 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
|
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. Joseph Welch
Introduction Mr. Chairman and Members of the Committee, thank you for this opportunity to appear before the Committee to present my Company's views on the electrical blackout that began on August 14, 2003. My name is Joe Welch and I am the President and Chief Executive Officer of the International Transmission Company (ITC) with headquarters in Ann Arbor, Michigan. This past February 2003, ITC acquired the transmission system formerly owned by DTE Energy. Regardless of the cause of this occurrence, all of us in the electric power industry need to commit to take whatever action is necessary to prevent such a widespread event from happening again. We designed this vast interconnected grid to increase the reliability of service to customers and on August 14 that system failed us. With all the changes that have occurred in the electric power industry in the past twenty-five years, we must find a way to operate our systems in a coordinated manner lest we invite more massive blackouts in the future. We must also find a way to assure that we have adequate infrastructure in the form of transmission lines and related equipment to sustain the customer demands on our electrical grid. My testimony today will describe the factors and events ITC witnessed that led up to the blackout. I will also provide my views on the policies that are important to understanding how this blackout happened and how it might be prevented from recurring. International Transmission Company's (ITC) response to Billy Tauzin Testimony of Joseph L. Welch My name is Joseph L. Welch. I am President and Chief Executive Officer of International Transmission Company. ITC is a truly independent stand alone transmission company with no ties to any market participant or company that brokers electricity, owns electric generating facilities, or has an obligation to serve end-use customers. Our sole mission is to provide the transmission infrastructure necessary to reliably support the electric market in a fashion that minimizes the total delivered cost of electricity to customers. ITC, jointly with the Michigan Electric Transmission Company (METC) (which is another independent transmission company), operates the Michigan Electrical Coordinated Systems (MECS) Control Area in Ann Arbor . This Control Area is responsible for ensuring that generation and load within the Michigan peninsula remains in balance and reports when it is not. ITC is also a member of the Midwest Independent Transmission System Operator (MISO) Regional Transmission Organization (RTO). MISO is the Transmission Provider for the Michigan transmission systems, and is the Security Coordinator responsible for the safe operation of the transmission grid. As Transmission Provider, MISO also schedules the use of the Michigan transmission grid (within its physical limitations) and bills the transmission customers for their use of the transmission grid. ITC became the sole owner of the transmission lines formerly owned by DTE Energy in southeast Michigan on February 28, 2003. DTE Energy's distribution utility subsidiary, Detroit Edison, physically operates, repairs, and maintains all of the ITC transmission assets under contract for a period of one year which began on February 28. On February 28, 2004, ITC will be solely responsible for such physical operation and maintenance in accordance with the February 20, 2003 Federal Energy Regulatory Commission (FERC) Order in Docket Nos. EC03-40-000 and ER03-343-000 which approved the sale of ITC. On August 14, 2003, with absolutely no warning, the ITC transmission grid experienced severe electric flows (which were a result of energy demands of electric customers other than those residing in Michigan) which collapsed our grid and the grids of our interconnected neighbors, ultimately blacking out over 50,000,000 customers. This event is akin to a "tsunami" hitting an unsuspecting costal community. These severe electric flows described above are known in the electric industry as loop flow which is electric energy flow that travels over a transmission system without that flow being scheduled on the transmission system. It can be the results of another transmission provider scheduling and selling more capacity than its own transmission system(s) will accommodate without regard for its impacts on other interconnected transmission systems. Such loop flows also occur when an entity fails to curtail its transactions (imports and/or exports of power) when the transmission needed to support those transactions is no longer available. The letter inviting my testimony correctly notes that "all indications are that the electric power supplies in regions affected by the blackouts have generally been more than adequate to meet peak summer demands." As I will discuss in my testimony, this blackout did not arise from a lack of electric generation supply. Rather, this blackout was rooted in a disconnect between the use of and the capability of the transmission system to deliver that supply. This disconnect in turn is rooted in institutional failures to properly regulate and monitor such transmission usage such that the transmission system stays within its physical limitations. Ultimately, more transmission infrastructure will be required to accommodate increased usage of the transmission system, but, until it can be provided, the proper and safe use of the transmission grid must be enforced. 1. What exactly were the specific factors and series of events leading up and contributing to the blackouts of August 14th? See Attachment 1. 2. At what time did your company first become aware that the system was experiencing unscheduled, unplanned or uncontrollable power flows or other abnormal conditions and what steps did you take to address the problem? Were there any indications of system instability prior to that time? August 14, 2003 began as a typical summer day in Michigan. The only notable generation event was that Detroit Edison's Greenwood #1 unit shut down in a controlled fashion at 1:14 pm EST and returned to service at 1:57 pm EST later that day Electric system metrics :such as system voltages and frequency, as seen from Michigan, were completely with normal limits. Attachment 1, slides 19 through 21, areplots of voltage beginning at 7 am, and Attachment 1, slide 15, is a plot of system frequency for the same time period. Likewise, tie line flows across Michigan's three interfaces with the Eastern Interconnection (METC lines connecting to American Electric Power (AEP), ITC lines connecting to FirstEnergy (FE), and ITC lines connecting to Ontario) were all within normal parameters (Attachment 2) throughout the day, up until the blackout event (Incident). The MECS Control Center has a disturbance monitoring system which collects large amounts of data related to the operation of the transmission system. When the Incident occurred, this system was triggered and began collecting very comprehensive data throughout the Incident in very small time increments, tracking power flows, voltages, frequency, and generator outputs and status. This data enabled ITC to determine: a. A very large demand (2200 MW plus voltage support demand) was suddenly thrown on the ITC's three 345 kV interconnections to FirstEnergy. b. This sudden demand forced power flows to drastically increase across the entire state of Michigan and to a lesser extent, via the ITC-Ontario 230 kV interconnections. This in turn caused depressed voltages on the ITC transmission system (leading to the total voltage collapse). c. These extreme power flows caused the four METC 345 kV lines connecting the METC transmission system to ITC's transmission system to disconnect resulting in the disconnecting of the remaining lower voltage connections between METC and ITC as well. This occurred in a matter of seconds. With METC and ITC disconnected, there was no other supply route for the sudden demand on the (FE) ties except for Ontario. The power demanded by FE subsequently caused the existing flow across Michigan to reverse and flow around First Energy and then through systems such as AEP, Pennsylvania - New Jersey - Maryland (PJM), New York, then into Ontario, Canada via the Ontario-New York ties. d. The voltage collapse within Michigan in conjunction with the power swing through Canada was accompanied by the sudden loss of generators connected to ITC's grid. All of these events and consequences were viewed from within Michigan and I can attest to the data that documents the event which we witnessed. Subsequent reports from various entities including AEP, the MISO RTO, the PJM RTO, and the North American Electric Reliability Council (NERC) indicate that areas in northern Ohio were experiencing serious internal problems for some time prior to the event (approximately two hours before the Incident). AEP and PJM reported they were also experiencing problematic high electricity demand on their connections to FE. While FE is connected to the three different transmission systems of ITC, AEP, and PJM, system flows and voltages within ITC and the rest of Michigan were well within nominal limits all day, notwithstanding the problems to the south. When the AEP and PJM systems disconnected from FE without warning to Michigan, the electricity demand that appeared to have been overloading the AEP-FE connections was thrown onto Michigan. Michigan is a peninsula and the Michigan transmission system was never designed to support northern Ohio on its own, and the results were devastating. The Michigan system collapsed under the strain, followed by the Ontario system shortly thereafter. PJM reported it had disconnected itself from the trouble areas to its west and north, which would make New York and New England a peninsula, isolating them from the Eastern Interconnection. When isolated in this fashion, portions of New York and New England were unable to avoid collapsing when the Ontario system disconnected. 3. What systems operated as designed and which systems failed? Physical systems within ITC operated substantially as designed. I cannot speak to the systems belonging to other entities. The protective relays on transmission lines are designed to disconnect lines for "faults" (for example, a wire touching the ground). Great care is taken to set them so that they do not inadvertently disconnect when there is no such fault (known as "overtripping"). They are also set to reclose automatically, following a safety check, to ensure that the overall grid remains reliable. However, these relays will disconnect lines when voltage collapses, as occurred within Michigan, because voltage collapse presents conditions which are similar to a fault. Transmission line protective relays within ITC appear to have operated properly in response to the conditions presented. The earlier NY blackouts of the mid-sixties resulted in the installation of technical safety devices ("underfrequency relaying") throughout the transmission grid. which undoubtedly have protected the security of the grid in many cases in the past. Unfortunately, in this instance, such equipment was designed to address an imbalance in load and generation (a frequency event), not overuse of the system resulting in voltage collapse as we saw within Michigan, and had little value in mitigating the August 14 event. Black start procedures were generally effective in restoring operation of the grid after this blackout. Protective relays on generators largely disconnected generators before they were seriously damaged. I cannot speak for systems outside of Michigan but ITC transmission lines were undamaged and ready for restoration when the generation needed to supply the load was brought back on line. The systems which did fail were the ones underlying communication. (The communication failures were themselves a predictable outcome of new institutions even now being promulgated by a few parties, notwithstanding substantial objections.) Had Michigan been warned of the problems, a number of actions which would have forestalled the blackout were available. Michigan, in concert with AEP and PJM, could simply have opened its ties to FE as well. The FE system may have survived with some load loss, but more importantly, no cascading would have occurred as the problem would have been localized to the FE system. A better option, given advance warning, would have been for Michigan to prepare for the oncoming tsunami by interrupting air conditioner load in Detroit Edison, by interrupting the large voluntarily interruptible industrial load in Detroit Edison's area, starting Michigan peakers and other available generation), all basically reducing the initial loadings on the Michigan grid and bolstering the voltage support. The Michigan system would not have collapsed, and the cascading blackout would not have occurred. The worse case would have been the collapse of the FE system but FE's problems would have been localized. The best option of all, given an appropriate advanced request, would have been for Michigan to take the same steps outlined above; these steps could have strengthened FE sufficiently that FE may have survived; it would not have been necessary for AEP and PJM to disconnect their systems to save themselves. However, no such call was made or warning given. I have confirmed that by having my staff listen to control room operator tapes. I hope that the DOE task force will review all control room tapes for all the systems that were involved in any way. Phone calls are not the only means of communication. Within MECS at least, there are three electronic systems through which Control Area Operators and Security Coordinators communicate system status, convey warnings, etc. I asked my staff and MECS operators to determine what information was conveyed via that route. They informed me that there were no records or reports of the line outages which were so critical to this event. Without such information, there is no way for Control Area Operators or Security Coordinators to take actions necessary to mitigate problems, especially those events in other systems which could affect our system. I would expect that DOE will review this matter and determine why information was not communicated via those systems. The fact that no such calls or communications were undertaken or warnings extended or even properly reporting of the (subsequently reported) line failures in FE and AEP and PJM illustrates the number one cause of the blackout in my opinion. 4. If events similar to those that occurred on August 14, 2003 had happened a year ago, would the results have been the same? Yes. The infrastructure components underlying this event were the same a year ago. 5. If similar events occur a year from now, do you anticipate having in place equipment and processes sufficient to prevent a reoccurrence of the August 14 blackout? ITC will proceed immediately to implement a plan that will protect ITC and its users, and Michigan as well, from further blackouts. It is unlikely that the physical infrastructure will be implementable within a single year, but we will proceed as soon as possible. The external processes necessary to avoid a reoccurrence will have to be undertaken at the national level; at the moment, a number of entities are attempting to institutionalize the underlying structure which sets up conditions which led to the blackout. 6. What lessons were learned as a result of the blackouts? On August 14, it was apparent that parties were choosing to operate the grid within their sphere of influence for their own purposes without regard to rules, procedures, or the impact of their actions on other users of the grid. Further, the convoluted RTO configurations which major entities have contrived to create virtually guarantees that communication, when it occurs, will be a matter of luck. As MISO Market Monitor Dr. David Patton warned in a March 2003 MISO market monitor presentation to FERC, "The electrical configuration between the PJM and the MISO also raises substantial gaming concerns." Entities will have the means to game the system to their own ends to the disadvantage of all other users. The regional RTOs have proposed to "paper over" this "seam" which is the focal point of the blackout with even more convoluted operational procedures and protocols, when there is insufficient evidence that even the current more elemental protocols have been followed. The result of the 1965 and 1977 blackouts in the Northeast resulted in many fine reliability standards of operation and planning that were followed with very good results until relatively recently. Loop flows such as those onerously imposed on Michigan allow over scheduling of the grid on fictitious contract paths without regard to the consequences. Operational practices such as "parking" and "hubbing" of transactions (scheduling of transactions using intermediary third parties rather than transacting directly between buyer and seller), cause actual use of the grid to be cloaked. This is because the park/hub transaction, with its fictional flow of electricity, can fall beneath the screen whereas the original transaction would have been visible. Entities responsible for ensuring proper use of the grid ignore threats to reliable operation in response to pressure from market participants wishing unfettered use, regardless of actual infrastructure capability-to substitute operational procedures for infrastructure-to ignore the rules when it is advantageous.
7. How can similar incidents in the future be prevented? Mandatory reliability standards, developed and enforced by non-market participants, and funded independently of market participants are absolutely critical. Mandatory RTO participation is essential to ensure elimination of unscheduled loop flow. No seams between RTOs can be allowed, and no seams which overlay natural markets can be tolerated. Reliability plans such as the proposed MISO-PJM plan which embeds loop flows on the transmission systems of Michigan companies will virtually assure additional blackouts. The communications mishmash underlying the August 14 blackout must be unwound. MISO is Michigan's and FE's Security Coordinator, and PJM is AEP's Security Coordinator. Michigan companies are members of MISO but FE is not. (The Security Coordinator is the entity which oversees the reliability of the grid within his footprint, acts to ensure that action is taken to maintain safe and reliability operation, and communicates to other Security Coordinators within other regions to ensure overall safe operation of the grid). AEP is not a member of any RTO but the Southwest Power Pool (SPP) is AEP's transaction scheduler. PJM does not report its internal flows and circuits to the systems which allow tracking and unwinding of transactions when necessary to resolve overload problems; MISO does, but only within its footprint. Unfortunately, Commonwealth Edison (an Exelon operating company) (ComEd), for example, is embedded within the MISO grid, so that ComEd transactions across the AEP grid into its affiliate in PJM are not subject to MISO oversight. As part of PJM, ComEd flows are no longer visible to the entities outside PJM. While these flows contribute significantly to the loop flows through Michigan, they are no longer curtailable through the current TLR (NERC's Transmission Line Loading Relief) process. When these RTO configuration issues were first raised at the July 17, 2002 FERC meeting, NERC's Mr. Gent, in discussing the concerns raised, stated "is this the configuration as you would have designed it? Probably not. Is it the configuration that I would have designed? Probably not. But it is the configuration that the participants have chosen,,..Therefore, our recommendation to you is that you condition your approval of any configuration on the participants successfully convincing the industry, through our NERC Operating Committee, that reliability is not impaired." However, notwithstanding the forceful, unanimous, and continuous objections of the Michigan companies, the NERC Operating Committee and NERC regional council, ECAR, have approved, and continue to approve the proposed reliability plan. In fact, ECAR voted to approve the plan on August 15, 2003, while major areas of Michigan were still blacked out, when none of the Michigan companies were present. Ultimately, the safe and reliable operation of the grid can be restored by ensuring that the standards and procedures required to do so are developed and enforced, independent of market participants. Where the market desires transactions which the current grid cannot safely accommodate, new infrastructure investment must be made, rather than rely on luck and prayer. Some required infrastructure improvements will span multiple traditional utility footprints. Regulatory and rate changes will be required to get those facility investments made. Some of these investments will require significant time to obtain rights of way and address environmental issues. The institutional and regulatory changes I have described must come now so that the existing infrastructure can be optimized within its capabilities without repeating August 14. My findings are based on the data that we collected within Michigan which I will make publicly available. I urge that others do the same. At ITC, we chose to work in the open because our job is to serve the market to the benefit of all electric users.
<
Attachment 2
|
|||||||||||
|
