M.Sc.

Thesis Presentation

Travelling Wave Based DC Line Fault Location in VSC HVDC Systems

K.P.A.N. Pathirana

Department of ECE University of Manitoba Canada.

Outline

Introduction Surge detection method Modelling of Rogowski coil Line fault location performance Conclusion and future work

Background
HVDC transmission lines and cables need repairs quickly as possible after a fault. Travelling wave based fault location is the common fault location method applied in HVDC transmission lines. IGBT based voltage source converter (VSC) HVDC systems are gradually gaining ground.

Problem definition

No publications dealing with the fault location in VSC HVDC schemes with such long cable connections. The large DC capacitance at the converter terminal. Measurement bandwidth of the transducers.

Objectives
Development of a method of measurement for detecting travelling wave arrival times in a VSC HVDC scheme. Testing and verification of the proposed measurement system through simulations. Investigate the effect of different parameters on the accuracy of fault location.

Line fault location methods

Techniques based on impedance measurement Techniques based on high frequency spectrums of the currents and voltages Machine learning based approaches Techniques based on travelling waves

Line fault location methods

Techniques based on impedance measurement Techniques based on high frequency spectrums of the currents and voltages Machine learning based approaches Techniques based on travelling waves

Travelling wave based fault location

. ( ) = 2

Current LFL technology

Detection methods

Current LFL technology

Detection methods Time stamping

Current LFL technology

Detection methods Time stamping Typical accuracies

Line Termination in LCC and VSC Schemes

LCC HVDC

VSC HVDC

Travelling waves incident on junction

Travelling waves incident on junction

= =

, = . , 2 + +

, = . ,

Travelling waves incident on junction

= 1

, =

, = 1 + . ,

= +

Travelling waves incident on junction

=
2

= 1

Voltage magnitude

1.6 1.2 0.8 0.4 0 0 2 4

V(Xo,t) Vo(Xo,t)

, = 1 + . , =

= +

Time [S]

Travelling waves incident on junction

0

= -1

= +

, =

, = 1 + . ,

Travelling waves incident on junction

0

= -1

Voltage magnitude

1.6 1.2 0.8 0.4 0 0 2 4

V(Xo,t) Vo(Xo,t)

, =
8

, = 1 + .

= +

Time [S]

Test network

Terminal voltage

205

200

Voltage [kV]

195

190

No inductor
185 0.595 0.598 0.601 0.604

Time [S]

Solid P-G fault 70 km away from Converter-1

Terminal voltage
Gradual Change
205

200

Voltage [kV]

195

190

No inductor
185 0.595 0.598 0.601 0.604

Time [S]

Solid P-G fault 70 km away from Converter-1

Terminal voltage

205

200

Voltage [kV]

195

190

No inductor 1 mH inductor
185 0.595 0.598 0.601 0.604

Time [S]

Solid P-G fault 70 km away from Converter-1

Terminal Current

0.6

No inductor 1 mH inductor

Current [kA]

0.3

-0.3

-0.6 0.595

0.598

0.601

0.604

Time [S]

Solid P-G fault 70 km away from Converter-1

Terminal Current
Less sharp terminal Current
0.6

No inductor 1 mH inductor

Current [kA]

0.3

-0.3

-0.6 0.595

0.598

0.601

0.604

Time [S]

Solid P-G fault 70 km away from Converter-1

Problems with line voltage and current measurements

Transducers need to be installed at very high potentials.
Insulations requirements. Electrical isolation between sensor output and the data acquisition system. Bulky and expensive instrumentation.

Surge capacitor current

Rate of change of terminal voltage

0.01

Current [kA]

-0.01

-0.03

-0.05

No inductor 1 mH inductor
-0.07 0.6001 0.6003 0.6005 0.6007 0.6009

Time [S]

Solid P-G fault 70 km away from Converter-1

Rate of change of the surge capacitor current

Small effect on value of inductance
Rate of change of surge capacitor current

9000

No Inductor 1 mH 10 mh

4000

-1000

0.6004

0.6005

0.6006

Time [s]

Solid P-G fault 70 km away from Converter-1

Proposed termination

Inductor Converter side Cable Side Surge Capacitor Rogowski Coil

vr

Experimental results

Dorsey converter station LCC HVDC 500 kV 900 km Overhead line

0.5 H Converter side Cable Side

Inner radius Outer radius Resistance Self-Inductance Capacitance

260 mm 284 mm 468 3.5 mH 60.93 pF

55 nF Rogowski Coil

Mutual-Inductance 0.55 H
vr

Experimental results
6

Rogowski coil voltage [V]

5 4 3 2 1 0 -1 0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

(a)
x 10
-3

Time [ms]

Rogowski coil voltage [V]

1.5 1 0.5 0 -0.5 -1 -1.5 -2 -2.5 -3 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045

(b)

Time [ms]

Rogowski coil voltage for a fault 356 km away from Dorsey converter station.

Remarks
If there is no series inductor voltage or surge cap cannot be used Current can be used With series inductor voltage or surge cap can be used The value of the series inductor is not that important as long as it is above 1 mH.

Modelling of Rogowski Coil

(). cos . = () . () = . . 0 . (). cos()

Modelling of Rogowski Coil

() () () = = 0 . . .

() = 0. . . ()

Equivalent Circuit of Rogowski Coil

= () . . = . +

Parameters of the designed Rogowski coil

Inner radius Outer radius Number of Turns 51.37 mm 57.49 mm 870 measured calculated Resistance Self-Inductance Capacitance * 4 81 H 3.9 81 H 13 pF 0.093 H

Mutual-Inductance 0.093 H * Capacitance is too small to measure

Test setup

Verification of the Rogowski coil model

40 20

Current [A]

-20

Current Through the Rogowski coil

-40 -0.05 3 -0.01 0.03 0.07 0.11 0.15 0.19 0.23 0.27 0.31 0.35 0.39

Time [mS]

Voltage [V]

-1

Simulated Experimental
-3 -0.05 -0.01 0.03 0.07 0.11 0.15 0.19 0.23 0.27 0.31 0.35 0.39

Time [mS]

Verification of the Rogowski coil model

40 20

Current [A]

-20

Current Through the Rogowski coil

-40 -0.05 3 -0.01 0.03 0.07 0.11 0.15

Time [mS]

Voltage [V]

-1

Simulated Experimental
-3 -0.05 -0.01 0.03 0.07 0.11 0.15

Time [mS]

Line Fault Location Performance

Line Fault Location Performance

Terminal voltages and Currents

Positive pole
Voltage [kV]
200 -170 -190 -210 603 600 601 602 603

Negative pole
Voltage[kV]
Con. 1 Con. 2

180

Con. 1 Con. 2
601 602

160 600

(a) Time [mS]

1.5 1

(b) Time [mS]

Current[kA]

0.5 0 -0.5 -1 600 601 602

Current[kA]

0.5 0 -0.5 -1 -1.5 600 601 602

Con. 1 Con. 2
603

Con. 1 Con. 2
603

(c) Time [mS]

(d) Time [mS]

solid pole-to-ground fault on positive pole 130 km from Converter-1

Surge Capacitor currents and Rogowski coil Voltages

Positive pole
0.02 0.07

Negative pole
Current[kA]
Con. 1 Con. 2

Current[kA]

-0.01 -0.04 -0.07 600 1 0.5

0.04 0.01 -0.02 600 1

Con. 1 Con. 2
601 602 603

601

602

603

(e) Time [mS]

0.5 0 -0.5 -1 -1.5 600

(f) Time [mS] Con. 1 Con. 2

Voltage[V]

0 -0.5 -1 -1.5 -2 600 601 602

Con. 1 Con. 2
603

Voltage[V]

601

602

603

(g) Time [mS]

(h) Time [mS]

solid pole-to-ground fault on positive pole 130 km from Converter-1

Threshold setting

Threshold setting

Actual fault location (km) 30 50 130 160 230 260 visual inspection 0.233 0.721 0.578 -0.476 -0.327 -0.863

Fault location errors (km) Threshold 1 0.209 0.707 0.567 -0.394 -0.286 -0.807 Threshold 10 -0.209 0.326 0.453 -0.172 -0.019 -0.424 Threshold 25 0.097 0.123 0.193 -0.115 0.106 -0.165

Threshold setting and fault resistance

0.8

Error[km]

0.5

0 ohm 50 ohm 100 ohm

0.2

-0.1

10

15

20

25

30

35

40

Threshold

Solid fault 30km the Converter -1

Threshold setting and fault resistance

0.8

Error[km]

0.5

0 ohm 50 ohm 100 ohm

0.2

-0.1

10

15

20

25

30

35

40

Threshold

Solid fault 50km the Converter -1

Threshold setting and fault resistance

0.8

Error[km]

0.5

0 ohm 50 ohm 100 ohm

0.2

-0.1

10

15

20

25

30

35

40

Threshold

Solid fault 220km the Converter -1

Threshold setting and fault resistance/low Thresholds

0.8

Error[km]

0.5

0 ohm 50 ohm 100 ohm

0.2

-0.1

10

Threshold

Solid fault 220km the Converter -1

Possibilities of improving the accuracy

Modal Transform
Remove the coupling between conductors.

Filtering
Selecting frequency band.

Modal transform
= . = .

1 1 1 = 2 1 1

1 1 1 = 2 1 1 .

1 1 = 2 1

1 . 1

Fault Location errors /Modal transform

Actual fault location (km) No M.Trans. 30 50 130 160 230 260 Solid-Fault 0.209 0.707 0.567 -0.394 -0.286 -0.807 Fault location error (km) Mode 0 0.172 0.707 0.567 -0.467 -0.286 -0.807 Mode 1 0.209 0.707 0.567 -0.431 -0.286 -0.807

Fault Location errors /Modal transform

Actual fault location (km) No M.Trans. 30 50 130 160 230 260 100 Fault resistance -0.088 0.427 0.474 -0.182 -0.100 -0.499

Fault location error (km) Mode 0 -0.119 0.402 0.432 -0.179 -0.080 -0.508 Mode 1 -0.095 0.452 0.479 -0.404 -0.120 -0.527

filtered and unfiltered Rogowski coil voltages

No filter 100 kHz L.P.

0.5 0

Voltage [V]

-0.5 -1 -1.5 -2 600.6

600.7

600.8

600.9

601

601.1

601.2

Time [ms]

Solid P-G fault 130 km away from the Converter-1.

Line Fault Location Performance

0.5 0

No filter 100 kHz L.P.

Voltage [V]

-0.5

x 10
-1 -1.5

-3

No filter 100 kHz L.P.

Voltage [V]

-2 600.6

600.7

600.8

600.9

601

601.1

601.2

0 -0.5 -1 -1.5 -2 593.6

Time [ms]

594.6

595.6

596.6

597.6

598.6

599.6

600.6

Time [ms]

Solid P-G fault 130 km away from the Converter-1.

Fault location with filtered signals (Threshold-1/Solid fault)

Actual fault location (km) 30 50 130 160 230 260 No filter 1MHz 0.172 0.707 0.567 -0.394 -0.286 -0.807 0.161 0.641 0.510 -0.31 -0.278 -0.731 Fault location error (km) 500 kHz 100kHz 0.112 0.576 0.452 -0.190 -0.197 -0.619 -0.095 0.163 -0.004 -0.089 -0.203 -0.216 50kHz 0.071 0.114 0.030 -0.015 -0.011 -0.129 10kHz 0.159 -1.63 -1.121 1.164 52.462 67.948

Fault location with filtered signals (Threshold-1/100 )

Actual location (km) 30 50 130 160 230 260 fault No filter -0.088 0.427 0.474 -0.182 -0.100 -0.499 1MHz -0.136 0.362 0.38 -0.172 -0.056 -0.424 Fault location error (km) 500 kHz -0.117 0.359 0.182 -0.164 -0.068 -0.414 100kHz 0.058 0.206 -0.003 -0.071 -0.135 -0.204 50kHz 0.140 0.129 0.011 0.012 0.064 -0.152 10kHz 0.804 -2.195 -1.51 1.723 53.374 68.969

Fault location with filtered signals (Threshold-10/Solid fault)

Actual fault location (km) 30 50 130 160 230 260 No filter -0.209 0.326 0.453 -0.172 -0.019 -0.424 1MHz -0.221 0.297 0.176 -0.125 -0.011 -0.349 Fault location error (km) 500 kHz -0.165 0.258 0.015 -0.117 -0.024 -0.302 100kHz -0.038 0.041 0.03 -0.015 -0.048 -0.056 50kHz 0.040 0.057 0.019 0.009 0.034 -0.012 10kHz 0.373 -0.057 -0.305 0.03 -0.039 -0.333

Fault location with filtered signals (Threshold-10/100 )

Actual fault location (km) 30 50 130 160 230 260 No filter 0.031 -0.016 0.011 -0.097 0.028 -0.008 1MHz -0.016 -0.044 0.028 -0.050 0.035 0.030 Fault location error (km) 500 kHz 0.069 -0.021 -0.019 -0.045 0.003 0.013 100kHz -0.004 0.015 -0.019 -0.009 0.039 0.012 50kHz 0.051 0.012 0.002 0.035 0.133 0.094 10kHz 2.354 -7.37 -2.04 0.816 6.053 -2.886

Fault location errors with cable connection

0.1

Threshold 10/ 100kHz L.P

0.05

Error[km]

Solid fault

-0.05

-0.1 0 50 100 150 200 250 300

Distance to Fault fron Con.1

0.1

Threshold 10/ 100kHz L.P

0.05

Error[km]

100 fault resistance

-0.05

-0.1 0 50 100 150 200 250 300

Distance to Fault fron Con.1

VSC HVDC scheme with overhead lines

0.14

0 ohm 100 ohm

Error [km]

0.09

0.04

-0.01

10

20

30

40

50

60

70

80

Threshold

Solid fault 300km the Converter -1

VSC HVDC scheme with overhead lines

0 ohm 100 ohm

0.05

Error [km]

0.03

0.01

-0.01

10

20

30

40

50

60

70

80

Threshold

Solid fault 600km the Converter -1

VSC HVDC scheme with overhead lines

0.1

0 ohm 100 ohm

0.08

Error [km]

0.06

0.04

0.02

10

20

30

40

50

60

70

80

Threshold

Solid fault 100km the Converter -1

Fault location errors with overhead line

0.09

0 ohm
0.07

Error [km]

0.05

0.03

0.01

-0.01

200

400

600

800

1000

Actual fault location [km]

Fault location errors with overhead line

0.05

100 ohm

0.03

Error [km]
0.01 -0.01

200

400

600

800

1000

Actual fault location [km]

Remarks

Simulation results indicated that there is an optimum range of threshold settings. Accuracy improved by filtering the signal from Rogowski coil with a low pass filter with a cut-off frequency of 50-100 kHz.

Conclusions
Proposed termination enables successful detection of travelling waves in VSC HVDC schemes. Fault location accuracy can be improved by filtering and selecting a optimum threshold setting. Fault location accuracy of 250 m for a 1000 km overhead line or 300 km long cable in a VSC HVDC system with the proposed method.