By Peter Hartley, Ph.D.
George A. Peterkin Professor of Economics, Rice University; Baker Institute Rice Faculty Scholar, Center for Energy Studies
Jim Krane, Ph.D.
Wallace S. Wilson Fellow for Energy Studies, Center for Energy Studies
Michael Maher, Ph.D.
Senior Program Advisor, Center for Energy Studies
and Kenneth B. Medlock III, Ph.D.
James A. Baker, III, and Susan G. Baker Fellow in Energy and Resource Economics; Senior Director, Center for Energy Studies
The mysterious causes behind February’s disastrous power failures are finally becoming clear, but important questions remain.
Grid operator ERCOT’s April 6 preliminary report pins much of the blame on sheer unpreparedness of the state’s energy and power generation infrastructure, mostly for cold weather.
But the report also contains revelations that should give Texas legislators reason to pause their proposed remedies until the full causes of the crisis are understood.
ERCOT’s report, based on a survey of its generation providers, demonstrates that early attempts to assign blame were misguided. Most tended to focus on single causes when many different failures produced the complex disaster.
The biggest cause of unplanned outages were factors ERCOT labelled “weather related.”
Among them were “frozen sensing lines, frozen water lines, and frozen valves—ice accumulation on wind turbine blades, ice/snow cover on solar panels, exceedances of low temperature limits for wind turbines, and flooded equipment due to ice/snow melt.”
Elsewhere, “equipment issues” knocked out power for reasons other than just cold air. They included plants tripping offline or ramping down due to control system failures and turbine vibrations.
Some of these breakdowns probably arose from the strain on plants from the unfolding disaster. The dangerous drop below normal 60Hz electrical frequency on the grid in the wee hours of Feb. 15 caused many plants to trip automatically to prevent mechanical damage. This was a big factor in getting the problems started, but according to ERCOT was a small overall contributor to the outage.
Also, since winter demand is typically low in Texas, about 15% of generating capacity was offline for routine maintenance or seasonal shutdown ahead of the peak summer season. Having so much capacity unavailable was painful.
Another point of failure was the drop in natural gas deliveries that knocked out plants that were otherwise operable. Power outages took down several electric compressor stations that were not identified as critical infrastructure, a major oversight given gas’ dominance in the Texas power sector. The lack of compression prevented pipelines from transporting enough gas to keep plants in operation, causing even more blackouts. This is an obvious, and very addressable, regulatory failure.
Unfortunately, a lack of public data prevents us from being sure how big a factor the gas deficiencies were, and where the key failures took place. For instance, storage send-out capabilities in Texas should have been sufficient, even with wellhead freeze-offs in the Permian basin. But we still don’t know when, where, and why compression failed, or whether there would have been enough gas to supply all the power plants and still meet heating demand. This knowledge is critical to understanding how gas flowed during the winter storm.
In short, pinpointing the full causes of these failures is ongoing. ERCOT’s report is too brief to provide sufficient guidance for new policy.
The report unhelpfully bundles all generation forms together and measures outages by deviations from a plant’s maximum output. A more useful report would give us the deviation from its rated output potential in the winter season. In other words, we want to know how “unusual” a plant’s performance was during the cold snap.
For example, ERCOT acknowledges that wind and solar output “is typically much lower than the specified nameplate capacity” so the outages reported are “much higher than the actual amount of power that would have been available.” It is also true that the more “reliably unreliable” a generation source is, the more backup capacity is required to keep the system operating during extreme events.
So, even though gas generation exceeded that of a “normal” winter day, it still fell short of what was needed during extreme weather.
Finally, a major breakdown in market design went unmentioned in the report. ERCOT’s scarcity pricing incentives did not convince generators to deliver power when Texans needed it. Something like two-thirds of the outages were due to preventable problems, mainly equipment that stopped working because of the cold.
The way Texas addresses such problems now is by allowing generators to sell power for as much as $9,000 per megawatt when blackouts are used to force load reductions. This compares to normal rates of $25/MW.
This add-on incentive is meant to encourage generators to keep emergency “operating reserves” available. Evidently, $9,000/MW was not enough to convince plant owners to harden their plants for tough conditions, or even to convince them to store backup fuel.
The legislature could improve those incentives by tweaking market design.
Policymakers should require plant owners to nominate “operating reserve” capacity in advance. Only “nominated” capacity would get the add-on payment when it is called upon in an emergency. That enticement should be bundled with a penalty: Firms with nominated reserve capacity should be charged $9,000/MWh if their plant is unavailable when called on.
Such a mechanism would strengthen incentives to weatherize and maintain resilient plants that can supply power on short notice. Our suggestion would be for legislators to set the parameters—incentives and penalties—but leave it to firms to figure out the technical details of hardening their operations.
The Berkshire Hathaway Energy proposal for additional gas-fired power generation misses the mark, since installed capacity was not an issue, operable capacity was. But the state’s gas-fired generators may want to explore Berkshire’s proposal to co-locate gas storage near power plants.
We hope Texas legislators now taking action on the freeze will seek more details before finalizing laws under consideration. The serious failures in February should not be treated as a political issue. At a minimum, lawmakers need data on peak gas availability for simultaneous spikes in heating and power demand, along with weather-related outages by generation type, and how much those outages deviated from seasonal norms.
This information should point to the true causes—and remedies—worthy of their attention.
This post originally appeared in the Forbes blog on April 9, 2021.