| 7.12.04 | John Hanger, President and CEO,
Citizens for Pennsylvania's Future (PennFuture) Peter
Adels, General Counsel, Citizen's for Pennsylvania's Future (PennFuture) |
| Keeping Our Cool | ||||
|
||||
PJM is called upon to detail what the chances are that a blackout will strike some or most of PJM because peak electric demand caused by summer heat waves overwhelms supply. After crunching all the numbers about expected peak demands as well as generation and transmission capacity, PJM concluded that the chances of a blackout this summer are remote, no more than the typical probability of a blackout occurring one day every ten years.
PJM projects that it will have a reserve margin of 18.6 percent this summer, exceeding the required 16 percent reserve margin needed to keep the probability of summer blackouts remote. Unit operational reliability, demand side response, transmission adequacy, system expansion, imports and exports, and operations and communications are a key part of a useful current assessment. Indeed, the summer 2004 reliability assessment is good, better than last year. PJM anticipates record peak demand, reduced bulk sales out of PJM, increasing demand response program participation, and improved communication.
That's good news, but it is not a guarantee that a blackout won't happen during the summer of 2004. Murphy's Law still lives. Despite the best planning and operation, enough things can go wrong at the same time - for example six days of record temperatures in a row, especially in a world now affected by global warming, plus fuel interruptions or mechanical breakdowns - to cause what should be at worst rotating or planned blackouts that bring demand back into balance with supply. A cascading or uncontrolled blackout like that of August 14, 2003 should never happen.
Load Forecast
PJM is forecasting a summer peak load of 64,927 MW, higher than the all-time
peak of 63,762 MW experienced on August 14, 2002 (yes, but a different August
14). For 2003, PJM had forecast a peak load as high as 64,064 MW, although the
actual peak only reached 61,499 MW during a generally mild summer.
Bad forecasting last year? Hardly. Although the core of a reliability assessment, forecasts are essentially best guesses largely dependent on the inexact science of weather assumptions plus fairly accurate (short term) predictions of population growth and economic activity. In fact, on a day-to-day basis, PJM predicts electric demand with remarkable accuracy and uses its considerable experience to plan for each summer's peak load. The predicted peak and/or other high loads will be reached if a string of very hot and humid days in the middle of the week occur. If it doesn't get that hot and humid, and/or if it only happens on the weekend, there won't be a record demand.
Capacity
The primary way to meet the potential peak load is PJM's installed capacity base
of 77,030 MW. That helps produce a reserve margin of 18.6 percent more than is
needed to meet the forecast peak, well in excess of the current 16 percent
required reserve margin. And of course, it is way, way more than is needed to
meet load on most days that don't approach the forecast record peak.
The 77,030 MW is a net increase of 1,465 MW since last summer. There is 2,750 MW from new generation and 200 MW from increased rating for existing plants, but 1210 MW in retirements and 75 MW of decreased ratings. For the first time, the capacity includes 130 MW of wind energy. Wind capacity has been available for the last few years, but intermittent renewable resources such as wind were not previously eligible for capacity credits.
Forced Outages
Having sufficient generation available to meet demand means having plants that
operate. A plant that is built but not producing energy for PJM does not keep
the lights on. As has long been the case, planned outages are not permitted
during peak periods and scheduled maintenance is coordinated to minimize any
impact during peak periods. Unexpected "forced outages" always occur
though. Before restructuring, when owners of plants could get paid even during a
forced outage, forced outage rates were higher than ten percent. That meant that
we couldn't rely on about ten percent of installed generation to be able to
generate electricity at any given time.
Annual forced outage rates dropped steadily after 1996 from more than ten percent, reaching less than five percent in 2001. But forced outage rates have since begun to creep back up, although still less than 7.1 percent for all of 2003. During the hot summer months, 7.1 percent of PJM capacity is historically forced out of service.
While the forced outage rate remains historically low, many generators have been arguing that lower energy and capacity prices during the last few summers have made generation owners less willing or less able to make needed investments to keep their plants in tip-top shape. PJM staff is reluctant to speculate on why the data is what it is.
Demand Response
Voluntary participation in PJM sponsored load response programs is increasing,
with 2.9 percent of total demand enrolled in a load response program.
PJM now counts on 1,100 MW of Active Load Management (ALM), the capacity-based demand response program, to contribute to reliability. The ALM program provides participants a capacity payment for allowing PJM to interrupt their service. With plenty of supply, ALM was only initiated twice during 2000, 4 times in 2001, 3 times in 2002, and never during 2003. Yet, the program will provide 1100 MW of "capacity" during the summer of 2004, reducing the forecast peak load from 66,027 MW to 64,927 MW.
The demand response programs on the energy side also are growing in importance. PJM has 663 MW located at 257 sites enrolled in the Economic Demand Side Response Program that provides an incentive to reduce consumption when prices are high. In addition, PJM has 726 MW at 174 sites enrolled in the Emergency Demand Response Program that compensates customers for voluntarily reducing load during an emergency.
This continuing growth of demand response during a period in which the market is long on generation and prices are low is a testament to the significance and value of demand response. Effective demand response is a key to reliability. It also keeps down peak prices, thereby reducing the value of generation assets. To state the obvious, most but not all benefit from lower peak prices.
Size and Connections
The expanded PJM is better able to meet reliability partly because there are so
many more generating plants and transmission options, but also because the
extreme weather that produces peak demand is less likely to affect a very large
area at the very same time. The reserve requirement in PJM dropped to 16 percent
when PJM West became integrated. The lower reserve margin is a huge benefit of
restructuring for all consumers. Consumers get the same amount of reliability
with less generation capacity needed. Over time, this reduction in required
reserve margins will produce lower electricity bills than would otherwise be the
case.
Similarly, PJM may now import excess generation from neighboring areas. For this summer PJM is expecting to be able to import 1,100 MW from the West, 2,300 MW from the North, and 2,100 MW from the South. Conversely, PJM currently has only 871 MW scheduled to be sold out of PJM this summer, compared to the historic average of about 1200 MW.
So Close Yet So Far
All of the foregoing numbers reflect the original boundaries of PJM as well as
PJM West. Although the Northern Illinois Control Area (NICA) was added to PJM as
of May 1 to include the ComEd territory, NICA is not yet sufficiently integrated
to be a major factor in reliability planning. The new NICA region has reserve
sharing with the rest of PJM, but there is only a limited ability to transfer
generation between the two areas. (See E3, May 5, 2004, Land of Linkin.) There
is no mutual load shedding between the two regions.
But the expansion nevertheless is huge and will affect reliability more in the future. The forecast peak load for ComEd for this summer is 22,225 MW, with 505 MW of mandatory interruptible generation and deliverable generation of 28,645 MW. The interim capacity construct in place is different than the rest of PJM until after AEP is integrated (on track for this October) and is currently expected to yield a whopping 31.9 percent reserve margin. If ComEd has any problems this summer, it won't be due to any lack of generation.
Transmission Keys
Ample generation is not reassuring if the transmission lines can't get it to the
customers who need it. All of PJM generally uses the same "one day in ten
years" adequacy criterion: the system is designed to be adequate all but
one day every ten years. PJM expansion now makes things a bit more complicated,
though, as PJM now includes three different reliability councils that use
different specific criteria to meet the same standard. PJM reported to the
Pennsylvania PUC that the transmission system is adequate to meet the "one
day in ten" standards in each area - the Mid-Atlantic system based on
Mid-Atlantic Area Council (MAAC) criteria, the PJM West system based on East
Central Area Reliability Coordination Agreement (ECAR) criteria, and the NICA
region based on Mid-America Interconnected Network (MAIN) criteria.
Meeting the "one day in ten" standards is usually a no-brainer, but it became a bit of a last minute challenge this year. Although not part of the summer assessment report, a confluence of events that could not be predicted raised the possibility that the criteria would not be met in some areas, mostly in northern New Jersey. A critical transformer substation was derated due to damage found on the transformers. Reliant retired some generation units in New Jersey. There also were some delays in transmission upgrade projects. PJM and Public Service Enterprise Group (PSEG) staff were able to scramble to make some last minute upgrades and the criteria were met, but not without some anxiety. Some new generation also is about to come on line in the affected area, providing an additional cushion.
Lessons from August 14, 2003
PJM's presentation to the Pennsylvania PUC ended with a discussion of PJM's
response to the recommendations of the FERC/DOE investigation into the blackout.
As August 14 made clear, reliability is not only a factor of adequate
generation. The blackout had nothing to do with lack of generation, of which
there was plenty.
PJM certainly learned to walk with a stiff spine. The investigations pointed the finger at PJM for inadequate communication with neighboring areas. PJM was more aware of the events in the FirstEnergy territory that were leading up to the blackout than FirstEnergy. PJM's attempts to communicate its information to FirstEnergy didn't work very well. PJM certainly was a bit stung by criticism of efforts that PJM considered "above the call of duty" to inform FirstEnergy of problems in the FirstEnergy system.
But PJM also learned that improvement is still possible. PJM went through an extensive process of reviewing and modifying its communications protocols both internally and with neighboring regions to assure that critical information is communicated quickly and clearly to those who need it. To assist in this effort, PJM accelerated the implementation of sophisticated new visualization tools - computer generated dynamic maps of the PJM system and neighboring regions. These tools allow PJM operators to better monitor the status of the system and even more quickly identify the location and magnitude of potential problems.
PJM remains at the forefront of restructuring that supports both more economic and more reliable electricity systems. The level of communication between PJM and its member utilities, as well as its neighboring regions, is superb and getting better. From planning the system for the future to ensuring that problems are identified and resolved quickly, we are blessed with a system that can allow customers within PJM to enjoy the summer with the expectation that the lights, and air conditioning, and everything else will stay on.
Yet, the biggest lesson is that things just happen no matter how much we build and how much we plan. During those moments, distributed generation and demand response are the last line of defense. By the summer of 2005, both should be more deployed and developed than they are this summer. Then we'll know for sure that we can keep our cool.
To subscribe or visit this site go to: http://www.energypulse.net