Forced Oscillation Event Analysis

Transcription

1 Forced Oscillation Event Analysis January 11, 2019 Event Ryan D. Quint, PhD, PE Senior Manager, Advanced System Analytics and Modeling, NERC February 2019

2 Background Oscillation observed across entire Eastern Interconnection from 08:44:41 UTC (03:44:41 EDT) to 09:02:23 UTC (04:02:23 EDT) 2 Oscillation frequency of 0.25 Hz o Aligns with inter-area mode frequency across EI Power swings around Florida of 200 MW, around ISO-NE of 50 MW RCs identified oscillation on PMU data, notified RC Hotline UTK provided videos of oscillation event GOs noticed oscillation on power plants across EI Source tentatively determined as Bayside #2 Steam Unit NERC issued PMU data request, working with possible source NERC will perform oscillation analysis on wide-area data set NERC working with industry to study event, develop mitigating actions

3 NERC PI System 3

4 Overview of Oscillation Strange oscillation in Florida around 08:44:21 UTC (03:44:21 EDT) Oscillation strengthens at 08:44:41 UTC (03:44:41 EDT) 4 Oscillation ends at 09:02:23 UTC (04:02:23 EDT)

5 FNET Videos Video of onset of oscillation:

6 FNET Information 6

7 Generator Analysis 7

8 FPL 8

9 FPL 9

10 FPL 10

11 FPL 11

12 TVA SCADA Data Flow on 500 kv tie line with Southern Company (Bradley-Conasauga) Event 200 MW Peak to peak 12

13 TVA 13

14 ATC Plot is sum of 5 EHV lines connecting from Wisconsin to Illinois 14

15 ATC 15

16 ComEd Current on an east-west 765kV line 16

17 AEP Antioch at JacksonsFerry 200 MW :45:00 03:50:00 03:55:00 04:00:00 Time (Eastern) 60 FFT Magnitude Frequency (Hz) 17

18 Duke Energy 18

19 ISO-NE ISO-NE Phasorpoint generated 14 alarms (20+ Alerts) from 3:46-4:03 AM, Hz oscillation, magnitudes up to MW RMS OSL application automatically processed Alarms and Alerts, identified source of oscillations located beyond ISO-NE area Plot: MW flow on 345KV line between NE and NY. 19

20 PMU Data Collection NERC issued data request at 9:05 PM EDT on to gather PMU data from RCs 20

21 WSU FFDD TVA-only data analyzed Damping of estimated mode with frequency 0.25 Hz is close to zero. Well-damped mode before and after event with frequency of 0.24Hz. Appears that well-damped mode is the well-known 0.22 Hz North-South mode of EI Conclusion that event caused by forced oscillation with frequency close to system mode. Analysis of data with FSSI engine in the next section will make the analysis clearer. Oscillatory modes of 0.75 Hz and 1 Hz dominant estimated modes before and after event. Both modes show low damping, may be related to other forced oscillations. Further analysis of these modes requires more PMU data. 21

22 WSU FSSI: TVA-only data analyzed Once event starts, mode with frequency of 0.25 Hz with damping about zero percent is clear in time plots. Analysis of voltage phase angles and current phase angles shows that during event there are two estimates with close frequencies about 0.25Hz. o One of these estimates, which seems to be the system mode with frequency about 0.23Hz, is a well-damped mode. This mode is persistent before, during, and after the event. o Another estimation with the frequency of 0.25 Hz appears during the event with damping about zero. The oscillation seems to be a forced oscillation which is happening at the frequency close to the system mode frequency. Since frequency of forced oscillation and system mode is very close, there is strong evidence that this event was related to inter-area resonance effect between system mode and forced oscillation. 22

23 WSU WSU Conclusions: 23 Online tool detected 0.25 Hz oscillation with high energy, ~0% damping. All 3 engines FFDD, FSSI, and EAO detect and alarm the low-damping oscillations with frequency of 0.25 Hz once the event starts. Before and after event, system mode with f = Hz well-damped. Analysis of phase angle data by FSSI showing two persistent modes with close frequencies of about 0.25 Hz. o Well-damped mode seems to be the system mode and the one with damping near zero seems to be a forced oscillation. o More detailed studies with additional PMU data is needed to determine the effect of possible inter-area resonance phenomenon. Before and after the event, poorly damped 0.75Hz mode is likely caused by the well-known forced oscillation in EI that has been present for years. Phase angle signals (both voltages and currents) are showing a very lowdamping mode with frequency of 1 Hz which may be related to clock issues in some of the PMUs. These should be fixed.

24 WSU 24

25 Source Oscillation Repowered combined cycle plant steam turbine oscillating Power-load imbalance (PLI) controls Failed potential input feeding feedback Feedback reading 2/3 power output vs. input perceived imbalance Drives shut intercept valves 4 second timer to reopen valves open/close cycle Different potential inputs for relaying and controls/metering Relaying hence did not pick up the imbalance/issue Plant manually tripped by operator Upon inspection, failed wiring in PT cabinet Damaged intercept valves, needed replacement, unit off-line for relatively extended period (couple weeks) 25

26 Overview of Forced Oscillation Errored Signal Forcing Function 26

27 Ryan Quint, PhD, PE Senior Manager Advanced System Analytics and Modeling North American Electric Reliability Corporation Office (202) Cell (202)