During the recent winter storm Heather (January 13-16), the Texas grid was able to produce enough energy to meet the demand, but just barely. Since winter storm Uri (February, 2021), it appears that improvements have been made to the reliability of natural gas generation. Also, there have been significant additions of renewables capacity, primarily solar, and a small amount energy storage (batteries). All of these helped to prevent another grid collapse. But we also got lucky. This storm was not nearly as severe as Uri in its intensity, duration, and precipitation.
During the storm, Texans used a wintertime record amount of electricity at 78,138 megawatts on January 16 at 7:50AM.1 At that time, ERCOT’s data showed that the system had 5,229 megawatts in Physical Responsive Capability (PRC).2 ERCOT begins calling alerts when operating reserves drop below 2,500 MW. At 1,500MW. ERCOT begins "controlled outages", aka rolling blackouts.
So, it appears to me that at the worst of the storm we had about 7% in reserves. The amount of reserves as a percentage of the total demands varies dramatically, even within a day, but typically, runs in the 10-20% range. So while we were never in danger of the grid failing this time, I think it is likely that if Heather had been as bad as Uri, the grid would have likely failed again, but perhaps not as badly as it in Uri.
The recent storm did, however, once again underscore the dilemma Texas faces with its current mix of generation sources. The industry jargon for this is called the “fuel mix.” Texas has the most diverse fuel mix in the country, and many may be surprised to know that Texas generates the highest level of renewables in the country, with wind and solar contributing over 30% of total generation.3
The Texas grid was primarily designed to perform in the summer heat when demand peaks are generally limited to several hours late afternoon and early evening. In the winter during periods of extreme cold, demand peaks can be much longer. In the case of Uri the demand peak lasted for several days.
Wind and solar perform better in the summer. Typically, renewables do not perform very well during winter storms. Because of the poorer performance of renewables and prolonged periods of high demand, the Texas grid is much more at risk of a catastrophic failure in the winter.
The contribution of wind and solar fell close to zero for several hours during Heather and for several days during Uri. For example, on January 15 at 7:00AM, wind was only producing 4,700 MW and solar was at zero. At that time, wind and solar were contributing less than 7% of the electricity Texans were using.
In these situations, the load falls back on other sources, primarily natural gas. Several times during the recent storm, natural gas was generating almost 70% of Texas’ electricity.
This dynamic means that for a reliable grid we must have sufficient dispatchable generation capacity to make up the difference when the sun is not shining, and the wind is not blowing. Simply building more wind and solar will not solve the reliability problem, in fact, more wind and solar capacity would exacerbate it. For now, at least, we need more natural gas generation to insure reliability.
But when the wind is blowing and the sun is shining, wind and solar produce electricity at a significantly lower cost, making it difficult for natural gas to compete during those times. For wind, the cost advantage is almost entirely attributable to government subsidies. But solar comes in a little below natural gas without any subsidy and the cost of solar continues to drop.
In a market-based system, like ERCOT, the competitive problem for natural gas is even more challenging during times when the wind and solar are producing a lot of electricity. That is because wind and solar have high upfront capital costs, but the marginal (input) costs are nearly zero. Even when electric prices are very low, wind operators are incentivized to continue to produce because their primary government subsidiary is based on the amount of electricity they produce.
In some cases, wind operators will continue to produce at even at negative market prices (i.e. the grid is oversupplied with electricity) because they still collect the production tax credit. In contrast, natural gas generators, which have significant fuel and operating costs, cannot afford to continue to produce when the price drops below their input costs.
As a result, there is very little incentive to invest in natural gas generation in Texas today. Texas recently offered a package of incentives to build additional natural gas generation, including low-rate loans. There were no takers.
So, if Texas want to be assured that it will have adequate dispatchable power for extreme weather conditions, the State is going to have to subsidize it with taxpayer money or require it by regulation which would ultimately increase rates. I will leave the irony of one government subsidy causing the creation of another government subsidy or additional regulation for another day.
The only other alternative is to build storage capacity for the electricity. Currently, that means batteries. At one point in the recent storm, batteries were contributing over 1,000 MW. The all-time record for battery contribution to the grid was 2,172MW in February, 2023. So, currently, batteries are making a very small contribution.
There is a building boom in battery storage as electricity users have become increasingly suspect of the grid’s reliability. And we are likely to see that number continue to grow. However, as you can see on the chart above, the cost of battery storage is very high. That will probably come down over time, but it has a long way to go.
Because of cheap natural gas prices and the significant contribution of wind and solar, Texans enjoy some of the lowest electricity prices in the country, running 18% below the national average. But that low cost comes at a price – an inherent lack of reliability in its grid. And it is a savings that is wiped out many times over when a storm like Uri does billions of dollars in damage, not to even mention the human toll.
The Texas grid held in this storm, but the long-term issue of building a reliable grid is Texas in far from over.
_________________________________
Note 1 – Much of the data I have used in this post comes from the Gridstatus.io website. It is an incredible resource and much more user friendly than trying to decipher the ERCOT website and data.
Note 2 - ERCOT reports "operating reserves" and PRC. Historical data on PRC is readily available but I have been unable to find it for operating reserves. There appear to be some technical difference between these two metrics, although I was unable to get a clear explanation from ERCOT of the difference. But from what I was able to discern, PRC is a good proxy for how close the grid is to having outages.
Note 3 – As you can see coal and nuclear contribute significant shares to Texas’ generation. However, it is clear that no new coal plants will be constructed in the United States and the future of nuclear is very much up in the air. There is a lot of discussion on about some new nuclear technology known as small nuclear reactors (SMRs). However, only SMR project I am aware of in Texas is a plan the Dow Chemical has to build three SMRs to powers it Freeport complex.
