Freezing Texans In the Dark: Engineering Rolling Blackouts

Last Wednesday, thousands of Texans experienced something that, to the best of my knowledge, is unprecedented in the history of the state. A combination of extremely low temperatures, weather-caused generating plant failures, and poor planning led to the need to cut off electric power for several hours or more in widespread regions of Central and North Texas. My house was in one of the affected regions, and so about 5:20 AM that morning I found myself hunting in the dark for flashlights and wondering if someone had driven into a power utility pole nearby. We experienced more blackouts intermittently the rest of the morning, and my university cancelled afternoon classes out of concerns that people would get trapped in elevators. Later that day I learned the reason: a short-term shortage of generating capacity forced the Electric Reliability Council of Texas (ERCOT) to order its member utilities to shed loads systematically in a series of rolling blackouts to prevent the whole system from going down in an uncontrolled way. There is both good news and bad news in the reports of why this happened and how decisions were made.

The good news is that the old “hold on till you can’t anymore” attitude that led in the 1960s to regional or national power blackouts is a thing of the past. With modern instrumentation and modeling software, operators can tell when their power grid is getting close to the brink and organize deliberate actions such as rolling blackouts to prevent a total system collapse. That is what operators in California had to do a few years ago during extremely hot weather and an energy crisis, and that is what ERCOT did last week. Relatively short local blackouts of a few “circuits” (distribution areas) at a time are much preferable to a disorganized collapse that affects everybody, including critical loads such as hospitals, rest homes, and semiconductor plants, for whom a power failure means endangering lives or the immediate loss of millions of dollars of product and equipment.

But the bad news is that we had rolling blackouts at all. Texas is unique among the 48 contiguous states in that its power grid is largely independent of surrounding regions. When other grids have trouble, this helps us get through unscathed, but by the same token, the state has to generate the vast majority of the power it uses within its own borders. Consequently there are some 500 or so generating plants in Texas, about 50 of which were off line last Wednesday for one reason or another. Reports are still coming in about why so many plants were down, but the most significant factor was the weather: it was 18 degrees F here in San Marcos, halfway between San Antonio and Austin, and proportionally colder as you went north. Evidently power-plant operators, whose machinery is mostly outdoors and exposed to the wind and icy temperatures, did not uniformly plan in advance for such low temperatures. Pipes froze or burst, machinery failed to start, and even many of the natural-gas-fired emergency plants designed for short-term supplemental use in just such a crisis couldn’t be started. The reason? Atmos Energy, the main natural-gas supplier, was having its own problems keeping gas pressure up to residential customers, so it exercised its contractual right to reduce pressure to large-scale industrial users, including—you guessed it—power plants. So we shot ourselves in the foot on that one.

ERCOT and its member utilities have since come in for a lot of criticism about the way the blackouts were distributed. It turns out that the firms had lists of “protected” loads which were not to be interrupted under rolling-blackout conditions: places like the aforementioned hospitals, nursing homes, and semiconductor plants. There were so many of these protected loads in so many circuits that the burden of the blackouts fell on a relatively few residential and commercial districts, with reports of some sections losing power for as long as eight hours.

What lessons can be learned from this experience? Surely a lot of power-plant owners are reviewing their cold-weather contingency plans, and the next time such an unusual cold snap hits I hope more plants will stay on line. Everybody now knows about the lists of protected loads, and after such public exposure perhaps a dialog about the wisdom of such lists can lead to improvements or changes if necessary. And clearly, just because something is in a contract doesn’t mean that it’s a wise thing to do. Cutting gas pressure to power plants in an emergency when you need more power plants, not less, is just the kind of bureaucratic messup that needs coordination at a higher level, perhaps with state government involvement if necessary.

But beyond these tactical issues lies a more strategic question: does this experience tell us something about changes in the level of commitment and planning in the electric-utility industry after several years of deregulation? Compared to thirty years ago, the industry is much more diversified, independent of government, and competitive, although these changes are only a matter of degree. The concern I have always had about extensive utility deregulation is that in the struggle for profits, the customer’s needs would be left behind. Under normal conditions this concern has largely proved groundless, and at least in many parts of Texas customers now have a choice about who they buy their electricity from. (That is not the case for people who live in cities that own the electric utility, such as San Marcos.) But the relationship between one customer and a particular electric provider proved illusory when ERCOT exercised what amounted to dictatorial control over the entire system to preserve its integrity.

On the whole, this control was exercised wisely. One wonders whether the problem would have occurred under the old regulated system of guaranteed profits, when generating, transmitting, and distributing equipment was typically under one ownership and profits were generous enough to allow overkill in maintenance and cold-weather protection, as well as a little surplus for extravagances like research, for instance. We will never know. I confess a little hurt pride at the thought that rolling blackouts, which I associate (rightly or wrongly) mainly with developing countries, actually affected my home state of Texas. I hope this is not a trend, and that the lessons learned from this unique experience help us avoid another one in the future.

Sources: I used reports from various issues of the Austin American-Statesman over the last week (see http://www.statesman.com/ for specific reports on the rolling blackouts of Feb. 2, 2011).