* Update* Where was the back-up power on Wednesday?

Earlier this decade when TXU Energy was pushing hard to build nearly a dozen new coal-fired power plants in Texas, much was made of the state’s “reserve margin”– the difference between the amount of power plant capacity on hand compared to what was expected to be the summer peak demand.

TXU and its supporters would point ominously to the annual reports from ERCOT that regularly predicted the reserve margin would dip dangerously low — below the then-mandated limits of 12.5 percent to as low as 5 percent — several years into the future. The message: if Texas doesn’t build these plants, we risk regular blackouts and turning Texas into a Third World nation.

That issue quickly dissolved when a group of private equity investors swooped in to buy TXU and promised to build many fewer coal plants. And, not surprisingly, subsequent ERCOT reports on reserve margins did what they usually do — push the crisis-level dates farther into the future as power plant builders continued to build new capacity to meet the state’s growing needs.

While the root cause of Wednesday’s rolling blackouts seems to be a massive failure of some 50 power generating units at once, it appears the state was well below the newly mandated 13.75 percent reserve margin going into the day, did ERCOT have enough power capacity waiting in reserve?

Technically, it appears so. These numbers are not exact, but here’s the math:

  • Texas has about 72,500 megawatts of power plant capacity.
  • ERCOT told me about 12,000 megawatts of capacity was offline going into Wednesday for regularly scheduled maintenance. This is work that has to be pre-approved weeks or months in advance. It usually happens in the cold-weather months.
  • That means about 60,500 megawatts of power were expected to be ready to go on Wednesday.
  • The projected peak demand for Wednesday morning was 57,200 megawatts. That means there was a 3,300 megawatts cushion, or a reserve margin of about 5.7 percent.
  • If you add the two demand-response programs ERCOT has, where large power users agree to shut-down in an emergency (which appears to be about 1,400 megawatts) you get about 4,700 megawatts or a reserve margin of about 8.3 percent.

Losing 7,000 megawatts of capacity — about 11.5 percent of what was expected to be online — is certainly unprecedented and unexpected. But was ERCOT operating under an unusually low reserve margin going into Wednesday?



As ERCOT has now explained to me, the daily operating reserves are not the same as the reserve margin.

The minimum margin required  for daily operating reserves is 2,300 megawatts, according to spokeswoman Dottie Roark.  ERCOT typically requires around 3,000 megawatts on line reserves and up to 2,000 megawatts of reserves than can be on line within 30 minutes. 

So in short, the day-to-day cushion is different from the long-term planning cushion. And ERCOT had at least that much power scheduled to be at the ready.

But is that enough?


According to a report from the North American Electric Reliability Council — a quasi-private entity that oversees all of the country’s power grid operators — ERCOT’s winter maintenance was not expected to top more than 4,568 megawatts at any one time this winter. If that was the case, ERCOT would have had about 10,700 megawatts of capacity in reserve, or a reserve margin of 18 percent.

What explains the difference?

The NERC study was done in November, so perhaps ERCOT allowed for many more service outages after the report was published.

It’s also possible I’m interpreting this data incorrectly. I contacted NERC earlier today (Friday) but have not heard back.

A spokeswoman for ERCOT said she wasn’t sure how NERC reached its numbers either, but she said the 12,000 megawatts of maintenance isn’t unusual: The last time Texas had rolling blackouts, in April 2006, there were 14,000 megawatts offline for planned maintenance.