Two months ago, on the evening of February 26, power grid controllers at the Electric Reliability Council of Texas (ERCOT) found themselves in an uncomfortable position: they were rapidly running out of power. As consumption outstripped supply the frequency of the alternating current — nominally 60 hertz — began to slide, threatening to damage utility and customer equipment. At 6:41 pm the grid controllers declared a grid emergency and began ‘shedding load’ to restore the grid frequency. Which is to say, they shut off the power to some customers.
These customers had agreed in advance to participate in such “demand-response” situations and would be compensated for their trouble. Nevertheless, saying no to a buyer is as much a measure of last-resort for the power industry as for any other.
Wind power got the blame early on, because wind turbines in West Texas were delivering less power than their operators had projected. But subsequent studies showed that other factors were more important. In the 40 minutes leading up to the emergency conventional power plants delivered 350 megawatts less than they had promised, while wind generation slipped just 80 megawatts relative to plan. At the same time consumption rose by a whopping 1,185 MW more than ERCOT had forecast. ERCOT’s report to the Public Utility Commission of Texas highlights that electric load growth as a key cause.
Still, smarter integration of Texas’ wind power could have prevented the trouble. As ERCOT’s report shows, an independent wind power forecast prepared for ERCOT on February 25 under an ongoing pilot project predicted the February 26 wind power drop with “good fidelity.” Unfortunately ERCOT’s grid operators never saw the forecast and hence could not take steps in advance to ensure that alternate power supplies were available.
My story on TechReview.com today, Scheduling Wind Power, shows that grid controllers increasingly get the message: Integrating wind forecasting into grid planning is not only key to reliably accomodating much greater levels of wind power. It will also maximize the pollution reductions achieved in the process.