Advertisement
Opinion

ERCOT is naive about how much power Texas can generate when we really need it

Texas needs better planning and execution from ERCOT.

This op-ed is part of a series published by The Dallas Morning News Opinion section to explore ideas and policies for strengthening electric reliability. Find the full series here: Keeping the Lights On.

Texas has a serious problem: not enough electricity when we really need it. While a lot needs to be fixed, three issues need immediate attention.

The Electric Reliability Council of Texas is responsible for managing the flow of electric power to meet about 90% of the state’s electricity demand. During the recent severe winter weather event, ERCOT planning and execution issues proved problematic.

Advertisement

Let’s first examine two planning problems.

Opinion

Get smart opinions on the topics North Texans care about.

Or with:

The first deals with ERCOT naiveté. According to ERCOT, the Texas grid has a little more than 100,000 megawatts of “installed” capacity, of which about 28% is made up of intermittent renewable sources — wind and solar. ERCOT’s baseload capacity comes from coal and natural gas, together making up 67% of installed capacity, and nuclear, which is about 5%. ERCOT includes a significant portion of each of these sources when calculating what’s called the reserve margin, or the amount of capacity that exceeds demand.

In early November 2020, ERCOT issued its Seasonal Assessment of Resource Adequacy report for this winter. “ERCOT anticipates there will be sufficient installed generating capacity available to serve system wide forecast peak demand this winter season, December 2020-February 2021,” ERCOT said. The assessment estimated that “nearly 83,000 megawatts of resource capacity is expected to be available for the winter peak,” well in excess of the “57,699 megawatts” forecast for peak demand.

Advertisement

Then came the arctic cold.

As temperatures began to plummet across Texas on Sunday, Feb. 14, ERCOT projected that demand would peak at about 74,000 MW, significantly higher than originally forecast. As night approached, wind turbines began to freeze, some thermal plants tripped offline due to the extreme cold, and natural gas production and delivery were constrained, causing additional thermal plants to fail.

By 1 a.m. Monday, Feb. 15, coal and natural gas-fired generation were supplying 79% and 77% of their installed capacity, respectively, while wind was supplying 21% of its installed capacity. By Monday night, those numbers had dropped. Coal and natural gas-fired generation were supplying 59% and 56% of their installed capacity, while wind was supplying only 3% of its installed capacity.

Advertisement

It is a serious planning mistake to include renewable sources when estimating how much power — actual generation — will be available at times of peak demand. When a thermal generator is unable to meet its obligation to serve demand, the generator must pay a penalty. Renewable resources have no such obligation. If the wind is not blowing or the sun is not shining, these generators have no obligation and pay no penalty. That has to change.

A second planning problem is somewhat related to the first. Texas has been exceptionally successful in promoting the development of renewable energy resources. When the wind is blowing and the sun is shining, Texans reap the benefit in the form of cheap electricity. But this comes at a cost in the form of lost baseload capacity, that is, power plants that generate electricity constantly. In the last five years alone, some 8,500 MW of baseload capacity has closed down. Those plants will not come back.

As the Texas population and economy continue to grow, we must build more baseload capacity. Winterization measures alone will not solve the problem. At a minimum, Texas needs to build 10,000 MW of additional thermal generation.

To incentivize investment in new baseload generation, changes in the Texas energy-only market may be required. Whether or how the market should be tweaked is a question for policymakers. But regardless, the decline in baseload generation must be reversed — and soon.

Now let’s turn to a potential execution problem.

In the U.S., all power plants operate on an established frequency of 60 hertz. Think of it as a spinning top designed to remain upright at exactly 60 revolutions per second. If the frequency of power plant turbines moves above or below 60 hertz for just a few minutes, really bad things can happen. Turbines and generators can overheat, vibrate severely, become inoperable and even explode.

About 1:50 a.m. Monday, Feb. 15, the frequency on the ERCOT grid began to decline. Generation capacity was falling, and demand was growing. The ERCOT control room operators had an emergency on their hands.

During the Texas House and Senate hearings of Feb. 25-26, industry executives testified that as ERCOT took action to bring the frequency back up by abruptly and significantly reducing demand (called shedding load), some power plants had to trip themselves offline due to the frequency instability on the grid. This happened not because of the cold conditions, but because ERCOT did not act to methodically shed load earlier to stabilize frequency. As a result, ERCOT come dangerously close to a catastrophe that would have shut down the entire grid for weeks.

Advertisement

An independent third party should be retained to conduct a thorough investigation of exactly what was happening in the ERCOT control room before and after 1:50 a.m. on that Monday and if the operators had full understanding of the operational risks that generators face when frequency changes that fast. If the investigation concludes the ERCOT operators had the appropriate knowledge and performed as they should have, this question will be resolved. However, if such an investigation finds mistakes caused thermal generation to automatically trip offline due to the frequency event, then significant changes should be made in ERCOT operational protocols.

In sum, three big issues — two dealing with planning and a third dealing with real-time execution — need to be addressed and fixed. Failure to do so is not an option.

Phil Wilson is the general manager of the Lower Colorado River Authority. He wrote this column for The Dallas Morning News.

Got an opinion about this issue? Send a letter to the editor, and you just might get published.

Advertisement