The Unruly Grid: One Year Later
Image: Lester Lefkowitz/Corbis
A year after the U.S. electricity blackout of 14 August 2003, the lights are back on and everything seems to be humming along smoothly, at least as of this writing in early July. But have the proper corrective actions in fact been taken? The blackout certainly seemed a wake-up call at the time, affecting a record 61 800 megawatts and costing consumers an estimated US $7 billion. Even more troubling, it came in the wake of the widespread U.S. power outages in 1999 and the two big western U.S. outages in 1996.
Altogether, there have been 130 significant blackouts since 1999 alone, by the U.S. Department of Energy's reckoning. And that's not counting the California electricity trading crisis of 2000-2001.
At least in principle, last summer's system failure ought to have prompted disciplinary measures against the parties now known to have caused it, if only to send a clear message to other utilities. The final U.S.-Canada report on the blackout, issued in April, found four major categories of causes, the first three associated entirely with FirstEnergy Corp., the Akron, Ohio, utility responsible for the generators and transmission lines where all the trouble started. FirstEnergy was charged, in descending order of importance, with failure to assess and understand inadequacies in its own system, absence of effective contingency planning, and rudimentary neglect of tree trimming in its transmission corridors. The language was blunt, as in point 1(A): "FE failed to conduct rigorous long-term planning studies of its system, and neglected to conduct appropriate multiple contingency or extreme condition assessments."
Though the U.S.-Canada report concluded that the blackout was wholly preventable, the reason nobody has been called to account is not hard to find. Since the late 1990s, the North American Electric Reliability Council (NERC) in Princeton, N.J., has sought the authority to enforce its own rules--the rules North American utilities currently adopt purely on a voluntary basis. But NERC's efforts to put teeth into its regulations came to grief earlier this year in the pettiest of partisan squabbles over long-stalled national energy legislation.
Speaking about the 2003 blackout at a Dallas meeting sponsored by IEEE's Power Engineering Society last September, NERC president Michael R. Gent said that "if we had had a solid compliance program in place to see to it that rules were being followed, we would not have had this blackout."
Of course, electric reliability is not just a U.S. problem. Within two months of the U.S. outage last year, the entire Italian power system went down, and there were big blackouts in London, and in Denmark and southern Sweden. But the United States is unique in having a power system that is neither fully deregulated nor fully regulated. In this legal limbo, despite a lot of promising new technology that could make the nation's grid systems much more secure--smarter grids, faster and more flexible controls, better backup and storage, automated overload protection systems--nobody knows who ultimately will own what. As a result, as one eminent expert puts it in this issue, you'd be "silly" to put any serious money into grid innovation. So the U.S. power industry lumbers on, spending less on R and D than pet food manufacturers do.
To be sure, even if last year's outage had jolted everybody into doing everything possible to avert another one, it is also possible that such big system failures are inevitable. In a provocative article in this issue, "The Unruly Grid," Contributing Editor Peter Fairley reports on a new system analysis in which it is argued that the biggest power outages will stubbornly keep happening. If that's right, it might be best if everybody is prepared for the occasional breakdown--then, ironically, the largest power failures would actually be less likely because local operators like FirstEnergy would be quicker to turn off the lights, to stop the cascading outage in its tracks. Yet even simple calls for concerted national blackout preparedness have been lacking. Under the circumstances, it's a good idea to keep some flashlights on hand, stash cash, and rummage around in the basement for that old windup clock.
Scientists who ponder the origin and fate of the universe naturally have to deal with some very big numbers, as Alexander Hellemans and Madhusree Mukerjee show in this issue with their article "Computing the Cosmos." But sometimes the names of these numbers can cause confusion, especially in the case of "billion," which in some parts of the world is read to mean a million million, or 1012. The dominant trends, however, in English-language publications is that a billion means a thousand million, or 109, and a trillion means 1012. With apologies to those who prefer to supersize their billions and trillions, IEEE Spectrum is formally embracing these trends. A billion will be 1000 000000: no more, no less.
The editorial contents of IEEE Spectrum magazine do not represent official positions of the IEEE, its entities, its volunteers, or its members. Please address comments to Forum at