1. Large Disturbance in the European Power System on the 4th of November 2006 CAMS/RRPA Panel Session Mitigation and Prevention of Cascading Outages: Methodologies and Practical Applications PES General Meeting, Vancouver, Canada July 24, 2013 Janusz Bialek (Durham), Vladimir Terzija (Manchester), Taiying Zheng (Manchester), Walter Sattinger (Swissgrid) janusz.bialek@durham.ac.uk

2. Site of two UNESCO Wold Heritage Sites 3 rd oldest university in England Durham University

3. PMAPS 2014, Durham International Conference on Probabilistic Methods Applied to Power Systems, 7-10 July 2014 Paper submission deadline 30 Nov 2013 www.dur.ac.uk/pmaps.2014, pmaps.admin@durham.ac.ukwww.dur.ac.uk/pmaps.2014

4. Source of data and diagrams: UCTE Final Report System Disturbance on 4 November 2006 15M households affected, 16,724 MW lost The worst disturbance in 50 years of UCTE/ENTSOE in terms of the number of TSOs affected and frequency deviations involved Duration 1 hour and 35 minutes (22:10-23:45) but only 38 minutes three island operation

5. East-West transfer 15 GW of wind (5.5%) High flows around Germany Source: UCTE Voltage phase angle differences in the UCTE system at 22:00 Source: UCTE Abschlussbericht zur Systemstörung 4. Nov. 2006 Loading pre-disturbance

6. Note the difference between scheduled and actual flows (e.g. FR-D, FR-BE) due to loop flow phenomenon Especially important D-NL, D-PL due to high wind Source: UCTE

7. Timeline 18 Sept: a shipyard request EON for a routine disconnection of double circuit 380 kV line in Northern Germany on 5 Nov 3 Nov: the shipyard request to bring forward the disconnection by 3 hours EON agrees provisionally but does not modify Day Ahead Congestion Forecast (DACF) distributed to all TSOs 4 Nov, 7 pm: EON informs RWE and TenneT about new time for the line outage 9.30 pm: EON concludes empirically, without updated (N-1) analysis, that the outage would be secure (it wasnt!) Image: http://www.cruise-ship-report.com/News/110506.htm

8. RWE does (N-1) analysis of its area which indicates high but secure loading 9.38: EON disconnects the line 9.39-41: warnings of high flows EON assesses the situation empirically, without simulations, and decides to couple a busbar to reduce the current by 80 A Result: the current increases by 67 A and the line trips Cascading line tripping all over UCTE and separation into 3 regions with different frequencies Source: UCTE

9. 0.8 GW deficit 49.7 Hz 8.9 GW deficit 49 Hz 10 GW surplus 51.4 Hz Source: UCTE

10. Western Europe: 8.9 GW deficit Drop of frequency halted by load shedding But frequency drop caused tripping of 10.7 GW of generation (40% wind)

11. North-Eastern Europe: 10 GW surplus Initial rise of frequency halted by AGC and tripping of frequency-sensitive generation (mainly wind) As frequency started to drop, windmills started to reconnect automatically worsening the situation Situation stabilised by manual action of TSOs Source: UCTE

12. South-Eastern Europe: modest 0.8 GW deficit No load shedding activated, subsystem (N-1) secure Source: UCTE

13. Resynchronisation A number of uncoordinated unsuccessful attempts made without knowledge of the overall UCTE situation Full resynchronisation after 38 minutes Source: UCTE

14. UCTE root cause analysis Main points: – (N-1) security rule, inadequate inter-TSO coordination – Lack of situational awareness – Other factors (wind farms, lack of coordination)

15. Improvements since 2006: situational awareness Web-based visibility of cross-border flows in Europe, ACE, generation traffic light system to indicate security, control reserves and state of IT infrastructure RAAS – real-time awareness and alarming system, EAS (ENTSO-E awareness) – all TSOs have the same view – the information maintenance is done on two central points with highly redundant infrastructure.

16. Improvements since 2006: coordinated (N-1) security analysis All national files are merged into one common CE load flow file each TSO downloads the complete system and perform complete (n-1) calculation. Evening phone/web/video conference of all the TSOs to coordinate remedial actions

17. Further Measures: Synchronized Measurement Technology, WAMS WAMS HSE/PMU-SE Situational Awareness Dynamic Security Assessment Decision Making

18. All measurements are synchronised by GPS signal Data Records during 39 Minutes of the Islanding System Operation Source: W.Sattinger, Swissgrid

19. Large Disturbance in the European Power System on the 4th of November 2006 CAMS/RRPA Panel Session Mitigation and Prevention of Cascading Outages: Methodologies and Practical Applications PES General Meeting, Vancouver, Canada July 24, 2013 Janusz Bialek (Durham), Vladimir Terzija (Manchester), Taiying Zheng (Manchester), Walter Sattinger (Swissgrid) janusz.bialek@durham.ac.uk