 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
|
Prepared Witness Testimony The House Committee on Energy and Commerce
Blackout 2003: How Did It Happen and Why?
Full Committee on Energy and Commerce
Mr. H. Peter Burg
Mr. Chairman: Thank you for the opportunity to testify today. I am Pete Burg, chairman and chief executive officer of FirstEnergy Corp., a registered public utility holding company headquartered in Akron, Ohio. FirstEnergy's seven electric utility operating companies provide electric service to 4.4 million customers in Ohio, Pennsylvania and New Jersey. We commend your determination, Mr. Chairman, to understand the events of August 14. Operation of the electricity grid is an extremely complex matter. Knowing all of the facts is vital to arriving at the policy decisions that could mitigate the risk of a repeat of this kind of outage. We are committed to helping determine what went wrong in the Eastern Interconnection on August 14 and are pleased to have the opportunity to tell you directly what we know about our system and our region within the Interconnection. Summary Notwithstanding the service interruptions on August 14, the United States has a reliable electric system, and I am particularly proud of FirstEnergy's 345 kV transmission system, which has achieved a top-quartile ranking among companies in the 2003 SGS Transmission Reliability Benchmarking Survey. FirstEnergy has been the subject of a great deal of speculation during the past three weeks regarding the outage. Clearly, and as we have said from the outset, events on our system, in and of themselves, could not account for the widespread nature of the outage. After much more evaluation, we continue to believe this is true. We strongly believe that such a widespread loss of power could only result from a combination of events, not from a few isolated events. Industry experts share our view. Dave Nevius of the North American Electric Reliability Council (NERC), an organization charged with working to maintain electric reliability in the country, said, "It's a more complicated problem than just one utility." Alan Schriber, chairman of the Public Utilities Commission of Ohio, said, "This has been the perfect storm of electricity. It's the confluence of a lot of bad things going on that day, and I don't mean just after 4 o'clock. There had been noticeable aberrations throughout the system all day." Today, I will highlight for you some of the significant events that we know happened. Bear in mind, however, that no one has all of the information yet, and it will take some time before we do. Much of the information regarding the events of the day is still being collected and analyzed. But from what we do know, a number of events occurred throughout the day, all of which could have combined to affect the Eastern Interconnection's ability to perform. It is understandable that everyone is looking for the straw that broke the camel's back. But there is no one straw - they're all heaped together. And the camel's ability to support the load cannot be overlooked. The reliability of the system - maintained by built-in reserve margins, operating protocols, sharing arrangements, communication systems, sophisticated electronics, and human vigilance - is a marvel. However, all those protections were not sufficient to prevent the problems that arose on August 14. The electric system is designed to handle contingencies that are bound to occur. Redundancies and protective devices are built into the system to protect equipment, maintain service to customers, and ensure safe operation. And, the entire interconnected network was built to provide support in emergencies. The interaction of all of these complex elements must be considered. The role of these protective devices - as well as their automated operation on August 14 - should be a focus of the investigation. In addition, the investigation should take into account the fact that the transmission system was not designed to serve regional wholesale electricity markets as an integrated transmission superhighway. Like other transmission owners, FirstEnergy built and maintains its transmission facilities to reliably meet the requirements of customers in its own service area. While we support federal policies to adapt the existing transmission system to the needs of regional wholesale markets, no one's transmission system was constructed with this purpose in mind. Data Collection Process As I mentioned, August 14 data are still being analyzed. Our transmission experts are studying millions of data points. We have digital fault recorder data, oscillographic data, analog charts, and other recordings. Some of the computer records have readouts in millisecond intervals, some in two-second intervals, and others in 30-second intervals. Additionally, the times for this information need to be calibrated and synchronized. The industry calculates the precision of time for its systems relative to the atomic clock. If the grid runs slightly above or below 60 Hertz, clocks will deviate slightly from the atomic clock, so the system must run slightly slower or faster to adjust. In fact, on August 14, the Eastern Interconnection was being run slightly below 60 Hertz to slow clocks. The relevance of the synchronization is that, when precipitating events occur in rapid succession, it is necessary to establish precise times to gain a clear picture of sequences and interrelationships. This is a tedious and lengthy process that is still being completed. We have shared the information we have gathered to date with the Department of Energy and the NERC, and we will continue to fully cooperate with them and with your Committee. Personnel and System Reliability FirstEnergy system dispatchers operating the facilities at the time of the events on August 14 have an average of more than 10 years of experience and are NERC-certified professionals. They are a dedicated group that takes immense pride in ensuring the safe and reliable operation of our system. From 1999 through 2002, we have spent $433 million system-wide on transmission operations, maintenance and capital, with nearly $200 million spent on transmission in Ohio. More specifically, the four FirstEnergy 345 kV lines that failed on August 14 had a history of good performance. There were no sustained outages in 2001, 2002 and through August 13, 2003 on the Chamberlin-Harding, Star-South Canton and Sammis-Star lines. The Hanna-Juniper line had six outages in 2001 ranging from four minutes to 34 minutes but none was tree related. The Hanna-Juniper line had no sustained outage in 2002 or 2003 through August 13. In short, we have qualified operators and we are consistently making the expenditures necessary to improve and maintain our transmission infrastructure. Grid Operation and Oversight Responsibilities The NERC and its affiliated regional reliability councils have the mission to ensure that the bulk electric system in North America is reliable, adequate and secure. That system is designed to maintain the interconnected network in order to provide emergency support and to prevent actions on one system from having unintended consequences on another. Since its formation in 1968, NERC has operated as a voluntary organization, relying on reciprocity and the mutual interest of all those involved. In recent years, NERC has made a significant effort to respond to changes in the regulation of the electric utility industry. For example, NERC has been aggressively seeking passage of legislation, which FirstEnergy also supports, to enable it to become an industry-based, self-regulatory organization enforcing mandatory reliability standards. Utilities operate their systems and maintain interconnections consistent with standards and guidelines adopted by NERC and its regional reliability councils. The systems are designed to withstand single and multiple outages while still performing reliably. The regional councils conduct assessments of the interconnected systems and continuously revise the standards that utilities and other industry participants observe to enable daily operations and, increasingly, electricity trading to be conducted in a reliable manner. The Federal Energy Regulatory Commission (FERC) has required that Regional Transmission Organizations (RTOs) take responsibility for regional reliability. The Midwest Independent System Operator's (ISO) role as one of the reliability coordinators for the region is an example of how that responsibility is discharged. While the physical control of the system remains with the control area, the reliability coordinator shares responsibility for assuring that the bulk electric system is reliable, adequate, and secure. The Midwest ISO is the reliability coordinator for our transmission assets in Ohio. Clearly, a common understanding of the cumulative events of August 14 will contribute very significantly to the consideration of reforms that can and should be made in reliability assessments and standards, and in protocols and procedures for commercial operation. One issue that comes to mind, however, is whether the reliability standards related to protective systems and their interactions with one another should be examined in light of the new ways we are using the interconnected networks to support inter-regional and international trading and marketing of electricity. For example, according to the Electric Power Research Institute, the number of wholesale transactions has increased by 400 percent over the past decade. The East Central Area Reliability Council (ECAR) transmission systems in particular have been used for increasing volumes of area-to-area and region-to-region transactions, supplying deficit areas within the Independent Electricity Market Operator (IMO), the New York Independent System Operator (NYISO) and the Mid-Atlantic Area Council (MAAC), predominantly from resources south and west of FirstEnergy. To fully understand the events of August 14, and more importantly, to evaluate what needs to be done to redesign the transmission system, these changes in usage of the system must be considered because they impact power flows and add stress to existing facilities. System Conditions and Events Attached to my testimony is a chronology of events and summary of power flows that describe, as we know them, the condition of the regional transmission system on August 14, events occurring on that system, and the changes in power flows through FirstEnergy's system. This information has been compiled by a collaboration of our transmission and generation personnel and other company and industry experts. It should be noted that it depicts only a partial picture of that day, because all of the conditions and events that took place throughout the Eastern Interconnection are not available to us. That broader information ultimately will be required to fully understand what happened and what actions will be needed to mitigate the risk of such an outage in the future. However, the following summarizes what we know as of this point. On August 14, our load was projected to be approximately 85 percent of our estimated peak summer load. Load and weather conditions for the day were typical for a mid-August day. The areas in our service territory affected by the outage experienced seasonable temperatures and no major storms. The ECAR region had adequate generation available, even with a number of large generating units owned by various companies off-line during the day, including Detroit Edison's 800 megawatt (MW) Monroe Unit 1, AEP's 1,133 MW DC Cook unit and FirstEnergy's 883 MW Davis-Besse unit. AEP's 1,300 MW Gavin Unit 2 was also off-line and was not scheduled to come back online until the afternoon of August 14. FirstEnergy's projected load was 11,958 MW. Our generating capacity for the day was 10,641 MW, and with net scheduled import power, we had adequate spinning reserves. Power generally was flowing from west to east, and from south to north - its typical pattern. Ontario was importing about 2,500 MW; New York was importing about 1,300 MW; and, the Mid-Atlantic Area Council (MAAC) area was importing about 2,500 MW. It is now evident that unusual system conditions, some of which were detected at the time, were occurring during the day. To get a better understanding of these conditions, NERC has taken the right approach by reviewing the events of August 14 beginning at 08:00. FirstEnergy's attached list of events outlines conditions existing within ECAR on August 14 and details certain occurrences beginning shortly after 12:00. Some of the unusual events include oscillations in flow, frequency dips and reversals in power flow between regions along major interconnections. As the afternoon progressed, a number of generation and transmission facilities in the region became unavailable. In the hours leading up to the outage, generation facilities that went off-line in our region include, in chronological order: AEP's 400 MW Conesville unit; DTE's 600 MW Greenwood unit; and FirstEnergy's 597 MW Eastlake unit. Also that afternoon, Gavin Unit 2, which is a major facility, was coming back online from an outage, though by 16:00 the plant was only supplying 50 MW of power. Conesville also was coming back online later in the afternoon but was not supplying its full load. Greenwood was also being returned to service that afternoon. Following the trip of the Eastlake Plant at 13:31:34, power systems and flows corrected themselves and FirstEnergy's system was balanced and stable. And even though, as the day proceeded, a number of other events occurred, our system remained in balance and power flows continued to be about the same as experienced earlier in the day. Between 15:00 and 15:30, we lost our Chamberlin-Harding 345 kV line. After that event and others during the same time frame, our system was still stable and importing essentially the same amount of power as earlier. During this time, our power flows to Michigan were approximately 346 MW. Between 15:30 and 15:45, a number of transmission lines in the area tripped out of service. These included our Hanna-Juniper 345 kV line, the South Canton (AEP)-Star (FE) 345 kV line, the Cloverdale (FE)-Torrey (AEP) 138 kV line, AEP's East Lima-New Liberty 138 kV line and our Pleasant Valley-West Akron/West 138 kV line. Even with the loss of these facilities, our net imports remained approximately the same, with power flows continuing into Michigan at 215 MW. From 15:45 to 16:05, the Cloverdale (FE)-Canton Central (AEP) 138 kV line, East Lima (AEP)-North Findlay (AEP) 138 kV line, the West Akron 138 kV bus, and the Dale (FE)-West Canton (AEP) 138 kV line all tripped off. The Canton Central (AEP)-Tidd (AEP) 345kV line tripped and reclosed, although two 345-138 kV transformers remained isolated. Following these events, our net imports dropped by about 230 MW, reflecting reduced loads within our service area. Even so, about 150 MW continued to flow to Michigan. At 16:06:03, FirstEnergy's Sammis-Star 345 kV line overloaded and tripped. At this point, 150 MW that had been flowing into Michigan reversed and 155 MW began flowing from Michigan into Ohio. A reversal of this magnitude would not, in and of itself, appear to be of particular significance. Then, at 16:08:58, AEP's Muskingum-Ohio Central 345 kV line tripped, and at 16:09:06, AEP's East Lima-Fostoria 345 kV line tripped and reclosed automatically after a less-than-two-minute delay. At this point, flows from eastern Michigan to Toledo increased by 1,855 MW; flows from FirstEnergy into AEP increased by 2,670 MW; and flows from AEP directly into western Michigan increased by 1,630 MW. Also at this time, because we had lost load, FirstEnergy's net imports actually were reduced by 1,100 MW. Then, events began occurring rapidly throughout the Eastern Interconnection. A critical fact is that the system kept working as it was designed to do, despite all of these circumstances. On the afternoon of August 14, before we began to see unusual loop flows - a result not only of power seeking ways around unavailable lines, but also of the existing power flows between regions - we did not see the system perform inconsistently with its design. When a path started to overload, the circuit breakers performed as designed to cut off flow on the line and protect components from overheating and sustaining significant damage. This all happened automatically. In fact, to the best of my knowledge, no manual intervention to disconnect from the interconnected network was taken by anyone. Our system remained interconnected with DPL, DQE, APS, and AEP. And, through APS, we remained connected with PJM through PJM West. None of these interconnected systems experienced major service interruptions. While we also experienced problems with our Energy Management computer system and are still evaluating the functionality of that system that was available to our dispatchers during this time frame, information about our system and the events that were occurring throughout the day were available to our reliability coordinator, the Midwest ISO. Also during this time, our dispatchers were in communication with other system operators and plant operators, as well as the Midwest ISO. The Midwest ISO did not call for any system interventions. Looking back, however, I do not believe it would have been appropriate for operators to intervene in this event. Again, these systems are designed to protect themselves and interact in a way that keeps power flowing to as many customers as possible. That is how they functioned. The automation in fact did protect facilities and customers in adjoining areas. NERC Policy Number 5, relating to Emergency Operations, states: "When an operating emergency occurs, a prime consideration shall be to maintain parallel operation throughout the Interconnection. This will permit rendering maximum assistance to the system(s) in trouble." Had the ties been disconnected, the negative impact on these other systems and their interconnections and customers could have been significant. Once systems were rapidly reacting to surrounding conditions, it was beyond the ability of operators to control. Even recognizing the point at which the automated response began moving too rapidly for operator intervention may have been impossible when the systems are doing what they are supposed to do. Even if such a point could have been quickly recognized, there was no time to react. The Washington Post reported last week that, according to a PJM spokesperson, the blackout spread too fast to employ emergency measures, and that "there were no conclusions reached" about transmission load relief (TLR) measures until it was too late to implement them. Everyone wants to know: "Why did the blackout happen? What was the precipitating event?" Some might like to say it was the first outage that occurred. If that were the standard, which of the above facilities was the "first" outage? Some might like to say it was the first outage that caused power flows to shift in response. But every outage causes flows to shift. Others might like to say it was the moment the system was "in trouble" and couldn't recover. However, we believe it was the cumulative effect of occurrences in the region that combined to impact the event, not unlike the boxer who gets knocked out in the tenth round. That last punch was important, but the accumulation of all of the previous blows led to his weakened condition. We need to understand the many small, chance events that played a role on August 14, but more importantly, we need to focus on the larger system conditions that imposed a strain. Observations I have several observations about the foregoing that I would like to summarize. First, the events of August 14 demonstrate that the facts are complicated and interrelated in ways that no one yet understands. Second, the systems appear to have responded as they should have to address successive issues. However, at some point events began to occur so rapidly that, even if human intervention were advisable, it likely would have been impossible to avert the widespread nature of the outage. As mentioned above, this was an automated event in which the system and protective devices appeared to work as designed. However, the broader question is whether the design is appropriate given what we are asking the system to do today. Localized grid conditions do not explain the widespread nature of the event. Third, competitive markets impact the operation of the interconnected grid. A significant amount of power passes through our area en route to areas that do not have the same generation resources that exist in ECAR. As many have observed, this puts additional stress on the grid. On August 14, this situation resulted in certain areas being left with excess generation or load when individual systems were automatically separating themselves from one another. In the past, utilities were responsible to match load and generation within their own service territories. The system was designed and constructed with that mission in mind. Interconnections with neighboring utilities were reliability enhancements, not on-ramps to an interstate highway. We support wholesale markets, and in fact strongly rely on them to help serve our retail customers. However, imbalances between regions are a new factor to consider in updating reliability standards for the interconnected grid. This is especially critical when power has to move through several systems to reach customers. Fourth, grid responsibilities are now in the hands of more entities than ever before. This puts a premium on identifying new designs and devices to better coordinate the operation across a wider region. Recommendations I have several recommendations to decrease the likelihood of these kinds of events occurring again. The first is to make the transmission system more robust to accommodate the growth in competitive markets. This requires investment and the ability to site new facilities. Transmission owners make regular investments in their facilities to maintain reliability to their utility customers. I have already noted FirstEnergy's investment and excellent performance. But the nation's transmission investment in general is not geared today toward development of facilities to sustain wholesale markets, which can change on a day-to-day or hour-to-hour basis. This will be challenging because, unlike building facilities to serve a relatively stable customer base as was done in the past, investments to meet wholesale market opportunities can be rendered "excess" by a change in those markets. Transmission rate reform is necessary to encourage investment in the construction of a more robust interstate transmission network. FirstEnergy has been a strong proponent of policies that encourage such investment. We support the applicable language in H.R. 6. Regarding transmission siting, we support provisions in H.R. 6 that would grant FERC with "backstop" eminent domain authority to help get critical transmission lines built in a timely fashion. My second recommendation is to establish mandatory reliability standards for the industry. FirstEnergy has long supported such standards as an element of federal legislation, including those in H.R. 6. Third, consistent with my previous observation, policy makers and the industry need to review automated and control systems to determine whether the right equipment is in place and whether new technologies would prevent this type of event. Over the years, we have supported proposals for the federal government to promote implementation of new transmission technology. It may be too soon to tell what advances in technology would have been necessary to mitigate the outage. Fourth, the government needs to review the significance of interstate power flows across interconnected grids. Experts agree that a more vigorous grid is needed to accommodate wholesale markets. But what does that grid and its protective equipment need to look like, based on what we see today, and where the market development trends are heading? In the course of reviewing the August 14 event, we may gain a better understanding of the impact of regional power flows, at least in the affected regions. But we must make sure that electric customers in one area are not burdened by the cost of transmission facilities that are being constructed because another region is not siting sufficient generation. Like in the old days, utilities built generation and the necessary transmission to get it to their customers. Generators should not be able to avoid these costs today. Conclusion It is not possible at this time to pinpoint the "causes" of the outage. There are many contributing factors. FirstEnergy is committed to working with Congress, the Administration, and industry organizations to promote changes that will decrease the likelihood of this kind of event occurring again. Chronology of Events
This chronology is based on currently available information. It therefore should be considered preliminary and subject to revision. Major FE and Related Transmission PerspectiveAugust 14, 2003 Event All Times in Eastern Daylight Savings Time The following is a summary of major events known to FirstEnergy ("FE") at this time related to the operation and system conditions on the afternoon of August 14. The summary of major events focuses on the FE and East Central Area Reliability Coordinating Agreement ("ECAR") systems where the data are the most complete at this time. It includes events known to FE in nearby interconnected systems where those effects have impacts on the operation of the FE resource supplies and transmission system. This chronology is based on currently available data from FE as well as data from the Midwest Independent System Operator (MISO), the Michigan Electric Coordinated Systems (MECS), the Pennsylvania-Jersey-Maryland Interconnection Company (PJM) and other sources. Data used to compile this timeline, whether from FE or other sources, should be considered preliminary and subject to revision. The time stamps provided within this chronology are likewise preliminary and approximate, but reflect the time recorded within the FE Energy Management System (EMS) or other local device. Time synchronization and corresponding adjustments have not been completed. Additional information for events earlier in the day is still being compiled. The following is a brief backdrop of conditions existing within ECAR on the day of the event.
Chronology of Regional Events known to FE at this time :
The FE transmission system conditions following these occurrences are within the range of acceptable operating limits with no overloads and no out-of-range voltages.
The FE transmission system conditions following these occurrences are generally within the range of acceptable operating limits with no overloads and voltages in the Akron and Massillon areas are at the lower end of the acceptable operating range. Sammis-Star 345 kV line loading is approximately 104% and there are some local area overloads.
The preceding sequence of 138 kV line trips began to shift line flows in the Michigan and Toledo area, increasing the flows on other tie lines between AEP, FE, and MECS. The FE transmission system conditions following these occurrences are generally within the range of acceptable loading and voltage operating ranges, although Sammis-Star 345 kV line loading is 111% and voltages and some loadings in the Massillon and Akron areas fall outside the acceptable range.
Akron and Massillon area load began to decline significantly. Some load in southern Cuyahoga County also begins to decline. FE load is reduced by approximately 1000 MW at this time due to distribution substation trips and under-voltage or over-current trips within customer sites. FE generating units along Lake Erie are operating and ties with other adjacent control areas are operating within line limits, although local voltages in the Akron and Massillon areas fall below 80% and the Mansfield and Marion area voltages below the acceptable range.
At this time, power flow between Michigan and Ontario reversed from exporting 400 MW from MECS to importing 250 MW from IMO, and flows increase between Michigan and Ohio to 2000 MW, while flows increase to approximately 3800 MW from AEP into Western Michigan. Approximately 1000 MW supports the Toledo load center while approximately 700 MW flows into the AEP system. This reverses the net flow between AEP and FE from imports into FE of 2200 MW to a net export to AEP of 450 MW. Meanwhile, to the east of the FE system, flows into the Cleveland area from PJM increased to about 1000 MW, which is within the line's rated capability. At the same time, flows increased to about 3600 MW from AEP and FE plants into Allegheny Power Systems (APS) and PJM.
The following additional blackout-related items are noted:
Timeline Presentation(Adobe PDF Required)
|
|||||||||||
|
