Just after the sun rose on January 19, 1994, the mercury fell to new record low temperatures. Lights went out, furnaces shut off, and computer screens went blank in parts of Delaware, Maryland, New Jersey, Virginia, Washington, DC, and Pennsylvania. The 11 electric utilities that compose the PJM power pool had begun rolling blackouts (em the first ever in winter, and the first since the summer of 1970.
Looking back at this extraordinary event reveals that much of the ensuing conventional wisdom about the rolling blackouts of 1994 is wrong. Looking back also provides valuable insight into the future of the electric industry as customer choice reshapes it.
The multistate PJM power pool was formed to increase reliability and efficiency through central and economic dispatch of the pool's 57,626 megawatts (MW) of installed capacity. Economic dispatch saved consumers millions of dollars. But the pooling of capacity made electric service so reliable that too many people took it for granted. It was, then, a shock to discover on January 19 that the PJM operators could not meet the morning peak of 40,554 MW. The five Pennsylvania electric utilities in PJM experienced the greatest difficulty in meeting demand. At the urging of the Pennsylvania Commission, the governor declared a state of emergency.
Members of the Pennsylvania Public Utilities Commission (PUC) joined with officials of the Governor's office and colleagues in other states, as well as the federal government, to discuss how best to conserve electricity. Each affected jurisdiction then took measures to reduce demand. This regional cooperation was important, because the rolling blackouts were apportioned according to the ratio of each company's installed capacity to the pool's total capacity. If a company's generation constituted 15 percent of the pool's capacity, that company dumped 15 percent of the demand that could not be supplied, even though it may not have been responsible for the regional capacity shortage. No company could escape the burden of blackouts if the regional pool's supply of electricity was inadequate to meet demand.
If we take for granted the reliability of the electric system, we do so at our peril.
What Caused the Blackouts?
Some say the supply emergency occurred because too little generation had been built. Others point to transmission constraints. Still others argue that the extreme weather caused the emergency. These answers are either flatly wrong or too simple.
First, the daily peak on January 19 hit 40,554 MW (em 17,072 MW less than the total built capacity of 57,626 MW. The installed reserve was 42.1 percent. From January 17 to 21, the weekly peak was 41,351 MW, which occurred on the evening of Tuesday, January 18, about 13 hours before the blackouts. That weekly peak was met without resort to rolling blackouts, though other emergency procedures proved necessary. PJM's historic peak was 46,429 MW; it too was met without rolling blackouts. Clearly, the blackouts were not caused by too little installed capacity.
The blackouts also were not caused by transmission constraints. On January 19, PJM utilities bought 4,124 MW from utilities outside the pool. At crucial moments, however, not enough power was available in the bulk-power markets, because many utilities in a very wide geographic area were coping with heavy demand and operational problems.
The real cause of the emergency was the incredibly high number of PJM units that were forced out of service. Of PJM's total installed capacity of 57,626 MW, 28 percent (16,248 MW) of PJM's capacity was unavailable on Monday, January 17. On January 18, the day of weekly peak demand, 30 percent (17,289 MW) of total capacity was unavailable. On January 19, the day of the rolling blackouts, unavailable capacity rose to and peaked at 37 percent of total capacity (21,223 MW). Of this figure, 16,248 MW had been forced out of service. The rolling blackouts coincided with the peak in the amount of capacity unavailable, not with the peak in demand.
At 37 percent of total capacity, unavailable capacity on January 19 rose a crucial 4,000 MW more than the day before, which meant that the daily peak could not be met without blackouts, even though it was less than that of the day before. Subsequently, on January 20 and 21, unavailable capacity amounted to 30 and 32 percent, respectively. No rolling blackouts were necessary on either day.
Why was so much capacity forced out of service?
Load Was Firm, Fuel Wasn't
It is too simple to say that cold weather caused the outages. This answer also begs the question of why electric generation could not operate under the severe conditions.
The forced outages stemmed from two main causes: 1) fuel supply and handling problems, and 2) mechanical failures. Plants were forced out of service because fuel either was not available or could not be handled efficiently in the cold weather. Fuel supply and handling problems occurred in oil, coal, as well as gas plants. During the rolling blackouts, coal plants generated 17,924 MW, about 33 percent of PJM's total capacity; yet coal plants accounted for 44 percent of the capacity forced out of service. Further, 60 percent of PJM's coal capacity was operational during the blackouts.
Obviously, performance at the coal plants was not adequate, though it was three times better than the performance of plants where gas was the primary or sole source of fuel. The fuel problem at coal plants was rarely no coal or too little coal. Plenty of coal was usually on site or being delivered by rail, but it sometimes froze, making handling and plant operation difficult. Steps to remedy these problems have been taken.
Plants where gas was the sole or the primary fuel provided about 8,642 MW of capacity, about 15 percent of PJM's total capacity; yet gas-fired capacity also accounted for about 44 percent of the capacity forced out of service. Consequently, although a proportionately small amount of PJM's total installed capacity, gas plants formed a disproportionately large amount of the capacity forced out of service. No more than 20 percent, and probably quite a bit less, of PJM's gas-fired capacity was operational during the blackouts. If 60 percent of gas-fired capacity had been operational, which was the coal plants' performance, rolling blackouts would probably have been unnecessary.
Based on these numbers, the performance of PJM plants using gas as the primary or sole fuel on January 19 can only be judged unacceptable. The problem at these plants was not fuel handling or transportation; gas that was available for electric generation was handled and transported through pipelines without difficulty.
The operational problem began with the loss of gas supply at plants using gas as the primary or sole fuel. Many of these plants were served by interruptible transportation gas contracts. For the 1,154 MW of generation that relied solely on gas, interruption meant no generation of electricity. Since the four rolling blackouts on January 19 were four separate
500-MW load dumps (em never more than 1,500 MW at one time (em the gas-only plants by themselves nearly made up the entire difference between an electric system under great stress and one in crisis.
We cannot ignore the threat to reliability posed by plants that use only gas and are supplied with interruptible contracts, even though such plants amount to a small percentage of total capacity. When a system is stressed, reserve margins shrink precipitously, and the margin for reliable operation may be just a few percentage points of total capacity. Every megawatt counts in an emergency.
At gas-fired plants where gas was the primary fuel, interruption of gas supply should have meant switching to an alternate fuel. But attempts to switch to backup fuels frequently failed for a variety of reasons: The backup fuel gelled. Oil ran out, and transportation problems made it difficult or impossible to replenish supply. Mechanical problems cropped up when plants tried to switch from gas to an alternate fuel.
In fact, the mechanical problems at combustion turbine units were so pervasive, and the plants so unreliable, that PJM now assumes 50 percent of this installed capacity will not start when needed (em a remarkable planning concession to the realities of unreliable equipment.
All these problems merged in a compressed period and pushed the people and businesses within the areas served by the PJM power pool toward the precipice. Only decisive government action, the cooperation of the public, the work of electric utilities and other power producers, the National Guard and Coast Guard deliveries of oil through frozen rivers and roads, and a slight warming trend prevented a plunge over that precipice. t
John Hanger, a Commissioner with the Pennsylvania Public Utilities Commission, experienced the PJM emergency first-hand. This article incorporates the results of the PUC's investigation as well as his own research.
Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.