Last month’s electricity outages in Texas meant that some consumers saw power costs spike by several thousand percent. What happened, and what have we learned?
Winter Storm Uri hit Texas unexpectedly on February 13–17, wreaking havoc across the entire state. Grocery shelves emptied as people stocked up; more than four million residents went without power and water; and dozens perished due to extreme cold, lack of heat, and other related tragedies. Some residents who did have access to power saw exorbitant electricity bills in the aftermath.
What happened? Could any of these adverse events have been prevented? What can we do to prepare for such disasters in the future? Especially given that the frequency and severity of storms are expected to increase due to climate change, we need answers to these questions. RFF scholars Kathryne Cleary and Karen Palmer discuss all this and more in the Q&A below.
Resources: For those Texans who have been able to access electricity during the widespread blackouts, why have their power bills risen so sharply?
Kathryne Cleary: Texas, like other parts of the country such as New England, has a deregulated electricity market with retail choice for their customers. Consumers in Texas must select a supplier, which leads to variety and competition. An example of this variety of service offerings is the option to purchase electricity at a variable rate, with the rate tied to wholesale market prices. One supplier in Texas offering variable rates (and which has come up a lot in the news) is Griddy; their customers saw a spike in electricity bills. Notably, however, most Texans opt for fixed rates, and thus did not face significant price spikes last month—but a small minority purchases electricity at variable rates.
Other grid regions, such as PJM, have suppliers that offer variable rates tied to wholesale market prices, too. But what makes Texas different from these other regions is its wholesale market structure. Texas relies on its energy-only market to encourage investment by allowing the market to reach very high prices in times of scarcity. Such price extremes are not allowed in other markets; consequently, customer bills tied to wholesale market prices can spike more significantly in Texas than in other areas of the country.
Wholesale prices spike when demand is very high or when supply is very low. What we saw during last month’s events in Texas was both a supply issue and a demand issue, because of the weather. Supply was very low because many power plants froze, and supply chains had problems, particularly for natural gas; in addition, demand was very high for electricity because people needed to heat their homes during the extreme cold. High demand compounded by shortages in supply led wholesale electricity prices to skyrocket—in some cases reaching the market price cap of 9 dollars per kilowatt-hour (which is $9,000 per megawatt-hour, as shown in Figure 1) and remaining that high for many hours. For reference, the average wholesale price in 2020 was about 2.2 cents per kilowatt-hour ($22 per megawatt-hour), so the wholesale price during the storm was several hundred times higher than normal, with this increased price reflected in the bills of Texans subject to variable rates.
Figure 1. Average hourly wholesale price each day in February for electricity on the Electric Reliability Council of Texas (ERCOT) market, per megawatt-hour
All this is not to say that variable rates can’t work. Most of the time, variable rates aren’t a problem, because wholesale rates do not typically fluctuate to the extremes we saw last month. Customers subject to time-varying prices can typically save money during periods of low prices, which occur when there’s a lot of supply on the grid or if demand is low for electricity. But in exchange, these customers do bear some risk of very high prices during certain periods—though such instances usually are rare.
Ideally, customers with variable rates would have real-time access to information about prices and would be able to adjust their usage accordingly—which, overall, can be efficient for the system if done correctly.
If all customers in Texas had variable prices, not just the small portion, then it’s possible that everyone would have seen the high prices and reduced their electricity consumption in response (to the greatest extent possible), and the price would have dropped as a result. We don’t know this for sure, but that scenario could have reduced the need for rolling blackouts, because demand would have been a lot lower. But the likelihood of all Texans in the future being subject to variable rates is very low, because variable rates are not very popular among consumers.
Has the grid—and have electricity bills—bounced back to normal?
Kathryne Cleary: Prices have come down a lot. Now that the weather has returned to what you would expect in Texas during this time of year, the demand for electricity has gone down for things like heating. They’ve also been able to restore power plants, so the supply is back up.
Like anything, it’s a market. The price is a result of supply and demand crisscrossing each other. Both supply shortages and demand spikes drive the price up, and both of those factors contributed to the high prices during the weather event. Now that those things have been restored, the prices should be back to what we would normally expect this time of year.
Texas has famously avoided joining other interstate electrical grids. To what extent did that unique arrangement contribute to the recent crisis? What happens when a state cannot as easily import power from its neighbors in times of need?
Karen Palmer: Transmission is a very important part of the whole electric delivery system, because it enhances access to distant sources of electricity.
Interestingly enough, Texas already had recognized the role of transmission investment. In the past couple decades, Texas greatly developed its wind resources—which are in the western part of the state—and built transmission capacity to bring the wind power to load centers in the central and eastern parts of the state.
But Texas is not connected with the rest of the US grid in any robust way. In the case of Winter Storm Uri, the entire state of Texas was hit by very uncharacteristic winter storm conditions and exceptionally low temperatures. As a result, both generators and fuel sources—including parts of the natural gas system that deliver fuel to gas-fired power plants—weren’t prepared the way that energy infrastructure in more northern latitudes are prepared, where cold weather is more frequent.
The extent to which greater access to fuel sources could have helped in this specific situation is uncertain, because the cold temperatures in Texas also were experienced by neighboring states, which caused increases in demand for electricity in those states, as well. However, upgrading transmission capability more generally across the country—beyond improving the interconnection between Texas and its neighbors—certainly could have helped all the regions hit hard by the cold. The need for a national transmission policy and ways to facilitate transmission investment and siting are part of a set of recommendations in a recently released study on the future of electric power in the United States, authored by a National Academies of Sciences, Engineering, and Medicine panel on which I had the privilege of serving.
For Texas, the prospect of greater interconnection with the rest of the US grid does raise the prospect of federal regulation—that’s what causes hesitancy in the state. But Texas already exports other forms of energy. With greater connection to the US grid, the state could potentially export electricity.
Unlike the rest of the country, Texas does not require utilities to maintain a certain amount of power that exceeds the expected demand. Could the state have prevented a crisis of this scale by implementing a more conventional system that ensures resource adequacy?
Karen Palmer: The short answer is that a solution would have to be about more than just capacity markets. It’s unlikely that having reserved margin requirements and an associated capacity market would have prevented this crisis from happening.
What capacity markets do is focus on peak electricity demand hours. Typically in Texas, peak demand is not in the winter, but in the summer. The markets make sure that enough capacity is on and available to supply energy during those hours. The markets do this by creating a mechanism to pay for that capacity to come online. Some electricity markets—such as PJM, which serves the mid-Atlantic and upper Midwest regions—impose penalties for any failure to deliver power, if you can’t do so when you’re called upon to do so. That does create an incentive.
However, it’s unclear to me that such a market—even including these additional penalties—would have created adequate incentives for suppliers in Texas to make the investments that would have been required to weatherize against the adverse conditions that we saw last month.
Another thing to point out is that it’s not just the electricity system, but also the natural gas system, that failed. The lines that gather gas out in the field froze. Pumps—many of which rely on electricity—failed to operate. Overall, gas production in the state fell during the peak of the crisis by something close to 50 percent. The tight links that exist in Texas (and other parts of the country) between the gas system and the electric grid mean that any vulnerabilities in one system will create vulnerabilities in the other system.
The National Academy of Sciences study includes a recommendation that the Federal Energy Regulatory Commission be given authority to designate an entity to oversee the reliability of the bulk gas system. The commission has that authority for the electric grid, but no such provision currently exists for reliability of the gas system.
Addressing these issues will involve planning, looking at what the worst-case scenarios are, assessing both the costs of experiencing the worst case and the costs of avoiding those terrible situations, and making hard choices about whether we should bear the cost of making the system more resilient. The uncertainty surrounding future extreme weather events is very high, which complicates matters. Figuring out what makes sense to do can’t be informed solely by past experience. We’ve got to think more about what situations will happen in the future, particularly as we experience the effects of a changing climate.
Electricity markets were designed at a time before the proliferation of renewable resources, when power sources such as coal and natural gas predominated. Has the introduction of renewables posed particular challenges for grid operators? To what extent did renewables exacerbate the crisis in Texas?
Kathryne Cleary: There’s a difference between reliability of the grid versus resilience. Reliability refers to the ability to meet consumer demand during what we consider to be normal times (or normal peak times) within the range of scenarios that grid operators are accustomed to planning for.
In Texas, renewables make up a large share of the grid, relative to a lot of other places in the country, at around 25 percent of the state’s capacity. We’ve seen renewables meet up to 60 percent of electricity generation in Texas at times and do so reliably, which speaks to how well the state’s market can work during normal times.
That being said, renewables can of course pose challenges for reliability, because they generate whenever they generate; they can’t be dispatched like traditional thermal generators. This intermittency will become a bigger issue as the penetration of renewable resources on the grid grows and will require investment in resources like batteries to help balance grid operations.
But the issue we're talking about with the recent events in Texas was not an issue of reliability with respect to renewables. We’re talking about an extreme weather event, which relates to the resilience of the grid: this idea of preventing disruption and bouncing back quickly if the system does get disrupted under extreme and rare circumstances that are difficult to plan for, as Texas has experienced.
During the extreme weather event last month, all types of power sources across the board experienced some failure—and most of it was thermal plants. During the storm, about 30 gigawatts of thermal plants were down, representing nearly a third of thermal capacity. Most of the downed sources were natural gas, because supply chain issues ran all the way from the wellheads to the generators themselves in the natural gas system. Lots of renewables were unable to contribute during the storm and its aftermath, though renewables accounted for only about 13 percent of the gigawatt outages that occurred; the biggest failures were wind turbines frozen with ice. As Karen noted earlier, in more northern latitudes, measures are taken to prevent wind turbines from freezing; but because Texas doesn’t typically face these kinds of low temperatures, preventive measures have not been taken.
Given that climate change is expected to increase the likelihood of severe storms, and severe weather drove the problems we saw in Texas, how can grid operators across the country prepare the grid for extreme weather events?
Kathryne Cleary: Planning for rare but very damaging events like this is definitely extremely important, for two main reasons. First, climate change will increase the frequency and severity of storms. Different parts of the country will be affected in very different ways, so it’s important to understand what those risks are and the likelihood of those risks occurring in different areas. Second, as we undergo an energy transition to fight climate change, relying more heavily on renewables can pose different risks in situations like this compared to traditional thermal power plants.
The first step toward a solution is for grid operators to prioritize resilience and long-term planning for the future. Doing so requires understanding the potential sources of disruption and their associated risks, and enacting solutions that reduce those risks cost-effectively.
Some regional transmission organizations have been concerned about our current heavy reliance on natural gas for both electricity generation and heating. They’ve talked about valuing fuel security—which is the ability to hold fuel on site for several days, and which typically is not the case for natural gas plants, but is true of coal and nuclear plants. But a solution like this, which just compensates certain technology types, could be very expensive and not necessarily effective—especially since we saw coal and nuclear plants fail during the event in Texas as well, revealing that thermal plants are just as vulnerable to extreme conditions as other plants.
The definition of what we consider to be ‘extreme’ weather could change as we become more dependent on renewables.Kathryne Cleary
As Karen has mentioned, a preferable way to deal with a problem like this might be through capacity performance rules, like those of PJM, in which generators get penalized for not being available during emergency times. PJM experienced similar, though far less consequential, issues like this, with generators unable to perform during the polar vortex of 2014. PJM then implemented capacity performance rules and saw much-improved results during the polar vortex in 2019. A solution like this, which is performance based and technology neutral, could be more effective.
Lastly, I’ll add that, as we become more dependent on intermittent renewables, planning for resilience becomes even more important. The definition of what we consider to be “extreme” weather could change as we become more dependent on renewables. An extreme future scenario could be that we experience several days without sufficient sun or wind, which could really affect the grid. Also, most power outages today occur because of impacts to the distribution and transmission system. In a renewable-dependent world, building more transmission lines to transmit renewable power from various parts of the state to load centers will be important; consequently, we’ll be even more dependent on transmission lines. And as we start seeing more distributed energy resources, we will be even more reliant on distribution infrastructure. So, understanding the risks to the entire grid—and not just focusing on generation—will be essential.
What other lessons can Texas policymakers and grid operators take away from these recent events? Could policy or market reforms reduce the risk of future blackouts? What options besides increased regulation might be available?
Karen Palmer: First, I think it’s important to point out that the deregulation that happened in Texas isn’t necessarily the root of the problem. Outages like this could occur in any state with regulated, vertically integrated markets that happen to experience temperatures they aren’t prepared for.
We can think about three main strategies as lessons to draw from what just happened in Texas. One, to take seriously what we’ll learn from our evaluations of this cold weather–related grid failure and assessments of future weather scenarios. Texas has been here before. Outages already have been associated with cold weather, both in 1989 and 2011. Both of those events resulted in some studies that recommended weatherization—and those recommendations were not adopted.
Perhaps part of the reason for inaction has been due to the infrequency of these kinds of extreme weather events. What’s different now is that we increasingly recognize that the climate is changing. We need to be more forward-looking and include potentially severe future weather events as part of the calculus, both in Texas and elsewhere, instead of just reflections based upon past experience.
Second, more distributed resources could help—including things like distributed solar, energy storage, microgrids, and the activation of flexible demand. As Kathryne mentioned, policies that enable distributed resources to make money or otherwise be worthwhile at other times would encourage their growth. For example, it could make sense to provide access to the wholesale market for distributed resources.
In the case of microgrids, we probably would need some changes in regulation if they act as a local source of power. Even though Texas does have a competitive market, the state also limits who can provide distribution service to others; so, that idea needs to be explored further to get full value from distributed resources.
We need to be more forward-looking and include potentially severe future weather events as part of the calculus, both in Texas and elsewhere, instead of just reflections based upon past experience.Karen Palmer
Third, we don’t want to give up on time-varying pricing just because of what happened in Texas. But we absolutely want to make sure that such plans come with the right tools and warnings to enable customers to deal with the potential consequences of extremely high prices in emergency situations. Being exposed to real-time prices is not for everyone. The range of possible prices needs to be communicated to people who take this option up, and perhaps those who sell such plans should be required to notify customers about the risk they’re taking on.
As Kathryne pointed out, Texas has retail choice. Most options for retail customers don’t include exposure to real-time prices. Conveying to some customers the times when electricity is scarce and when it’s abundant—particularly for some end uses like car charging or water heating—could be a low-cost way to help both integrate renewables and reduce the need for rolling blackouts in situations like this. Following these real-time prices in more typical cases of supply and demand fluctuations can effectively contribute to grid resilience—but with these extreme weather-related outages in Texas, moderating demand based on real-time prices wouldn’t have been sufficient to prevent the outages and associated hardships.