DALLAS, Sept. 11, 2003 -- "There is no excuse for that outage, it should not have happened," Michehl Gent, President of the North American Electric Reliability Council (NERC) told the Institute of Electrical and Electronics Engineers (IEEE) Transmission and Distribution conference this week.
A special session was devoted to providing an update on the August 14, 2003 outage that blacked out 62 GW of load in eight northeastern states-largest electrical blackout in US history.
Gent said that restoration of the 62 GW of lost load within 30 hours was "miraculous." He explained that the focus is now on the investigation to find out why it happened and to determine how to improve protection against such outages in the future. The final report of the official investigation into the blackout may not be finished for a year, Gent told the meeting.
"We need to avoid jumping to conclusions," said Gent, "and focus on the facts." He said that it appears the blackout resulted from a "perfect storm" of adverse conditions and events and that the likelihood of another such blackout in the near future is remote. He noted that it took three years to produce a final report on the 1965 blackout, the previous record holder.
The US investigation is being led by the federal Dept. of Energy. Technical analysis is being provided by Gent's organization, which provides voluntary reliability guidelines and standards for the US and part of Canada's high-voltage power delivery system. A key question in the investigation is to determine if transmission system operators complied with NERC's reliability guidelines.
The NERC-led team of engineers is currently collecting information and data records so they can develop a timeline of significant events and identify the root cause of the blackout. Gent said the investigators have obtained about 90% of the relevant data records-several hundred thousand records in all-and reviewed about 40% of them.
So far they've identified about 200 "significant" events. Gent said an investigation of this scope has never been done before. He said the investigation of the last big cascading failure outage in the US, the 1996 western system blackout, was "simple" compared to this one.
Gent said that generating plant outages at the time of the blackout were not a significant contributor. He also said that no equipment-with the exception of one generator-was destroyed as a result of the outage.
One problem the team has encountered is determining the reliability of time stamps on the data records. In order to develop an accurate sequence of events from data from all the companies involved in the outage, the time stamps must be consistent. "Were all computer clocks synchronized," he asked. He noted that if records from digital fault recorders are not linked to a standard time, "it's not worth having them."
If we leap to conclusions, said Gent, we might conclude that all 50 states should be isolated as Texas is. He said he feels it will be better to improve controls for the existing systems and have a more effective compliance program to enforce the rules of operating synchronized high voltage grids. Gent said NERC has been pushing for federal reliability legislation for five years. The legislation currently in Congress does give NERC teeth to better enforce its standards, ranging from exposure of violators to fines. He said the legislation is "very close to being law."
Another expert at the IEEE meeting, Damir Novosel, Senior Vice President of KEMA, said, "We believe protection was big part of this event." He said that protection system failures, misoperation, and outdated settings were involved in 70% of past North American blackout events.
Carson Taylor, principal engineer for transmission operations and planning at Bonneville Power Authority said reviews of four cascading failures in the western interconnection between 1994 and 1996 showed that the best protection is protection depth, "Do everything you can." The biggest western outage was the loss of 30 GW of load in August, 1996. He noted that a common problem in these blackouts was the tripping of generators at a time when they were in no danger of experiencing damage.
A spokesman for American Electric Power reported the sequence of circuit breaker actions on August 14 on the 138-kV and 345-kV lines that connect AEP with First Energy in northern Ohio, noting that the AEP system stayed up for the most part. He said that there was at least one tree contact related failure early in the sequence and a tree contact fire caused by a line sag later in the sequence. He said the AEP system survived because it kept load and generation in balance, its protective equipment operated as designed, operators kept in contact with their neighboring utilities, and the AEP grid is robust.