Rising gas prices spark a rush to wind farms, straining grid capacity and raising larger issues about market design.
When the Public Utility Commission of...
Industry experts debate whether so-called "price mitigation measures" miraculously solved the California crisis.
percent of the outages have to represent an economic decision" by the generators, he says. Even if the typical outage level was around 2,400 megawatts, and one assumes 50 percent more outages due to running plants at full capacity for extended time periods, the outage figure should have jumped to 3,600, Moore explains. Even at 100 percent more outages, the outage figure would rise to 5,000 megawatts. April 2001 saw 14,900 megawatts offline, which is nearly one-third of the state's total load.
The five major generators in California showed very low levels of generation in 2000, according to Robert McCullough, principal of McCullough Research. Of the 20,000 megawatts of gas-fired plant capacity in California, roughly 50 percent was in operation then, he says-a situation that lasted until the onset of the caps. McCullough acknowledges that reasons for taking plants offline include environmental restrictions, gas price increases, and breakdowns. But he notes that "by any normal standard, the dispatch rate was low."
Data that McCullough gathered from the California Independent System Operator (CAISO) lends some credence to the theory espoused by many regulators that generators were gaming the system by taking generation offline to create a perceived supply crisis, resulting in sky-high spot prices. In the entire month of May, 30 days saw forced or unplanned outages of better than 3,000 megawatts. The only day in May with less than 3,000 megawatts of unplanned outage was May 31, with roughly 2,800 megawatts-two days after the May 29 FERC price cap decision. In contrast, during June, there were only 11 days with outages over 3,000 megawatts, with 10 days between 2,500 and 3,000, six days between 2,000 and 2,500, and three days below 2,000 megawatts of unplanned outage. July saw even lower levels of forced outages, with only five days of more than 3,000 megawatts offline.
"There is a strong implication that FERC did little to control wholesale price, but much to change the [market] incentives." McCullough says. "In a perfect world, you expect availability to fall in November and May-these are shoulder months, used for repair and maintenance. In 2000, generators took down units in November and never brought them back up. It would be surprising to have extreme maintenance problems last year," he observed. He noted that an entire plant can be built in about six months from start to finish.
Despite the suggestiveness of the data, some regulators are withholding judgment on whether generators gamed the system before price caps. "You can reach the conclusion that caps gave market the right signal, or it could be that old plants were taxed by running at capacity for extended periods," says Richard Bilas, one of the CPUC's commissioners. He also points out that there has not "been anything other than allegations-you need an evidentiary hearing, for generators to be put on the record." Bilas says he "suspects deep down that the generators colluded-but how do you prove it?" Even if solid evidence showed that generators gamed the system, Bilas says there are better ways to signal generators not to manipulate the market. "Caps