Presentation Paper CEPSI 2006 Category of Paper: T-2 Transmission Systems Title of Paper: Characterization of Dynamic Phenomena on EGAT

T Tie Transmission Line

Caused by Major Disturbance
Keywords: Dynamic Phenomena, Power System Disturbance, Power System Characteristic, Phasor Measurement Unit Tormit Junrussameevilai Author: Power System Analysis Department, Control and Protection System Division Electricity Generating Authority of Thailand, Thailand email: tormit.j@egat.co.th Nitus Voraphonpiput Power System Analysis Department, Control and Protection System Division Electricity Generating Authority of Thailand, Thailand email: nitus.v@egat.co.th

Abstract: This paper aims to present dynamic phenomena under major disturbances of EGAT power system. The EGAT power system can be considered as a remote system connected to a main system via a long tie transmission line. Dynamic phenomena on the tie transmission line are observed using two Phasor Measurement Units (PMU) that were installed at the sending end bus and receiving end bus where the tie transmission line is connected. Dynamic phenomena on the tie transmission line were observed during one year and were selected to present in this paper. Static Var Compensator (SVC) and HVDC were installed in the EGAT power system to increase transient and dyamic stability. These devices have enhanced stability function, which were designed to maintain stability and cecurity of the remote system. Reactions of the SVC and HVDC under several disturbances were evaluated from PMU recorder. The PMUs offers accurate measurement of voltage and current phasors. Moreover, phase angle between two power systems are also calculated. These features allow engineers to analyses dynamic phenomena of the power system. Particular disturbance has special dynamic characteristic which are analyzed in the paper.

1. Introduction Thailand transmission grid is owned and operated by Electricity Generating Authority of Thailand (EGAT). It consists of a 50 Hz network with 500 kV, 230 kV, 115 kV transmission voltage levels. Thailand power system has generation about 26,459 MW and maximum demands of 20,744 MW. The EGAT power system can be considered as two systems one of which is a remote system (Southern area) and the remainder is a large system (Central area). Two areas are tied together via a 230 kV transmission line, which comprise two circuits 1272 MCM with 233.92 km long. Installed capacity of the Southern area is about 1,720 MW, which consists of thermal power plant, hydro power plant, and gas turbine power plant. There is also a 300 MW High Voltage Direct Current (HVDC) was installed at the lower part of the EGAT transmission grid. It was installed at the KHLONG NGAE (KNE) substation. This HVDC is an isolated link between Thailand transmission grid and Malaysia transmission grid. The maximum demands of the Southern area approximately 1,760 MW. Therefore, the tie transmission line pays an important role in the EGAT power system. It transfers energy from the Central area to the Southern area according to economic operation of the power system. Moreover, the southern area can import and export electric energy from Malaysia via The HVDC interconnection for security and stability of both transmission grids, and commercial purpose.

This SVC increases transfer capability of the tie transmission line. The Southern transmission grid is shown in Fig.1

Central Area

PMUs

Tie Transmission Line

Southern Area

Fig.1: The Southern transmission grid In order to maintain system security and stability, the enhanced control function of SVC is used to prevent

A +300 to -50 Mvar Static Var Compensator (SVC) was installed at the 230 kV intermediate substation (Bang Saphan) closed to the central area.

power oscillation on the tie transmission line. Moreover, the enhanced control function of the HVDC such as Runup, Run-back, and Frequency Limit Control (FLC) were implemented. Performance of the enhanced control functions could be observed by the Phasor Measurement Units (PMUs), which were installed at 230 kV Bang Saphan (BSP) and 230 kV Suratthani (SRT) substation in early 2005 AD. These PMUs offer voltage, current phasors and system frequency of the Central and the Southern areas with sampling data of 100 ms. Therefore, the PMU data could be used to evaluate the enhanced control functions and dynamic phenomenon of the power system on the tie transmission line. 2. System Disturbance due to the HVDC Tripped The HVDC received electric power of 300 MW from Malaysia transmission grid. The tie transmission line transferred real and reactive power from the Central system about +18 MW and -24 Mvar respectively. After that, the HVDC was blocked. Power flow of the tie transmission line increased from +18 MW to 460 MW with in 0.9 second and then oscillated with frequency of 0.52 Hz. The oscillation took approximately 13 seconds to settle. The final value of the real and reactive power flow at BSP were 300 MW and -17 Mvar. Dynamic phenomena of the real and reactive power flow of the tie transmission line are shown in Fig.2.
BSP Real Power Flow [MW] 500 BSP Reactive Power Flow [Mvar] RT Real Power Flow [MW] S SRT Reactive Power Flow [

time was 166.7 MW per 0.1 Hz. Response of the Central system frequency and the Southern system frequency were shown in Fig.3 and close look in Fig.4.
BSP Frequency 50.05 SRT Frequen

50.00

114.4 s (1.9 min)

50.0 Hz

49.95

49.9 Hz
49.90

49.85

49.81 Hz
49.80 350 400 450 Time in Seconds 500 550

r Characterisation of Large Disturbance\HVDC-Trip Feb-24-2006 2319-23

Fig.3: The Central and the Southern systems frequency

BSP Frequency 50.05 SRT Frequen

50.00

11.5 s SRT Frequency

50.0 Hz

49.95

49.90

BSP Frequency

49.9 Hz

49.85

400

BSP Real Power 300 MW 100 MW BSP Reactive Power

365 370 375 380 385 390 395 400

49.81 Hz
49.80 360 365 370 375 Time in Seconds 380 385 390

300
r Characterisation of Large Disturbance\HVDC-Trip Feb-24-2006 2319-23

200

100

10 s

0 360 -100

Fig.4: The Central and the Southern systems frequency (close look) Fig.5 presents phase angle difference between the Central system and the Southern system. The Central system is assigned as a reference bus. Therefore, positive angle value means power flow from the Southern system to the Central system. It could be seen that the phase angle swung and settled within 13 seconds.
Phase Ang 5

SRT Reactive Power SRT Real Power

Time in Seconds

-200

-300

-400

r Characterisation of Large Disturbance\HVDC-Trip Feb-24-2006 2319-23

Fig.2: Real and Reactive Power Flow of the Tie Transmission Line (230 kV BSP-SRT) System frequency of the EGAT transmission grid was maintained at 49.99 Hz. The Central system frequency and the Southern system frequency were equal. After the loss of the HVDC, the Southern system frequency decrease from 49.99 Hz to 49.89 Hz within 0.7 seconds and then swing back to 50.04 Hz. The Southern system frequency dropped together with the Central system frequency and swung with frequency of 0.52 Hz. The Central system frequency dropped to 49.81 Hz within 11.5 seconds and then recovered back to 50.0 Hz in 1.9 minutes due to response of an Automatic Generation Control (AGC). It might be said that frequency bias of the EGAT transmission grid at that

0.0
0 360 365 370 375 380 385 390 395 400

-5

-10

-15.0
-15

-20

10 s
-25 Time in Seconds

r Characterisation of Large Disturbance\HVDC-Trip Feb-24-2006 2319-23

Fig.5: Phase Angle difference between two systems

The +300 to -50 Mvar SVC was installed at the 230 kV BSP substation. Its main function is to regulate voltage at this substation. The addition function is power oscillation damping (POD). The POD will activate when the power oscillation occurs in the tie transmission line with an inter-area mode. The output of the POD function is modulated with the voltage reference signal to correct the power oscillation on the tie transmission line. According to the loss of the HVDC, the POD function was activated and generated a modulation signal to the voltage reference signal. Action of the POD resulted in voltage fluctuation at the bus where the SVC is connected (BSP bus). The voltage fluctuation at BSP leaded to mitigation of the power oscillation on the tie transmission line. Response of the BSP voltage and SRT voltage were illustrated in Fig.6.
BSP Voltage [kV] 246 SRT Voltage [k

then swung back to -60 MW. The power oscillation decayed and settled at 74.5 MW within 21.6 seconds. The BSP reactive power also swung from -9 Mvar to +13 Mvar and then turned back to -0.7 Mvar. It settled at 4.7 Mvar within 21.6 second. These phenomena were presented in Fig.7.
BSP Real Power Line 1 [MW] 20 BSP Reactive Power Line 1 [Mvar] BSP Real Power Line 2 [MW] BSP Reactive Power Line 2

Reactive Power Line No.1

0 30 35 40 45 50 55

10 s
60 65 70

-20

Power Line No.2 36.2 MW 20 MW

-40

21.6 s
-60

Power Line No.1

74.5 MW

-80

244

242 kV

POD Action of the SVC BSP Voltage

95.2 MW
-100 Time in Seconds

242

240

er Characterisation of Large Disturbance\BSP-SRT-circuit2_trip_Sep-6-0

238

236

SRT Voltage 234 kV

Fig.7: BSP Real and Reactive Power during the Circuit No.2 Tripped In Fig.8 presented dynamic phenomena during restoration of the circuit No.2. The circuit No.2 was energized after the fault was cleared. Firstly, the BSP CB was closed and then the SRT CB was closed at 12.7 second after that. When the BSP CB closed, BSP bus received -52.5 Mvar from VAR charging of the circuit No.2. At the time of the SRT CB was closed, the real and reactive power oscillated again but the oscillation frequency was 0.5 Hz and took 13.1 second to settle. The real and reactive power of both lines became equal again. The power oscillation records confirmed that the power system has better system damping when the tie transmission line was two circuits in service. Namely, the power oscillation in case of two circuits in service decayed and finished faster than the case of one circuit in operation.
BSP Real Power Line 1 [MW] 10 BSP Reactive Power Line 1 [Mvar] SP Real Power Line 2 [MW] B BSP Reactive Power Line 2

234

232

230

10 s
365 370 375 380 Time in Seconds 385 390 395 400

228 360

r Characterisation of Large Disturbance\HVDC-Trip Feb-24-2006 2319-23

Fig.6: Voltage Magnitude Measured at BSP and SRT The BSP voltage was 242 kV at the beginning whereas the voltage reference of the SVC was set at 240 kV. At the time of the incident, the BSP voltage started to swing from 242 kV to 234 kV and then turned up to 246 kV. The SRT voltage began at 237 kV and then also swung following the BSP voltage. Since the tie transmission line transferred more real power flow, the BSP and SRT voltage dropped to 240 kV and 234 kV at the final respectively. From the PMU records, they were also confirmed that the POD of the SVC could help the EGAT transmission system to damp the inter-area oscillation.

Reactive Power Line No.1

0 260 -10 265 270 275 280 285

Power Line No.2

290 295 300

3. System Disturbance due to One Tie Transmission Line Tripped The tie transmission line consists of two circuits of 1272 MCM ACSR conductors. The tie transmission line carried real and reactive power about -35 MW and -9 Mvar respectively (flowed from the Southern system to the Central system). The fault occurred on a conductor phase A of circuit No.2. and leaded to tripping of the tie transmission line circuit No.2. The real and reactive power on circuit No.2 transferred to circuit No.1. The real and reactive power on the tie transmission line oscillated with oscillation frequency of 0.42 Hz. The BSP real power swung from -36.2 MW to -95.2 MW and

12.3 Mvar 10 MW Reactive Power Line No.2 12.7 s -26.7 MW

-20

-30

-40

-52.5 Mvar
-50 -60

13.1 s

Power Line No.1

Time in Seconds

10 s

-70

er Characterisation of Large Disturbance\BSP-SRT-circuit2_trip_Sep-6-0

Fig.8: BSP Real and Reactive Power during the Circuit No.2 Closed

The BSP and SRT voltages during the circuit No.2 tripped were presented in Fig.9. At the time of the line tripped, the POD of SVC was not activated even the power oscillation occurred on circuit No.1. It might be concluded that the performance of the POD was reduced when the tie transmission line was not two circuits in operation. The BSP voltage behavior in Fig. 10 also confirmed that the POD of the SVC could perform when the circuit No.2 was in service. These incidents lead to modification of the POD function of the SVC in feature.
BSP Voltage [kV] 246 SRT Voltage [k

244

242 kV
242

240

238

236 kV
236

234

10 s
232 30 35 40 45 50 Time in Seconds 55 60 65 70

deviation of 150 MW. The FLC function was described in details in [1]. Fig.11 presents real and reactive power flow at BSP and SRT during the FLC was activated. The large power plant in Malaysia transmission grid was tripped and system frequency dropped over the acceptable limit. Therefore, the FLC was activated by an activation signal from HVDC converter station at Gurun (Malaysia). The HVDC started to draw the DC power from EGAT transmission grid. The HVDC drew the real power about 130 MW. The real power flow on the tie transmission line increased from 35 MW to 170 MW within 2 seconds. It could be seen that there was a small oscillation with frequency of 0.58 Hz. The tie transmission line flow was recovered back to the previous flow (35 MW) in 10 second. The real power swing occurred after FLC action with magnitude of 14 MW. It could be observed that the real power flow of BSP and SRT has large difference because the tie transmission line had changed configuration. It was sectioned to LANG SUAN (LSN) substation. The tie transmission line is now called 230 kV BSP-LSN-SRT rather than the 230 kV BSP-SRT.
BSP Real Power Flow [MW] BSP Reactive Power Flow [Mvar] RT Real Power Flow [MW] S SRT Reactive Power Flow [

er Characterisation of Large Disturbance\BSP-SRT-circuit2_trip_Sep-6-0

Fig.9: Voltage Magnitude Measured at BSP and SRT during Circuit No.2 Tripped
BSP Voltage [kV] 244 SRT Voltage [k

200

170 MW
150

100

242 kV
242 50

10 s BSP Real Power SRT Real Power BSP Reactive Power

325 330 335

240

0 300 305 310 315 320

340

238

POD Action
-50

SRT Reactive Power

Time in Seconds

236

234 kV
234

-100

232

r Characterisation of Large Disturbance\HVDC FLC Activate at Gurun Aug-25-

10 s
265 270 275 280 Time in Seconds 285 290 295 300

230 260

Fig.11: Real and Reactive Power Flow of the Tie Transmission Line (230 kV BSP-LSN-SRT)
BSP Frequency SRT Frequen

er Characterisation of Large Disturbance\BSP-SRT-circuit2_trip_Sep-6-0

Fig.10: Voltage Magnitude Measured at BSP and SRT during Circuit No.2 Closed

49.98

10 s
49.96 49.94

4. System Disturbance due to Frequency Limit Control (FLC) of the HVDC Action HVDC can offer fast DC power response. Thus, it is utilized for frequency stabilization of the systems. The FLC will be activated when the system frequency is over an acceptable band. It will increase or decrease the DC power transfer regarding a real power demand of the system to correct frequency excursion. Action of the FLC can consider as a disturbance of the system because the real power of the power system can be increased or decreased very fast with a maximum

BSP Frequency
49.92

49.92 Hz

49.90

49.88

49.86

SRT Frequency
305 310 315

49.85 Hz
320 Time in Seconds 325 330 335 340

49.84 300

r Characterisation of Large Disturbance\HVDC FLC Activate at Gurun Aug-25-

Fig.12: The Central and the Southern systems frequency

The system frequency of the Central system and Southern system are presented in Fig.12. Since the real power was drown by the HVDC, the system frequency of the Central system and the Southern system reduced. 0.07 Hz within 5.7 seconds and then the system frequency was recovered by AGC function. It is also observed that the Southern system frequency had an oscillation frequency of 0.58 Hz. Power flow deviation on the tie transmission line has difference in characteristic. It did not activate the POD function of the SVC even there was the power oscillation during the FLC action. Hence, the BSP voltage was regulated by the SVC and the SRT voltage had small dropped due to reactive power change, which were presented in Fig.13.
BSP Voltage [kV] 245 SRT Voltage [k

few second and then increased to 255.2 kV and 241 kV respectively. The BSP and SRT voltage behaved like step response of a first order delay system.
BSP Real Power Flow [MW] 150 BSP Reactive Power Flow [Mvar] RT Real Power Flow [MW] S SRT Reactive Power Flow [

100

Energized 500 kV SRT Real Power

100 MW

50

BSP Reactive Power

0 250 300 350

BSP Real Power

400 450

-50

-100

-100 MW 50 s SRT Reactive Power

Time in Seconds

-150

r Characterisation of Large Disturbance\OverVoltage_Jan-03-2006 1610-16

242kV
242

BSP Voltage
50.20

Fig.14: Real and Reactive Power Flow of the Tie Transmission Line (230 kV BSP-LSN-SRT)
BSP Frequency SRT Frequen

239

236 kV
236

Southern system frequency 50 s SRT Voltage

50.15

233

50.1 Hz
50.10

10 s
230 300 305 310 315 320 Time in Seconds 325 330 335 340

50.05

r Characterisation of Large Disturbance\HVDC FLC Activate at Gurun Aug-25-

50.0 Hz
50.00

Fig.13: Voltage Magnitude Measured at BSP and SRT

49.95 250 300 350 Time in Seconds 400 450

5. System Disturbance due to Energized a 500 kV Line BSP2-CBG To support demands in the feature, the 500 kV line BSP2-CBG was constructed. The 500 kV line connected to the BSP substation via a tie transformer and the other end was connected to the CBG substation, which located in the Central area. When the 500 kV line BSP2-CBG was energized from BSP2. The VAR charging from the line fed into the BSP substation and resulted in temporary over voltage at BSP substation. Phenomena of the VAR charging were described in Fig. 14. In Fig.14, the real and reactive power was presented. At the time of the energization, BSP and SRT real powers were not much change. The VAR charge injected into the BSP substation increased voltage at this bus and resulted in reactive power flow from BSP to SRT. It can be seen that the reactive power flow at BSP changed direction from -23 Mvar to 17.7 Mvar. The reactive power flow at SRT also reduced from -73 Mvar to -114 Mvar. The reactive power had slow increased. It took approximately 25 seconds to settle. The Central system frequency and the Southern system frequency were shown in Fig. 15. At the time of the energization, frequency was not changed because there was no real power change. The BSP and SRT voltage were shown in Fig. 16. The BSP and SRT voltage decrease in the first

r Characterisation of Large Disturbance\OverVoltage_Jan-03-2006 1610-16

Fig.15: The Central and the Southern systems frequency

BSP Voltage [kV] 260 SRT Voltage [k

255

BSP Voltage 250 kV

250

245 kV
245

240

SRT Voltage 50 s

235 250 300 350 Time in Seconds 400 450

r Characterisation of Large Disturbance\OverVoltage_Jan-03-2006 1610-16

Fig.16: Voltage Magnitude Measured at BSP and SRT When the 500 kV line was complete in service. The real power flows oscillated due to a configuration change of the Central system. It resulted in real power flows changes. Moreover, the reactive power flows of the tie line recovered close to the previous value. Both system frequencies were also oscillated, which could see in

Fig.15. Since the 500 kV line was completed in service, the BSP and SRT voltage were also reduced. However, the voltages were still high and did not recover close to the previous value. The POD action of SVC was not found even the power oscillation occurred on the tie transmission at the time of the 500 kV line energize completion. The reason of this incident might be limitation of the SVC. Namely, The BSP voltage was too high and Thyristor Controlled Reactor (TCR) of the SVC operated at full capacity. Hence, the POD could not perform at that time. 6. System Disturbance due to 80 MW load Transferring to Malaysia Transmission Grid The EGAT transmission grid can import electricity from The Malaysia transmission grid (TNB) via AC system and DC system. The AC system could not perform a synchronous link because Thailand and Malaysia transmission grids have separate control centre and Area Control Error (ACE) function did not construct. If both AC systems are connected, the power system will be losing system stability. Therefore, the EGAT transmission grid can import electricity by transfer partial load about 80 MW at SADAO (SDO) 115 kV substation to the Malaysia transmission grid. To prevent electricity interruption, it is necessary to do synchronous link in a few second during load transferring. Both systems will connect via the 115 kV AC transmission line and then the EGAT transmission grid will cut the partial load. The partial load is now connected to the Malaysia transmission grid.
BSP Real Power Flow [MW] 120 BSP Reactive Power Flow [Mvar] RT Real Power Flow [MW] S SRT Reactive Power Flow [

observed that the Southern system frequency began oscillate and trended to apart from the Central system even there were not change in real power. It could be confirmed that the synchronous link could not still operate.
BSP Frequency 50.05 SRT Frequen

50.04

50.03

10 s

Southern system frequency

50.02

50.01

50.0 Hz Central system frequency 49.97 Hz Frequency trend to apart

60 65 70 75 80 Time in Seconds 85 90 95 100

50.00

49.99

49.98

49.97

49.96

49.95

Characterisation of Large Disturbance\Import AC from TNB April 24-2006 0839-08

Fig.18: The Central and the Southern systems frequency during Load Transferring
Phase Angle d 3

10 s
1

0.0
60 65 70 75 80 85 90 95 100

-1

-2

-3

100

10 s 80 MW SRT Real Power 20 MW

-4.0

-4

80

-5

Time in Seconds

60 Characterisation of Large Disturbance\Import AC from TNB April 24-2006 0839-08

40

20

BSP Real Power

70 75 80 85 90 95 100

0 60 -20 65

Fig.19: Phase Angle Difference between BSP and SRT during Load Transferring
BSP Voltage [kV] SRT Voltage [k

BSP Reactive Power SRT Reactive Power

Time in Seconds 241

-40

-60

POD Action 240 kV

240

-80

Synchronous Link between EGAT-TNB

Characterisation of Large Disturbance\Import AC from TNB April 24-2006 0839-08

BSP Voltage
239

Fig.17: Real and Reactive Power Flow of the Tie Transmission Line during Load Transferring The dynamic phenomena of the load transferring were observed by the PMUs at BSP and SRT. Fig.17 was real and reactive power flow measured at BAP and SRT. The BSP real power reduced from 110 MW to 25 MW and then increased from 25 MW to 110 MW again. It could imagine that if the synchronous link between EGAT and TNB were still performed, the tie transmission line will oscillate. The Central system frequency and the Southern system frequency were presented in Fig.18. It could be

238 kV
238

SRT Voltage
237

10 s
236 60 65 70 75 80 Time in Seconds 85 90 95 100

Characterisation of Large Disturbance\Import AC from TNB April 24-2006 0839-08

Fig.20: BSP and SRT Voltage during Load Transferring It was also observed from Fig.19 that the phase angle between the Central system and Southern system

trended to unstable. The Southern system was very close to Malaysia. Therefore, the southern system frequency was affected by the Malaysia transmission grid. The BSP and SRT voltage were shown in Fig.20. It could be observed that the BSP voltage was not much change because the SVC regulated voltage at this bus. The SRT voltage had small fluctuation. The POD of the SVC was not activated because the real power on the tie transmission line was not a power oscillation. It was activated at the end of load transferring. It could be considered as EGAT transmission grid sudden lost load about 70 MW. 7. CONCLUSION This paper presented five categories of the large disturbance in the EGAT transmission grid, which consist of the 300 MW HVDC tripped, one circuit of the tie transmission line tripped, Frequency Limit Control action, large reactive power impact, and dynamic phenomena of the power system during synchronization of two large electric power systems. The PMUs records were utilized to characterize dynamic phenomena of the five disturbances. The observation on the tie transmission line offered better understanding on dynamic phenomena in a large electric power system. Dynamic behaviours of the power system during the enhanced control function were investigated. The observation results leaded to investigation of the SVC and HVDC performance. Moreover, this work could be used to confirm power system studies and adviced system operator in order to prevent system security and stability. Reference: [1] Nitus Voraphonpiput, Kittipon Chuangaroon, and Somchai chatrattana, Parameter Optimization for Frequency Limit Controller of EGAT-TNB HVDC Interconnection System Conference of Electric Supply Industry 2004 (CEPSI 2004)