You are on page 1of 8

Computer Technology and Application 4 (2013) 111-118

D

DAVID

PUBLISHING

Stability Improvement of Power System by Using PI & PD Controller
Habibur Rahman, Rafiqul Islam Sheikh and Harun-Or-Rashid
Department of EEE, Faculty of ECE, Rajshahi University of Engineering & Technology (RUET), Rajshahi 6204, Bangladesh Received: February 4, 2013 / Accepted: February, 15, 2013 / Published: February 25, 2013. Abstract: This paper presents the model of a SVC (Static VAR Compensator) which is controlled externally by a PI (Proportional Integral) & PD (Proportional Differential) controllers for the improvements of voltage stability and damping effect of an on line power system. Both controller parameters has been optimized by using Ziegler-Nichols close loop tuning method. Both single phase and three phase (L-L) faults have been considered in the research. In this paper, a power system network is considered which is simulated in the phasor simulation method & the network is simulated in four steps; without SVC, With SVC but no externally controlled, SVC with PI controller & SVC with PD controller. Simulation result shows that without SVC, the system parameters become unstable during faults. When SVC is imposed in the network, then system parameters become stable. Again, when SVC is controlled externally by PI & PD controllers, then system parameters becomes stable in faster way then without controller. It has been observed that the SVC ratings are only 50 MVA with controllers and 200 MVA without controllers. So, SVC with PI & PD controllers are more effective to enhance the voltage stability and increases power transmission capacity of a power system. The power system oscillations are also reduced with controllers in compared to that of without controllers. So with both controllers the system performance is greatly enhanced. Key words: SVC (Static VAR Compensator), PI (Proportional Integral), PD (Proportional Differential) Controller, AVR, TCR (Thyristor Controlled Reactor), voltage regulation, MATLAB Simulink.

1. Introduction
Power system stability improvements are very important for large scale system. The AC power transmission system has diverse limits, classified as static limits and dynamic limits [1-2]. Traditionally, fixed or mechanically switched shunt and series capacitors, reactors and synchronous generators were being used to enhance same types of stability augmentation [1]. For many reasons desired performance was being unable to achieve effectively. A SVC (Static VAR Compensator) is an electrical device for providing fast-acting reactive power compensation on high voltage transmission networks and it can contribute to improve the voltage profiles in the transient state and therefore, it can improve the qualities and performances of the electric services [3].
Corresponding author: Habibur Rahman, B.Sc. & M.Sc., research fields: power system, nano technology, communication. E-mail: habibieee@yahoo.com.

A SVC can be controlled externally by using properly designed different types of controllers which can improve voltage stability of a large scale power system. Authors also designed a PI controller [4-5] without properly tuned and system performances were investigated. With a view to getting better performance, PI & PD controller parameters has been designed Ziegler-Nichols method for SVC to injects Vqref externally. The dynamic nature of the SVC lies in the use of thyristor devices (e.g., GTO, IGCT) [5]. Therefore, thyristor based SVC with PI & PD controllers has been used to improve the performance of multi-machine power system.

2. Control Concept of SVC
A SVC is a controlled shunt susceptance (B) which injects reactive power (Qnet) into thereby increasing the bus voltage back to its net desired voltage level. If bus voltage increases, the SVC will inject less (or TCR will

If the thyristors are gated into conduction precisely at the peaks of the supply voltage. u (t ) = K p e(t) + In Laplace Form. many different types of tuning rules have been proposed in the literature [7]. and the current is the same as though the thyristor controller were short circuited. 1 [6]. 2 Steady state (V-I) characteristic of a SVC. From Fig. 4). 2 [6]. PI Controller Tuning Process The process of selecting the controller parameters to meet given performance specifications is called controller tuning. PI (Proportional Integral) controller parameters has been designed by Ziegler-Nichols close loop tuning method. V = Vref + Xs⋅I: In regulation range (-Bcmax < B < Bcmax) V = I/Bcmax: SVC is fully Capacitive (B = Bcmax) V = 1/Blmax: SVC is fully inductive (B = Blmax) Fig. Qnet.112 Stability Improvement of Power System By Using PI & PD Controller absorb more) reactive power. Kp Ti ∫ e(t )dt (1) For selecting the proper controller parameters. Thus the critical gain Kcr & the corresponding period Pcr are experimentally determined (see Fig. From V-I curve of SVC. is controlled by the magnitude of –Qind reactive power absorbed by the TCR. The basis of the TCR (Thyristor Controlled Reactor) which conduct on alternate half-cycles of the supply frequency.5 P s ⎟ ⎟ cr ⎠ ⎝ (3) (4) . At which the output first exhibits sustained oscillations (see Fig. Ziegler-Nichols close loop Tuning method [7] is described below. the parameter is selected at first in proportional action. Ziegler and Nichols suggested that the values of the parameters KpTi should set according to the following formula: Kp = 0. Fig. 4. ⎛ U(s) 1 ⎞ = Kp ⎜ + 1 ⎜ Ts⎟ ⎟ E(s) i ⎠ ⎝ (2) ⎛ 1 ⎞ Gc( s ) = K p ⎜ + 1 ⎜ Ts⎟ ⎟ i ⎠ ⎝ ⎛ 1 ⎞ 1 + Gc( s) = 0.6Kcr.5Pcr Notice that the PI controller parameters is tuned by this method of Ziegler-Nichols rules gives [7]. full conduction results in the reactor. (b) In VAR control mode (the SVC susceptance is kept constant). Most controllers are adjusted on-site. SVC based control system is shown in Fig. 3) only increase Kp from 0 to a critical value Kcr. 1). Here. therefore the magnitude of reactive power injected into the system. The PI controller has two term control signal [7]. 1 SVC based control system [6]. In this method. Ti = 0. SVC V-I Characteristics The SVC can be operated in two different modes: (a) In voltage regulation mode (the voltage is regulated within limits as explained below). +Qcap is a fixed capacitance value. So consider Ti = ∞.6K cr ⎜ ⎜ 0. 3. Td = 0. 4). Using the proportional controller action (see Fig. and the result will be to achieve the desired bus voltage (see Fig.

which can result in inducing electromechanical oscillations of the electrical generators. u (t ) = K p e(t) + K pTd de(t ) dt (5) (6) (7) U(s) = K p (1 + Td s ) E(s) Gc ( s ) = K p (1 + Td s ) Ziegler and Nichols suggested that the values of the parameters KpTd should set according to the following formula: Kp = 0. 6. 3 PI controller is in proportional action. Two types of faults: 6. The PI controller can be designed. Modeling of Power System with SVC This example described in this section illustrates modelling of a simple transmission system containing 2-hydraulic power plants. Any disturbances that occur in power systems due to fault. 6. By selecting proper values of Kcr & Pcr. Fig. PD (Proportional Derivative) controller parameters has also been designed by Ziegler-Nichols close loop tuning method. The remote 1000 MVA plant and a local generation of 5000 MVA (plant M2) feed the load. 5 Single line diagram of 2-machine power.6K cr (1 + 0. A single line diagram represents a simple 500 kV transmission system is shown in Fig. Another machine is swing generator. To maintain system stability after faults. the transmission line is shunt compensated at its centre by a 200MVAR SVC. PD Controller Tuning Process Similarly.1. . A 1000 MW hydraulic generation plant (M1) is connected to a load centre through a long 500 kV. The complete simulink model of this network is shown in Fig. Fig.To maintain system stability after faults. The phasor simulation method can be used.2 three phase L-L fault have been considered.125Pcr Thus. the transmission line is shunt compensated [8] at its centre by a 200MVAR SVC. PSS is used in the model to add damping to the rotor oscillations of the synchronous machine by controlling its excitation current. The PD controller has two term control signal [7]. Transfer Function Gc ( s ) = 0. 7).Stability Improvement of Power System By Using PI & PD Controller 113 Fig. 5.125Pcr s ) (8) By selecting proper values of Kcr & Pcr. 5. Machine & SVC parameters have been taken from [8]. excitation system. The PD controller can be designed. and PSS (Power System Stabilizer). total 40km transmission line. A load flow has been performed on this system with plant M1 generating 950 MW so that plant M2 produces 4046 MW. The SVC does not have any controller unit. 6. A 5000 MW of resistive load is modelled as the load centre.1 Single line to ground fault & 6. Such oscillating swings must be effectively damped to maintain the system stability and reduce the risk of stepping out of synchronism.1 Simulation Results (Without SVC) The load flow solution of the above system is calculated and the simulation results are shown below. SVC has been used to improve transient stability and power system oscillations damping. The two machines are equipped with a HTG (Hydraulic Turbine and Governor) (see Fig.1. 4 Determination of sustained oscillation (Pcr).6Kcr . The line carries 944 MW which is closed to its surge impedance loading (SIL = 977 MW). Td = 0.

06 0.01 0 -0.t 1 1.5 2 (a) 2000 1000 0 Power 2000 (b) 1500 Power .5 4 4.01 -0.8 kV/500 kV4 Regulators M 2 5 Line power 9 <Vm (pu)> <B (pu)> Vt 1 Vt2 PV Fig.5 1 time.04 Speed Deviation 0.01 0 0.6 0.5 0 0 0.t 0.9 0.02 S p e e d D e v ia tio n dw 0.8 0.p -1000 -2000 500 0 0 0.5 2 0 0.1 1 0.07 0.5 time.01 0 0 0.5 Bus Voltages 1 PSS.7 0.t 1 1.6 0.5 3 3.5 2 2.5 3 3.114 Stability Improvement of Power System By Using PI & PD Controller Vabc m m Pm Vf Pm Vf _ Vabc PQ 2 PQ 1 A B C a b c Pref1 1 Pref A B C i + - Vabc A Iabc B C a b c Iabc B2 aA bB cC CT 2 i + - PQ Vabc A Iabc B a b c Iabc3 + - 3 m a b c A B C A B C Pm Vf _ Pm Vf m Pref PQ1 i Pref2 1 Turbine & Regulators M 1 M1 1000 MVA 1000 MVA 13.5 time.05 0.8 1 0.5 time.5 0 0. HTG and excitation system block diagram for machine 1. 7 2 1.t 2 2.5 1 1.03 0. 1 d_theta 1 wref 2 Pref wref Pref we Pe0 gate dw Pm pm 1 1 Pm vs _qd d_theta 1 m m wm dw Peo Re Im mu |u| Vt 1 1 Vref 1 HTG vref vd Vf vq Machines Measurement Demux w1 dw 1 2 Vf dw Vs tab vs tab Multi -Band PSS EXCITATION Fig.02 0. 6 Complete simulink model of 2-machine power system.5 2 2.t (f) (e) .8 kV/500 kV powergui Phasors L1 350 km A B C L2 350 km Signals 6 w1 w2 w1 w2 C A B C V-I1 Fault Breaker SVC (Phasor Type ) V-I 8 7 Vt 1 Vt 2 V P SVC m Volt.P 1000 Power.2 0.5 Va Vb Vc 0.4 time.5 3 (c) dw (d) 0. B1 B2 B3 2000 MVA M2 5000 MVA Turbine & 13.t 1.5 1 1.5 time. Va Vb Vc Bus Voltages 1.

8a-8d.1. 8e) & power (P) becomes stable within 1s (see Fig. 7g) & machines speed deviation becomes stable within 5s (see Fig.u for 1-phase fault (without SVC).1 Single line to ground faults Consider a 1-phase fault occurred at 0. (g) Bus Voltage (Va) in p.1. 4) which is shown in Eq. ⎛ S + 10 ⎞ Gc ( s ) = 120⎜ ⎟ ⎝ S ⎠ 7. . 9. P in MW during fault (without SVC). 6).2 Three phase faults (line-line) During 3-phase faults. the system voltage.1s with 0% damping (see Fig. All results have been summarized in Table 1. 8h) & Power (P) becomes stable within 2s (see Fig. 7 (a) Bus voltages in p. 7. 8j). if PI is used as SVC controller then.1 Single line to ground faults During 1-phase faults. Two types of faults have been considered: 7.2 three phase L-L fault.1 Simulation Results (With PI Controller) (9) 7. after processing as shown in Fig.1. power all are The network remains same (see Fig. Without SVC.1s & circuit breaker is opened at 0. (c) Bus power. 8f) & Machines speed deviation becomes stable within 2. If no SVC is applied then system voltage & machines speed deviations becomes unstable But when SVC (without controller) is applied then the system voltage becomes stable within 5s (see Fig. (h) Machines speed deviation for L-L fault. line voltage. But if SVC (without controller) is applied then voltage becomes stable within 3s (see Fig. Thus the transfer function or parameters of PI controller is determined based on Ziegler-Nichols tuning method (Eq.5s (see Fig. 7a) is determined for this network (Kcr = 200) & corresponding period of Pcr (see Fig. if PI is used as SVC controller then. Design of PI Controller The proposed PI (Proportional Integral) controller parameters has deigned by using Ziegler-Nichols tuning method. 7a & found Pcr = 0. line current. During faults the machines angular speed deviation (dω) & mechanical power (Pm).1.2. 7. (f) Speed oscillations for 1-phase fault (with SVC).Stability Improvement of Power System By Using PI & PD Controller 115 (g) (h) Fig. All results have been summarized in Table 1. 7c and 7e). (b) Bus Voltages in p. 7h).5s (see Fig. During fault. 8g).u for L-L phase fault.1. power & machines oscillates goes on unstable (see Fig. 7d) & machines oscillation becomes stable within 4. 8b and 8d). The critical gain (Kcr) for which the plant output gives a sustained oscillation (see Fig. 6. power becomes within 3s (see Fig. the system voltage becomes stable within 3. (d) Bus Power (P) in MW for 1-Ø faults (with SVC). just simple SVC is replaced by PI controlled SVC (see Fig. PI reduces damping of power system oscillation. changed.2s (4-cycle fault). (e) Speed deviation for 1-phase fault (without SVC).1 Single line to ground fault and 7. The proposed PI controlled SVC simulink model is shown in Fig. 8i) & Machines speed deviation becomes stable within 4s (see Fig. the system voltage becomes stable within 2s with 0% damping (see Fig. 4) is also determined from Fig.1. Machines speed deviation (dω) & mechanical power deviation (pm) always monitored by PI controller & taking input of those oscillation. 6. 8a and 8c. 7b). 7f). 7a.u for 1-phase fault (with SVC). So dω & pm are taken as input parameters for improving the SVC performance of newly designed PI controller.2 Three phase faults (line-line) During 3-phase (L-L) faults.

5 (f) Va Vb Vc 1 B u s V o lta g e s 0.4 0.5 time. (e) Voltage (Va) in p.5 3 3. (h) Bus voltages in p.2 1 0.6 0.4 (d) Power 1.5 1 1.3 0.6 0. (d) PI controlled SVC Model with pm input.5 4 (i) (j) Fig.6 0.8 0.5 2 2.u for L-L faults (with PI).7 0 0.5 time.116 Stability Improvement of Power System By Using PI & PD Controller 2 w1 Vref s Gain Transfer Fcn 2 dw Saturation 1 5 A B PI COntroller (a) 1 pm 2 2 pm 1 120 Gain s+10 s Transfer Fcn 2 dw Saturation 1 1 Vqref (b) A B m 0. (b) PI controlled SVC Model with dω input.5 1 1.2 1.t 1.8 Vref pm 1 pm1 dw SVC C pm 2 pm2 Vqref 5 m w1 w1 dw SVC C 1 w2 s+10 120 1 Vqref w2 w2 Vqref SVC SVC PI COntroller 1.4 0.3 Va (c) 1.P 0.t 2 2.9 0.8 2 S peed Deviation Bus Power 2 0 -2 0 0.5 3 (g) (h) dw 4 3000 2000 1000 0 -1000 0 0. 8 (a) Internal Structure of PI controller with dω input.5 1 time. P o w e r.u for 1-phase fault (SVC with PI). (i) Bus power in MW for L-L faults (with PI).2 0.2 0.1 1 V o lta ge.8 time. (j) Machines speed deviation for L-L fault (with PI).2 1.t 0.t 1 1.6 1. (g) Speed deviation for 1.9 1 3 2 1 S p e e d D e v ia tio n 0 (e) dw 1.7 0.5 time.5 1 time.8 0.V a 0 0.4 1. .t 2 2.5 6 0 0.5 2 2. (f) Bus Power (P) in p.t 1.1 0.u for 1-phase fault (SVC with PI).phase fault (SVC with PI). (c) Internal Structure of PI controller with pm input.5 -1 -2 0 5000 Power 4000 0 0.

4 Va 1.t 0.u for 1-phase fault (SVC with PD).6 0. 4) is also determined from Fig.2 1.2.6 time.t 0.3 (g) (h) Fig. 6).Stability Improvement of Power System By Using PI & PD Controller 117 8. (f) Bus voltages in p. 9b) & the network is simulated in phasor simulation method.6 time.4 0. (d) Bus Power (P) in p.t 1 1. (e) Speed deviation for 1-phase fault (SVC with PD).1 Simulation Results (With PD Controller) (10) The network remains same (see Fig.Thus the parameters of PD controller is determined based on Ziegler-Nichols tuning method (Eq.5 3 .2 0. just simple SVC is replaced by PD controlled SVC (see Fig.4 0.8 1 1.2 0.2 three phase L-L fault.2 1000 1 800 `Voltage 0.5 1 time.2 Va Vb Vc 0 0 0.t 1 1. (b) PD controlled SVC Model with dω input. (c) Voltage (Va) in p.u for 1-phase fault (with PD). 1 w2 2 w1 120 Gain s+40 40 s Transfer Fcn 2 dw Saturation 1 1 Vqref ⎛ S + 40 ⎞ Gc ( s ) = 120 ⎜ ⎟ ⎝ 40 ⎠ 8.5 0. The critical gain (Kcr = 200) & oscillation period of Pcr (see Fig. 8) which is shown in Eq.4 1.4 (e) 2000 Bus Power 1500 5 Bus Power Speed Deviation 1000 10 (f) dw 0 500 0 0 0.7 0.2 0.8 -5 0 0. 9a-9b. The proposed PD controlled SVC simulink model is shown in Fig.6 5 4 3 2 Speed Deviation (c) 1.5 -1 -2 0 0. (g) Bus power in MW for L-L faults (with PD).2 1. 10.5 2 2.6 0.u for L-L faults (with PD).2 0.8 1 1. 7a & found Pcr = 0.5 time. Two types of faults has been considered: 8.8 time.t 0. 0.6 0. 9 (a) Internal Structure of PD controller.4 time.1.4 600 1200 (b) Power 400 0 0.1.8 Power 0. Design of PD Controller The proposed PD (Proportional Derivative) controller parameters has also been deigned by using Ziegler-Nichols tuning method. w2 w1 w2 w1 Vqref dw A Vref 5 B SVC m C SVC PD Controller (a) 1.t 1.5 dw (d) 1 Bus Voltages 1 0 0.5 0 0.1 Single line to ground fault and 8.4 0. (h) Machines speed deviation for L-L fault (with PD).1 0.

e. Rahman. STATCOM.4s 0. H.I. L-L fault (stability time) volt P dω α inf α inf α inf 5s 5s 5s 3. From above results. Garg. UPFC which may be controlled externally by designing different types of controllers for stability improvement of a power system. 9d) & Machines speed deviation becomes stable within 1.com.1s 2s 4s 1. then a small rating SVC can [4] [5] [6] [7] [8] . Online voltage level improvement by using SVC & PSS. Voltage control and dynamic performance of power transmission system using SVC.K. machine oscillation damping.5s SVC+PI 50MVA 2s 1s 2.6s 1. if PD is used as SVC controller then. A. IEEE PWRS 1 (4) (1986) 222-227. Rahman.5s SVC+PD 50MVA 1.6s 0.1 Single Line to Ground Faults During 1-phase faults. The proposed controller can take as input only change of machines angular speed deviation & mechanical power deviation. 2012. H. H. 4th ed. α (infinite time) means the system is unstable. pp. S.1. PD controller may be highly suitable.3s (see Fig. SVC with PI & SVC with PD controller. in: International Conference on Electrical. Yousef.O. K.R. 8. current. Rashid. Sheikh.5s 1.62s (see Fig. the system voltage becomes stable within 1.E. mechanical power deviation etc& taken those parameters as input & after internally processing it injects Vqref into SVC for stability improvements. Conclusions This paper presents the stability improvement of voltage level. 2008. 9g) & Machines speed deviation becomes stable within 3s (see Fig. International Journal of system & Simulation 6 (2) (2012). the system voltage becomes stable within 1. But a new controller can be designed which always monitor every network parameters i. 9.5s with 0% damping (see Fig. without SVC. Sabai.O. Computer & Telecommunication Engineering (ICECTE-2012). Journal of Engineering Sciences 32 (2) (2004) 665-677. A. RUET. by the Math Works Inc.mathworks. H. Agarwal. The network is simulated in four steps. [2] [3] 10. Ogata. Both are very cheapest & efficient controllers for SVC to enhance the power system stability. Prince Hall. With SVC. 2002. Rajshahi. 9e). Win. References [1] A. 9h). Advantages of those controllers is that only a small controller can handle a robust interconnected power system efficiently. Rahman. http://www. if a externally (PI & PD) controlled SVC is connected in the network. Bangladesh. Another FACTS devices namely SSSC.. M.. Engineering and Technology 18. if PD is used as SVC controller then. Rashid. Modern Control Engineering. Compared to those two controllers. December. Transient stability enhancement of multi machine using global deviation PSS. Modeling and Simulation of Static VAR Compensator for Improvement of Voltage Stability in Power System.6s 1-Øfault (stability time) Controller SVC rating volt P dω No SVC 200MVA α inf α inf α inf SVC 200MVA 3s 3s 4. International Journal of Electronics Communication and Computer Engineering 2 (2) (2011) 102-106. For stability improvements. T.118 Stability Improvement of Power System By Using PI & PD Controller Table 1 Performance Comparison of PI & PD Controllers. N. bus power in a large scale power system model of SVC with or without properly tuned PI & PD controllers for different types of faulted conditions. MATLAB math library user’s guide [Online]. Results & Discussions The performance of both proposed PI & PD controller of power system network with SVC has been summarized in the Table 1. 9f) & Power (P) becomes stable within 0.4s with 0% damping (see Fig.2 Three phase faults (line-line) During 3-phase(L-L) faults. this proposed PI & PD controllers which is properly & easily tuned by using Ziegler-Nichols tuning method may be highly suitable as SVC controller. In Table 1. change of bus voltage. Power system stability improvement by using SVC with PI & PD controller.3s 8. Analysis of power system stability enhancement by static var compensator. F. 425-429. Hammad.6s (see Fig. damp out all types of oscillation & the system becomes stable in faster way. 9c) & power (P) becomes stable within 1. machines angular speed deviation. All results have been summarized in Table 1.M. World Academy of Science.1.