumed inductive reactive powers are positive (note also that o is equal to 1.0 p.u.).

Voltage equations: d s = rs Is + js Vs dt d r = rr Ir + j( r ) r Vr dt (7) (8)

Flux linkages:
s = ls I s + l m I r

(9)

Long-term stabilisation of grid connected wind farms with FACTS controllers 261

r = l m Is + l r I r Equations of motion:
dr 1 = ( sd isq sq isd + t m ) dt s where Is, Ir Vs, Vr s, r lm rs, rr r, d, q-direct tm stator and rotor currents stator and rotor terminal voltages stator and rotor flux linkages mutual inductance (in per unit it is equal to Xm) stator and rotor resistances rotor angular speed, synchronous speed quadrature axis component mechanical torque.

(10)

(11)

The DFIG model used is a third order model [equations (7), (8) and (11)], the state variables being stator, rotor flux components and rotor speed. The variable frequency rotor voltage permits the adjustment of rotor speed to match the optimum operating point at any practical wind speed. In normal mode, the rotor side converter is used to control the real and reactive power outputs of the machine. Independent control of real and reactive power can be achieved through rotor current control. The rotor current controller is modelled using the model editor menu of EUROSTAG. From the basic equations of DFIG (seven to ten) setting all derivatives to zero (steady state) and with stator resistance rs = 0, we get:
(v jx m i r ) + x r ir Vr = rr i r + js x m s jx s

(12)

Considering a coordinate system where the d-axis is located along Vs, it follows that:
x Vrd = rr i rd + s m ( Vs ) i rq x r xs
Vrq = rr i rq + s i rd x r

(13) (14)

x2 where leakage coefficient = 1 m is introduced. x x r s s is the operating slip of the generator. The voltage drops over the rotor resistance in (13) and (14) can be interpreted as auxiliary signals, which are outputs of the intended rotor current controller. PI controllers are introduced to control the rotor voltages and hence rotor currents.

262

N.S. Kumar and M.A. Khan

1 Vrd = K I 1 + ( i rd ref i rd ) pT1 (15)

and

1 Vrq = K I 1 + ( i rq ref i rq ) pT1

(16)

The corresponding block diagram of the rotor current controller is shown in Figure 3. The wind turbines use a simplified control scheme and are not LVRT equipped.
Figure 3 Rotor current controller (see online version for colours)

The line or grid side converter has to transmit the active power from the dc link to the grid so that the dc link voltage is kept within limits. The corresponding controller and converter model is shown in Figure 4. The output is the active current that is injected into the grid node.
Figure 4 Block diagram of line side converter control

Long-term stabilisation of grid connected wind farms with FACTS controllers 263

2.3 Synchronous generators

In power system dynamic studies, the synchronous generator is commonly represented by using models of varying degrees of complexity the simplest being the classical (second order) model that assumes a constant voltage behind transient reactance. The synchronous machine model used for this dynamic analysis is the two-axis model with four state variables (Ed, Eq, , ).

2.4 Static var compensators

SVC is basically a shunt connected static var generator/absorber whose output is adjusted to exchange capacitive or inductive current so as to maintain or control specific power system variables; (Figure 5) typically, the controlled variable is the SVC bus voltage. One of the major reasons for installing an SVC is to improve dynamic voltage control, and thus, increase system loadability. The SVC bus is modelled as a PV bus in the load flow programme. It is modelled as a variable susceptance controller as shown in Figure 6 for the execution of the dynamic simulation programme.
Figure 5 Ideal SVC

Figure 6

Dynamic model of SVC

264

N.S. Kumar and M.A. Khan

An additional stabilising signal and supplementary control superimposed on the voltage control loop of an SVC can provide damping of system oscillations during transient faults and disturbances.

2.5 Static synchronous compensator

Figure 7 Block diagram of STATCOM

Figure 8

Dynamic model of STATCOM

Long-term stabilisation of grid connected wind farms with FACTS controllers 265

The STATCOM resembles in many respects a synchronous compensator, but without inertia. The basic electronic block of a STATCOM is the VSC, (Figure 7) which in general converts an input dc voltage into a three-phase output voltage at fundamental frequency, with rapidly controllable amplitude and phase. is the phase shift between the controller VSC AC voltage and its bus voltage V. Vref is the reference voltage setting. A phase control strategy is assumed for control of the STATCOM bus voltage and additional control block and signals added for oscillation damping as shown in Figure 8.

2.6 Thyristor controlled series capacitors

TCSC schemes typically use a thyristor-controlled reactor in parallel with a capacitor to vary the effective compensating reactance. In practice, several capacitor banks, each with its own thyristor-controlled reactor, may be used to meet the specific application requirements. For the purpose of analysis, the TCSC, regardless of its practical implementation, can be considered simply as a continually variable capacitor whose reactance is controllable in the range of 0 Xc Xcmax. The controllable series capacitive reactance cancels part of the reactive line reactance resulting in a reduced overall transmission impedance and correspondingly increased transmittable power. The single line diagram of the TCSC is shown in Figure 9.
Figure 9 Thyristor controlled series capacitor

Figure 10 Dynamic model of TCSC

266

N.S. Kumar and M.A. Khan

The variable reactance model of TCSC is as shown in Figure 10. Vcap is the voltage across the capacitor and Xc is the variable reactance inserted by the TCSC module into the line. Xref is the steady state reactance of the fixed capacitor under normal conditions to maintain the steady state power flow (Mathur and Varma, 2002).

2.7 Unified power flow controllers

The UPFC is the most versatile FACTS controller developed so far, with all encompassing capabilities of voltage regulation, series compensation and phase shifting. It comprises two VSCs coupled through a common dc link. The series converter (VSC2) is controlled to inject a voltage phasor Vse, in series with the line, which can be varied from zero to Vsemax and the phase angle of Vse can be independently varied from zero to 360. The single line diagram is shown in Figure 11. The UPFC is modelled in the power flow programme using the power injection model with two real and reactive power injections at two nodes of the system. The power injections at both the nodes are selected such that the base case power flow with induction generator is maintained. The active and reactive power flow control loops of the UPFC are shown in Figures 12 and 13.
Figure 11 Single line diagram of UPFC

Figure 12 Active power control loop

Long-term stabilisation of grid connected wind farms with FACTS controllers 267
Figure 13 Reactive power control loop

The stabilising signal for the UPFC is derived from a power oscillation-damping block, which uses active power flow as the input signal. Pflow ref is the reference value of active power flow in the line on which UPFC is connected. The reference value of the active power flow can be set depending on the operating condition of the power system. Pflow is the actual active power flow in the line. Vseq is the component of series injected voltage in quadrature with the line current. Qref from the above figure is the reference value of reactive power flow in the UPFC controller. Qflow is the actual reactive power flow in the line and Vsep is the component of AC voltage injected in phase with the line current.

2.7.1 Parameter tuning

The gains of the FACTS controllers in the forward path of the transfer function are tuned by using an optimisation algorithm which minimises the voltage oscillations of the induction generator bus. The tuning is posed as an optimisation problem with the objective as minimising the oscillations of point of common coupling (PCC) voltage from the desired value and is given by:
Minimise: SSDI = K min K K max

(V
k

ref

Vk ) 2

(17)

where Vref = 1.0 per unit and SSDI is the sum squared deviation index of the PCC voltage.

Long-term dynamic simulation results and stability investigation

The single line diagram of the test system with the wind turbine connected is shown in Figure 14. The test system consists of a seven bus system with two synchronous generators G1 and G2. The doubly fed induction generator (IG) is connected to the grid through a three winding transformer. IG denotes the stator of the doubly fed induction generator. At node ST, the stator of induction generator is connected, and at node RT, the rotor of the doubly fed induction generator is connected denotes the rotor of the induction generator.The dynamic performance of the FACTS controllers with wind turbines is investigated following the disturbance sequence mentioned below. 1 at t = 2 seconds, a three-phase fault on line 23 in the second line near bus 3

268 2 3 4

N.S. Kumar and M.A. Khan

at t = 2.125 seconds, the three-phase fault is cleared at t = 20 seconds, the wind speed increases from 2 m/s to 7 m/s (a ramp increase) at t = 50 seconds, the load at bus 5 is increased by 15 MW (load increase).

At bus 5, the load is represented as a combination of impedance and voltage frequency dependant load in the dynamic simulation. The shunt connected FACTS controllers (SVC and STATCOM) are located at bus 3 and the series connected FACTS controllers are located in one of the lines in grid wind farm line (23). The total MW load on the two load buses 5 and 4 of the test system is 500 MW and 5,000 MW respectively. The steady state active power generated by generator G1 800 MW and that of G2 is 5,000 MW. The wind genertor (IG) supplies 5.5 MW in the steady state. This specific test system is chosen for the dynamic simulation study as this system has only two synchronous machines which is good enough for conducting stability investigation on a wind farm. The doubly fed induction generator is modelled as two active power injections in the load flow programme of EUROSTAG at nodes ST and RT. The FACTS controllers are modelled as power injections or current injections in the load flow programme. The SVC is modelled as a shunt reactive current in the load flow programme. The TCSC and UPFC is modelled with two power injections between buses 23.
Figure 14 Single line diagram of the power system with wind turbine stator connected to node ST and rotor connected to node RT

BUS 4 BUS 1 G1 G2 BUS 5 IG RT ROTOR

The dynamic simulation results will be presented in the following sequence: 1 2 rotor angle deviations of synchronous generators with wind farm without and with FACTS controllers rotor speed deviation of induction generator without FACTS controllers and with FACTS controllers

BUS 2

BUS 3

ST

Long-term stabilisation of grid connected wind farms with FACTS controllers 269

3 4 5

active power injected by the wind turbine stator and rotor without and with FACTS controllers induction generator termianl (stator and rotor) voltages without and with FACTS controllers controllable outputs of the FACTS controllers following the disturbance and wind speed changes.

3.1 Rotor angle deviations of synchronous generators with wind farm and with FACTS controllers
Figure 15 shows the rotor angle deviation of the synchronous generator G1 without FACTS controllers in the network. From Figure 15, it can be observed that after the fault the generator rotor angle of G1 oscillates around the steady state rotor angle value of 65.8. It can be observed that after the load increase at time t = 50 seconds, the rotor angle of synchronous generator accelerates to a new equilibrium value. The rotor angle response of the generators with SVC is stable with a settling time of ten seconds after the transient fault. The controller parameters of the SVC/STATCOM are tuned to stabilise the oscillations. The dynamic performance of STATCOM and TCSC are almost comparable and give a stable response after the load increase at time t = 50 seconds. The peak overshoot following the load shedding is more with STATCOM than SVC.
Figure 15 Rotor angle deviation of synchronous generator G1 without and with FACTS controllers

270

N.S. Kumar and M.A. Khan

From Figure 15, it can be observed that there are no sustained oscillations with UPFC in the network even after the load increase at bus 5. Also, there are no oscillations following load increase at bus 5 in 50 seconds. This is due the fact that the shunt branch UPFC contributes reactive power to the network after the transient fault and subsequent changes in the network conditions. Table 1 shows the comparative analysis of different FACTS controllers in the network. The comparison is made on the basis of the settling time obtained after the three-phase fault is cleared with FACTS controllers in the network.
Table 1 Comparison of SVC, STATCOM, TCSC and UPFC for enhancement of dynamic stability in a multimachine power system with DFIG after the transient fault Without FACTS 20 sec SVC 10 sec STATCOM 1 sec TCSC 2 sec UPFC 1 sec

S. no. Settling time

All the FACTS controllers are located in the electical midpoint of the transmission line, so that intergenerator oscillations occuring between the two areas are damped out effectively.

3.2 Rotor speed deviation of induction generator without FACTS controllers and with FACTS controllers
Figure 16 Rotor speed deviation of induction generator FACTS

Long-term stabilisation of grid connected wind farms with FACTS controllers 271

Figure 16 demonstrates the rotor speed deviation of the induction generators without FACTS controllers in the network. From Figure 16, it can be observed that the rotor speed deviation without FACTS controllers in the network following wind speed increase at time t = 20 seconds is more than 1.12 p.u. It can be observed that the DFIG speed remains closer to the steady state operating speed (0.9 p.u.) with UPFC in the network. This is due to the reactive power contribution from both the series and shunt branches of UPFC. The DFIG speed with SVC is around 1.12 p.u. and with STATCOM it is near 1.06 p.u. This can be attributed due to the faster trasient response of the STATCOM than the SVC.

3.3 Active power injected by the wind turbine stator and rotor without and with FACTS controllers
Figure 17 shows the active power injected by the wind farm into the grid following the grid disturbance and wind speed increase. The stator protection system associated with the induction generator disconnects the stator from the grid if the terminal voltage of the induction generator is less than 0.75 p.u. for a period of 0.1 seconds, hence, the stator active power delivered comes down to zero after the fault. The stator of the induction generator is connected back to the grid after a period of 60 seconds by a reconnection mechanism. Observe the solid line in Figure 17 which shows the active power injected into the grid by DFIG without FACTS.
Figure 17 Active power injected by stator and rotor of induction generator without FACTS

272

N.S. Kumar and M.A. Khan

The active power injected comes down to zero from its initial value of 5.5 MW specified in the load flow, as the DFIG protection system disconnects the stator from the grid at two seconds because of the transient fault. The fault is cleared at time t = 2.125 seconds. The induction generator terminal voltage increases and the machine is connected back to the grid after a specified time delay at time t = 65 seconds without FACTS controllers in the network. With the FACTS controllers in the network, the voltage does not reduce below 0.75 p.u. Hence, the induction generator is not disconnected from the grid. At time t = 20 seconds, the wind speed increases to 7 m/s. Hence, the real power injected by the induction generator increases from its initial value of 5.5 MW to 18 MW. The power calculation according to (2) is based on a single wind speed. However, in reality, the wind speed may differ slightly in direction and intensity across the area traversed by the blades. To consider this effect, the wind speed is supplied through a lag block to the power conversion equation. This creates a slight change in the active power delivered to the grid before the disturbance at one second (Figure 17). It can be observed that the STATCOM has an improved transient response compared to SVC. There are no sustained oscillations in active power delivered to the grid with STATCOM (Figure 17). It can also be observed that the wind farm is not disconnected from the power grid with shunt connected FACTS controllers in the network. After the load increase at time t = 50 seconds, the stabilising loops of all the FACTS controllers (Figures 6, 8, 10 and 12) are activated which modulates the line active power flow and regulates the bus voltage. The active power injected reduces only by a fraction of an MW with UPFC and STATCOM in the network.

3.4 Induction generator termianl voltages without and with FACTS controllers
The response of PCC terminal voltage following the transient fault and wind speed increase without and with FACTS is shown in Figure 18.
Figure 18 Voltage at bus 5 (PCC voltage) without and with FACTS (see online version for colours)

Long-term stabilisation of grid connected wind farms with FACTS controllers 273

From Figure 18, it is clear that after subsequent fault clearing the voltage at the stator terminals of the induction generator lies at around 0.7 p.u. after the fault clearance. This leads to disconnection of DFIG stator terminals from the network as the under voltage relay is set with a voltage threshold of 0.75 p.u. for 0.1 seconds. However, with UPFC and STATCOM in the network, the dip in the PCC terminal voltage is only 0.8 p.u. after the transient fault. Hence, the induction generators are not disconnected from the grid.

3.5 Controllable outputs of the FACTS controllers following the disturbance and wind speed changes
Figures 19 and 20 show the change in transient ratings of FACTS controllers following the disturbance and wind speed changes with SVC and STATCOM controllers. It can be observed from Figure 19 that there are transients in the reactive power injected after the load shedding at time t = 50 seconds. From Figures 19 and 20, it can be observed that the dynamic MVAr rating of STATCOM is lesser than SVC for the transient fault and wind speed changes. The reactive power exchanged with the system swings from (10 MVAr to 50 MVAr) with SVC, whereas the swing in reactive power is only (0.08 MVAr to 0.06 MVAr) following the disturbance and wind speed changes. The corresponding change in reactive power at the TCSC and UPFC terminals is shown in Figures 2122.
Figure 19 Dynamic reactive power support provided by SVC

274

N.S. Kumar and M.A. Khan

Figure 20 Dynamic reactive power support provided by STATCOM

Figure 21 Change in reactive power at TCSC terminals

Long-term stabilisation of grid connected wind farms with FACTS controllers 275
Figure 22 Change in reactive power at UPFC terminals

Figure 21 shows the change in TCSC reactive power at its terminals. The dynamic reactive power support provided by the TCSC terminals swings from 350 MVAr to +350 MVAr after the transient fault and wind speed changes. The corresponding change in reactive power rating of UPFC is between (400 MVAr to +500 MVAr) after the disturbance (Figure 22). There is no appreciable change in reactive power delivered to the network following wind speed changes for both TCSC and UPFC (observe the graph between 1050 seconds).

Results and discussion

From the results, we conclude that among shunt connected FACTS controllers, the STATCOM provides better damping the oscillations compared to that of the SVC. Also, the dynamic reactive compensation (transient rating) required by the STATCOM is lesser than SVC. This is due to the fact that a STATCOM is basically voltage sourced converter-based shunt controller and SVC is a shunt connected passive thyristor switched capacitor/reactor. Among the series connected FACTS controllers, the UPFC damps both rotor angle oscillations of synchronous generators and rotor speed oscillations of induction generator very effectively when compared with TCSC. This is due to the reactive support provided by the shunt branch of the UPFC following the disturbance. However, the reactive power rating of UPFC is very high compared to that of the TCSC. It is suggested that a STATCOM of suitable rating may be installed at the PCC with or

276

N.S. Kumar and M.A. Khan

without a capacitor may be used for stabilising rotor speed oscillations associated with doubly fed variable speed induction generators following transient faults and wind speed increase.

Conclusions

The development of wind turbine and wind farm models is vital because as the level of wind penetration increases, it poses dynamic in stability problems in the power system. For the present work, we have taken a taken a doubly fed induction generator model and illustrated the presence of sustained oscillations with wind farms. This paper has described a comprehensive study of the application of both shunt and series connected FACTS controllers to a wind farm for enhancement of power system damping. The dynamic simulation study is carried out to take care of wind speed changes. All the FACTS controllers modelled make use of a steady state loop (primary objective of regulating bus voltage or maintaining power flow) and a stabilising loop. The transient reactive power ratings of FACTS controllers are arrived after carrying out long-term dynamic simulation study.

Acknowledgements
I sincerely thank my father Mr. S.K. Natarajan and wife Bhuvana for encouraging me at times of pressure during the research work.

References
Abdel-Magid, Y.L. and El-Amin, I.M. (1987) Dynamic stability of wind-turbine generators under widely varying load conditions, Electrical Power and Energy Systems, July, Vol. 9, No. 3, pp.180188. Akhmatov, V. (2003) Analysis of dynamic behavior of electric power systems with large amount of wind power, PhD thesis, Technical University of Denmark, April. Chompoo-Inwai, C., Lee, W-J., Fuangfoo, P., Williams, M. and Liao, J.R. (2005) System impact study for the interconnection of wind generation and utility system, IEEE Transactions on Industry Applications, Vol. 41, No. 1, pp.163168. Coughlan, Y., Smith, P., Mullane, A. and Malley, M.O. (2007) Wind turbine modeling for power system stability analysis a system operator perspective, IEEE Transactions on Power Systems, Vol. 22, No. 3, pp.929936. Elmoursi, M.S. and Sharaf, A.M. (2006) Novel STATCOM controllers for voltage stabilization of stand alone hybrid (wind/small hydro) schemes, International Journal of Emerging Electric Power Systems, Vol. 7, No. 3, Art. 5, pp.127. EUROSTAG Release 4.3 Package Documentation Parts 1, 2 and 3 (2004). Gaztanaga, H., Etxeberria-Ottadui, I. and Ocnasu, D. (2007) Real time analysis of the transient response improvement of wind farms by using a reduced scale STATCOM prototype, IEEE Transactions on Power Systems, May, Vol. 22, No. 2, pp.658666. Jauch, C., Sorensen, P., Norheim, I. and Rasmussen, C. (2007) Simulation of the impact of wind power on the transient fault behavior of the Nordic power system, Electric Power System Research, Vol. 77, pp.135144.

Long-term stabilisation of grid connected wind farms with FACTS controllers 277
Kumar, N.S. and Khan, M.A. (2008) Impact of FACTS controllers on the dynamic stability of power systems connected with wind farms, International Journal of Wind Engineering, March, Vol. 32, No. 2, pp.115132. Kundur, P. (1994) Power System Stability and Control, McGraw Hill, New York. Mathur, R.M. and Varma, R.K. (2002) Thyristor-based FACTS Controllers for Electrical Transmission Systems, IEEE Press, Wiley and Sons Publications. Mithulananthan, N., Canizares, C.A. and Rogers, G.J. (2003) Comparison of PSS, SVC and STATCOM controllers for damping power system oscillations, IEEE Transactions on Power Systems, Vol. 8, No. 2, pp.786792. Samuelsson, O. and Lindahl, S. (2005) On speed stability, IEEE Transactions on Power Systems, Vol. 20, No. 2, pp.11791180. Varma, R.K. and Sidhu, T.S. (2006) Bibliographic review of FACTS and HVDC applications in wind power systems, International Journal of Emerging Electric Power Systems, Vol. 7, No. 3, Art. 7, pp.116. Xiu, L. and Wang, Y. (2007) Dynamic modeling and control of DFIG based wind turbines under unbalanced network conditions, IEEE Transactions on Power Systems, February, Vol. 22, No. 1, pp.314323.

Appendix A Parameters
Base values for the per unit system conversion: base power: 100 MVA base voltage: 0.69 KV for low voltage bus bar, 150 KV for high voltage bus bar.

B Doubly fed induction generator

Rated apparent power MVA: 2 MVA Rotor inertia: 3.527 MW s/MVA Rs (p.u.) = 0.0693 Xs (p.u.) = 0.080823 Rr (p.u.) = 0.00906 Xr (p.u) = 0.09935 Xm (p.u) = 3.29 Rotor speed for synchronisation: 0.9 p.u. Minimum rotor speed: 0.56 p.u. Maximum rotor speed: 1.122 p.u.

278

N.S. Kumar and M.A. Khan

C Transformers
Three winding transformer (150 KV: 0. 69 KV) Primary rated apparent power = 25 MVA Secondary rated apparent power = 25 MVA Tertiary rated apparent power = 6 MVA.