A leading energy expert examines the causes of the Texas power crisis of February 2021. He finds that many of the contributing factors were well known beforehand and even caused a smaller power outage in 2011. He then lists some steps that could help reduce the potential for another, similar blackout.
Extreme weather is a growing threat to power grids everywhere. Weather—and the knock-on effects of long-term weather patterns such as drought—has caused blackouts in the northeastern United States from Hurricane Sandy in 2012, in the Midwest from a so-called polar vortex in 2014, and in California due to wildfire risks in 2020. It is expected that as climate change progresses, the weather at any given place on any given day will be farther from the gentle conditions thought of as average.
As a state with significant portions in semi-arid climate, we here in Texas prepare the grid for the hot summer afternoons when air-conditioners run full-blast. And there are plenty of them. But the state experiences Arctic fronts more often than Texans care to admit. These cold snaps are produced when the jet stream, a high-altitude wind which generally blows from west to east, develops large sinusoidal wobbles that pull warm, semitropical air north and plunges cold, Arctic air south.
In a recent February, cold fronts marched slowly southward, bringing cold temperatures into Texas. As the freeze swept across the state, natural gas demand spiked, since nearly 40 percent of Texas homes are heated with gas, and while electricity accounts for the other 60 percent or so of residential heating, gas fuels about half of the state’s electricity generation. At the same time, the cold temperatures caused freezeoffs at wellheads, processing facilities, compressor stations, and other equipment for the natural gas system. These freezeoffs occur when water that is produced alongside oil and gas freezes, clogging or damaging wells, pipes, and other equipment. As a result, supplies for natural gas started to drop at the same time that gas demand to heat buildings and drive turbines at power plants was spiking.
It wasn’t just natural gas production that was hit. Cold weather also caused equipment to fail at power plants, tripping offline several gigawatts of capacity overseen by the Electric Reliability Council of Texas (ERCOT), the state’s grid manager. Blackouts started, turning off power to parts of the gas production and distribution system, worsening the supply crunch and accelerating a series of cascading failures. In total hundreds of power plants that tried to fill the gap experienced an outage, were derated, or could not start, including a large fraction of gas-fired generators that could not get the gas they needed. Millions ultimately were without power and a non-trivial fraction had boil-water notifications because burst pipes and water mains caused leaks that raised concerns about contamination in the water supply.
Those events I just described occurred in February 2011.
The authoritative after-action reports by the U.S. Federal Energy Regulatory Commission (FERC) and industry stakeholders recommended that because gas-power-water interdependencies could trigger a downward spiral of interlinked failures in future cold-weather events, the energy system in Texas should take steps to winterize. For the most part, that advice was ignored: Too few planners in water, gas, or power invested to improve the robustness of their facilities against severe winter storms. The experience of the 2011 cold snap did not become a learning experience.
Flash forward almost 10 years to the day to February 2021. Winter Storm Uri moved slowly southward, but this time it was colder for longer and spanned the entire state such that all 254 counties in Texas were simultaneously frigid. The gas system froze up starting February 10, dropping Texas production in half at the depth of the crisis. Several GW of natural gas power generating capacity were not available because gas companies cut off their supply of fuel. Other power plants struggled to remain operational at full capacity. Ultimately, the power system started to “shed load”—that is, cut off power to customers whether they wanted to be disrupted or not—at 1:20 am on February 15, 2021.
At the depth of the crisis, about half of ERCOT’s generating capacity was unavailable. Because about 85 percent of Texans are served by a grid that is almost completely isolated from the other two grids in the continental United States—the eastern and western interconnections - we could not lean on other states for reliability support. Some of the bordering states, such as Oklahoma and Louisiana, were struggling with their own power systems, as well, so they could have helped a little bit, but not a lot.
Ultimately, more than 10 million Texans lost power for a few days and 14 million had a boil-water notification for several days. Hundreds of people died from the storm and the total damage is estimated to be between $100 billion and $200 billion (including crop loss, repairs to damaged property from burst pipes or falling ice, medical costs for thousands, and the loss of life for hundreds of people), making it as expensive as Hurricane Harvey and other major disasters.
Cautionary tales are useful, but only to the extent they are heeded, and looking for scapegoats rather than root causes leads to making the wrong decisions and taking the wrong steps. In the case of Texas, if the state carries out a program that papers over problems instead of addressing them, the power system will likely fail again when the next severe winter storm hits. The choices we make now will dictate how vulnerable we are in the future.
Some colleagues and I in June published a preliminary report on the 2021 Texas power crisis, and this is what we found. While the cold snap of 2021 was longer and deeper than the one in 2011, addressing the interdependencies between gas, power, and water might have averted the worst of the crisis. The natural gas system uses a lot of electricity to extract gas from the ground, separate it from liquids like oil and water, inject anti-freeze, transport it along pipelines, and pump it from storage caverns. Electrification helps keep production clean and cheap, but it also invites problems. Some gas producers sign interruptible power contracts in exchange for lower electric bills or large cash payments when the grid operator asks them to turn off production to save power. This configuration creates a potential moral hazard: Right when electric power and residential heating need gas more than ever, these gas producers volunteer to shut down—and we pay them to do it.
At the same time, modern gas power plants are cleaner and more efficient, but operate at much higher pressures. If the pipeline pressure sags and the power plant does not have on-site compression, the plant can trip off-line because of fuel supply issues. A lot of gas power plants had their gas supplies cut off, including some that had firm supply contracts.
And while the shale revolution has led to vast amounts of cheap gas to be available to Texas utilities—the state is the world’s third-largest natural gas producer, behind only Russia and the non-Texas part of the United States—the gas produced from the Permian Basin in west Texas is particularly liquids-rich, meaning a lot of water is produced along with the oil and gas. Excess liquids solidify at low temperatures, making the region especially vulnerable to freezeoffs.
Low temperatures lead to frozen liquids which leads to less gas which leads to less electricity which leads to less gas. And so on.
Weather conditions prevented other power sources from filling the gap. Winter Storm Uri was not just cold, but also wet and initially windless. In 2011, about 8 GW of installed wind capacity generated about 4 GW of power during the storm. In 2021, wind capacity had roughly tripled to more than 24 GW, but still only generated about 4 GW of power. Ice accreted on the wind turbine blades, posing a risk to equipment and nearby facilities as the ice can slide off the blade like a long, deadly ice javelin. Even when the wind farms worked just fine, some substations or transmission lines connecting them to their customers froze, blocking the wind farms’ power from traveling to market.
Snowfall initially reduced output of solar panels. For instance, the panels on the roof of my house in Austin were covered in snow and saw their production drop to almost zero for about a day. Utility-scale solar farms with tracking systems could turn panels away from the snow; once the snowstorm passed, the panels turned to face the sun and production recovered quickly. Low temperatures even improved the panel efficiency and the additional light reflected from snow-covered ground increased output.
Thermal plants—those that use heat from coal, gas, or nuclear fuels—couldn’t keep up. On top of the woes at gas-powered plants, coal-fired power plants endured broken equipment or frozen coal piles and one of the state’s four nuclear reactors tripped off-line when a water pump froze. And reservoirs froze, limiting the availability of cooling water.
Just about every fuel and technology struggled one way or another, and because of how closely these systems are coupled, failure rippled through every sector.
To those of us who endured the 2011 crisis, the response to the 2021 Texas power crisis was grimly familiar. There appears to be a three-phase process: Phase 1 is the finger-pointing and misdirection, lasting weeks. During and immediately after the 2021 storm, Texas politicians went on TV to scapegoat wind turbines, out-of-state politicians, or yet-to-be-enacted policies such as the Green New Deal for the failures.
Eventually, Phase 1 settled down and we entered Phase 2, which includes investigations and forensic analysis to get the facts straight about what went wrong and what the sequence of failures was. For example, FERC’s official assessment of the February 2011 storm, along with recommendations of how to avoid something similar in the future, was released six months after the storm in August 2011.
Phase 3 is the implementation of solutions, which takes a few years. Or in the case of 2011, not long at all, as too few entities followed FERC’s guidance and invested to winterize their systems to reduce vulnerabilities. It is too early to tell whether Texan energy stakeholders will take Phase 3 seriously this time.
Solutions at Multiple Scales
This crisis happened in Texas, but it isn’t a particularly “Texan” crisis. The entire country still relies on a far-flung system of large, remote power plants that need long-haul transmission systems to bring gas to the turbine or power to market. All too often if one pipe or transmission line goes down, gas can’t flow sufficiently to the power plant or the electricity can’t flow from the power plant to the customer.
To better handle this, energy planners across the country recommend modernizing the grid at multiple scales. At the largest scale, the country should build a more comprehensive network of underground high voltage direct current power lines that are robust against bad weather, connect wind from the plains and solar from the deserts to urban areas, and make the entire grid more reliable. Regionally, we should consider interconnecting ERCOT to the Eastern and Western electrical grids so Texans can sell its clean, renewable power for a profit most of the time and receive (or provide) reliability benefits those rare periods when our grid or our neighbor’s grid struggles.
Closer to electricity demand, utilities should implement more energy storage and microgrids to build flexibility and resilience. Also, utilities should move away from the “fail in place” model where parts are replaced only after they break, since multiple pieces of equipment can fail almost simultaneously during a region-wide weather event. Implementing efficiency and conservation into local building codes would reduce energy demand.
And as we saw in Texas in February, interdependencies between the gas and electricity systems can produce a vicious downward spiral that can cripple a region under certain conditions. Those power outages also can disrupt the water sector. A more resilient system design would avoid these cascading failures.
All these should be a national priority because local power crises can create continent-spanning ripples in an interconnected economy. A power outage in Texas might mean Atlanta doesn’t get jet fuel. Iced over roads in Ohio could leave the upper Midwest without propane. The failures are waiting to happen, and they are everywhere.
Above all, we need to stop designing the infrastructure of tomorrow based on the assumptions of yesterday. Climate science teaches us that weather events are likely to become more intense and more frequent. We should plan accordingly. Sure, northern climates need to prepare for hotter summers by planning for larger loads from air conditioners and helping to keep ancillary equipment cool. But because weather extremes in both directions are becoming more common, hot climates need to winterize the gas supply and power plants with insulation, heat tracing, temporary or permanent enclosures to keep out the wind and precipitation, on-site heaters for thawing equipment, cold climate packages for wind turbines, and snow removal for solar panels. These are steps that were called for 10 years ago and they are only more critically important today.
What happened in Texas can happen anywhere. Lawmakers need to use what happened here as a motivator for action so that we don’t have millions more stranded and shivering in the dark. In fact, the riskiest move would be to settle on scapegoats rather than address these issues. Choices made—and actions taken—now will dictate how vulnerable we are in the future.