Fault Analysis in HVDC Systems Using Signal Processing

  • Slides: 14
Download presentation
Fault Analysis in HVDC Systems Using Signal Processing Techniques Benish Paily School of Electrical

Fault Analysis in HVDC Systems Using Signal Processing Techniques Benish Paily School of Electrical and Electronic Engineering You Supervisors’ Names Here Dr. Malabika Basu Dr. Michael Conlon 29 th November, 2013

Presentation Overview 1 2 3 • HVDC - Next Dimension • HVDC Projects •

Presentation Overview 1 2 3 • HVDC - Next Dimension • HVDC Projects • Fault Analysis of HVDC System 2

HVDC-Next Dimension 1. Current Link HVDC System 2. Voltage Link HVDC System • LCC

HVDC-Next Dimension 1. Current Link HVDC System 2. Voltage Link HVDC System • LCC HVDC 1980 • VSC HVDC • MULTITERMINAL HVDC 1954 FUTURE Total 200, 000 MW HVDC Transmission Capacity in operation or under construction 170 Projects around the world ABB: 90 Projects, 120, 000 MW HVDC Light Technology Siemens: HVDC Plus Technology 3

HVDC Projects in Europe (ABB) East West Interconnector (HVDC Light)2013, connecting Ireland Wales 1.

HVDC Projects in Europe (ABB) East West Interconnector (HVDC Light)2013, connecting Ireland Wales 1. 2. 3. 4. 5. 6. 4 Link between Ireland Wales Power rating: 500 MW AC Voltage: 400 k. V DC Voltage: ± 200 k. V DC Underground cable: 2*75 km DC Submarine cable: 2*186 km

Fault Analysis of HVDC Systems HVDC system subjected to DC Line Fault HVDC system

Fault Analysis of HVDC Systems HVDC system subjected to DC Line Fault HVDC system subjected to AC line fault at Rectifier & Inverter side 1. Single line to ground fault (SLG) 1. Line to ground fault 2. Line to line fault (LL) 2. Line to line fault 3. Double line to ground fault (DLG) 4. Triple line to fault (3 L) 5

Faults in HVDC System DC Faults AC Faults DC Current increases to 2. 2

Faults in HVDC System DC Faults AC Faults DC Current increases to 2. 2 p. u. DC Current DC Voltage falls to zero at the rectifier 6

Fault Identification in HVDC System abc to dq 0 transformation 1. Computes the direct

Fault Identification in HVDC System abc to dq 0 transformation 1. Computes the direct axis, quadratic axis and zero sequence quantities in a two-axis rotating reference frame for a three-phase sinusoidal signal 2. Known as Park transformation 7

abc to dq 0 transform in DC and AC fault of HVDC PEAK MAGNITUDE

abc to dq 0 transform in DC and AC fault of HVDC PEAK MAGNITUDE OF DQ UNDER DC FAULTS Parameter Normal Operation Distance (km) DQ magnitude 0 12. 38 50 26. 58 100 27. 48 150 28. 09 200 28. 18 250 28. 40 DC Fault 9

abc to dq 0 transform (conti. ) PEAK MAGNITUDE OF DQ VALUES UNDER AC

abc to dq 0 transform (conti. ) PEAK MAGNITUDE OF DQ VALUES UNDER AC FAULTS AT RECTIFIER SIDE AC Faults at Rectifier side Peak Magnitude of dq in unit No Fault 12. 38 Single line to ground fault 99. 23 Line to line fault 150. 60 Double line to ground fault 144. 96 Triple line to ground fault 223. 18 10

Wavelet Transform 1. Mathematical Technique 2. Analysing signals simultaneously in time and frequency 11

Wavelet Transform 1. Mathematical Technique 2. Analysing signals simultaneously in time and frequency 11

Wavelet transform in DC and AC fault of HVDC DC faults at 150 km,

Wavelet transform in DC and AC fault of HVDC DC faults at 150 km, 250 km Absolute maximum value of five levels wavelet coefficients of dc line current for dc fault at various fault distances DC Fault Normal Operation 50 km 100 km 150 km 200 km 250 Km Max. value of wavelet coefficients in five level 0. 09 0. 18 0. 28 0. 67 0. 78 0. 83 12

Wavelet transform (cont. ) SLG and LL Fault Absolute maximum value of five levels

Wavelet transform (cont. ) SLG and LL Fault Absolute maximum value of five levels wavelet coefficients of dc current for various ac fault at rectifier side AC Fault at Rectifier side Normal Operation Maximum coefficients 0. 09 SLG LL 0. 35 0. 43 13 DLG 0. 25 LLL 0. 41

Conclusions • 1. Fault Identification in HVDC system is a challenging process because it

Conclusions • 1. Fault Identification in HVDC system is a challenging process because it should be accurate and fast • 2. abc to dq 0 and wavelet transform can be applied to fault identification • 3. In fault identification abc to dq 0 transform performed very well but the accuracy of fault distance estimation was poor. • 4. Wavelet transform is considered as a powerful signal processing tool for transient analysis of signal • 5. Wavelet transformation effectively proved that it can detect the abrupt changes of the signal indicative of a fault. 13

Thank you

Thank you