Power Struggle - The Texan Energy Grid

The energy grid in Texas is isolated, increasingly green, and under strain.

The Electric Reliability Council of Texas, or ERCOT, operates the Texas Interconnection, the electricity grid covering most of Texas, supplying electricity to over 26 million Texan consumers. In 2021, the grid experienced blackouts that killed hundreds of people. It was later revealed the grid came within four minutes of a total shutdown, which could have taken days or weeks to restart. Dysfunction continued in 2022 and 2023, with blackouts coming close nearly a dozen times this year alone. The issue has not been fixed despite billions of dollars of funding, and Texas risks yet more grid disruption this winter. 

In the aftermath of the 2021 disaster, Texans were caught in the crossfire of a media war as different interest groups rushed to blame one another. Renewable advocates were quick to point to frozen gas pipelines as the guilty party, whilst advocates for fossil fuels blamed the unavailability of wind and solar for the tragedy. The truth is more complex. Unpicking the dysfunction of the Texan grid provides vital lessons for a Western world that is rushing to switch to carbon-free generation.

Alone in the Current

Like other large nations such as China, India, Canada, and Australia, the United States has several electrical grids. In the continental US, there are two main electricity interconnections and three minor. The Eastern Interconnection serves a large portion of the population, including major cities like New York, Atlanta, and Chicago, and crosses into Canada through the midwest states. The Western Interconnection covers most of the Western seaboard and connects to the Mexican grid. These interconnections are managed by a number of regional entities. The Quebec and Alaska interconnections are two of the three minor interconnections.  The last primary grid is the Texas Interconnection, which the Electric Reliability Council of Texas manages. There are parts of Texas not managed by ERCOT - El Paso, the upper Panhandle, and some of East Texas. Still, around 90% of Texans get their electricity from a grid not substantially connected to the rest of the US grid systems. In Texas, ERCOT manages the interconnection/grid and the market serving it. 

Texas is not just independent from the other major electricity grids but also avoids being overseen by federal regulators. Indeed, Texas has a separate grid because of a desire by various Texan utility firms in the 1930s to avoid regulations created by President Roosevelt in the aftermath of the Great Depression. By only providing electricity within Texas, the companies did not cross state lines and, therefore, did not fall under the new Federal Power Commission (FPC) (now the Federal Energy Regulatory Commission). In 1970, in response to the North American Electric Reliability Corporation being formed after the 1965 Blackout, the Texan-based companies set up an interstate power pool named ERCOT. Although there was a legal battle in the mid-70s to try to prove that ERCOT was interconnected with the other grids prompted by the deliberate sending of power to Oklahoma by a company that operated grids in both Texas and Oklahoma, ERCOT managed to avoid being forcibly connected to the Southwest Power Pool.

In 1999, then-Governor George Bush signed the first renewable energy mandate for utilities, directing 1280 megawatts (MW) of generating capacity from renewable technologies by January 2003. Texas smashed through this target, leading Gov. Rick Perry to sign an expansion in 2005, roughly doubling the targets set in by the first mandate, calling for the state to obtain 5,880 MW, or about five percent of the state’s electricity, by 2015. Due to state and federal subsidies and favorable conditions for generators, who do not have to pay to connect to the grid, Texas also blew through this target. 

The current Texas energy mix is still dominated by combined cycle gas turbine units (CCGT)¹, which accounted for 37% of generation in 2023. Non-CCGT gas then produces a further 9%, totaling 46%. Gas-fired generation in the state has a maximum capacity of 69.9 GW.

Wind is the second largest contributor to the grid at 24%. Texas has the most wind turbines in the US and is only just shy of having more than the next three states put together, with more than 15,000. Total installed capacity is over 40GW, more than a quarter of all wind generation in the US, and if Texas were a country, it would have the fourth-highest capacity in the world.

The Situation Today 

A near blackout at the beginning of September of this year again highlighted some of the problems with ERCOT’s current energy mix. As twilight approached on September 7th, Texans continued to use air conditioning in the unseasonal heat as solar power became less and less available due to the sun setting. Wind generation conditions were poor, and ERCOT issued a Level 2 Energy Emergency Alert in response, meaning all available generation had to be brought online immediately. Along with ramping up what little generation was available, Texans were sent text messages asking them to reduce their electricity consumption to avoid what ERCOT and other independent system operators (ISO) call a Level 3 Emergency Alert. In a Level 3 alert, managed blackouts occur as there is insufficient power to supply the grid, but the frequency must be kept at a certain level so as not to damage generation inputs. This was the most severe incident of the past few months, but Texans have been requested to conserve energy on ten separate occasions over the summer due to the elevated demand for electricity. With increased temperatures due to climate change on the horizon, the Texan grid will likely continue to be tested in subsequent summers.

In the end, enough power could be brought online to stop the emergency from escalating further, and blackouts were avoided. Had uncontrolled blackouts occurred, they would have wreaked havoc on the grid's physical infrastructure, on industries relying on energy to keep their industrial processes stable, and, most seriously, put lives at risk. The blackouts in February 2021 resulted in the tragic loss of hundreds of lives due to extreme cold and power outages. Samsung lost over $290 million and a month's production at a semiconductor plant outside of Austin. The 2021 blackouts, as bad as they were, could have been much worse. As previously mentioned, the grid came within four minutes of a total shutdown. Restarting an electric grid from a complete shutdown is known as a black start and can be done by turning on dispatchable power such as gas, oil, or hydro. All grids are supposed to maintain backup generators to get the grid back online as soon as possible - even the biggest total blackout in US history was resolved in under 24 hours. Unfortunately, Texas does not have any significant hydrogen or oil-fired power generation, so its black start capacity is entirely gas-fired power plants.

Gas-fired power plants in Texas typically rely on supplies delivered through pipelines, not through storage on site. In the winter 2021 blackouts, not only were over 30% of ERCOTs rated winter capacity offline, but 14 of the 28 backup generators were also offline. If more of these backup generators had failed, it could have taken days or weeks to bring the grid back online rather than a number of hours.

Renewable energy advocates were quick to point the problems with gas-fired generation out in response to the Governor's claims that the lack of renewable supply had caused the blackout, with many claiming that the “pipelines froze.” This is inaccurate, as pipelines are buried underground, and natural gas cannot freeze. However, facilities that help pipelines carry gas were cut off from power, and this issue (among others) saw so much gas supply drop off, meaning gas plants couldn’t operate. It was a cascading failure. Much of Texas’s wind generation, which is not a dispatchable power source in the first place, was also offline. Regarding total generating capacity offline at the peak of the crisis, gas and wind were equal at 41% of their rated capacity being offline. The problem was not simply gas versus wind but a system failure.

On the 7th of November, 2023, billions of dollars of state funding were approved to support a more reliable grid. These costs include a $7.2 billion low-interest loan and grant program to build and maintain more dispatchable power, well as “$1.8 billion of state funds for grants or loans for the operation of stand-alone, “behind-the-meter” multiday backup power sources” to guard against grid failure at critical locations and another $1 billion “grants to transmission and distribution infrastructure and electric generation facilities located within Texas, but outside of the ERCOT power region.” This is on top of another $7 billion in funding set aside the previous year to deal with the 2021 blackout.

Furthermore, ERCOT, not generators, is responsible for building and maintaining the transmission lines from generation sites to population and industrial centers. This results in projects that would otherwise be unviable due to location continuing to be constructed. There are over 52,700 miles of transmission lines and 1,100 generation units, but so much generation capacity has been added to the grid far away from where it is needed (principally the West Texas plains) that ERCOT has to limit the production of power not to risk transmission congestion overheating high voltage lines, and in the worst case scenario, melting them. Upgrading and building new transmission lines cost $3.4 billion in 2023 alone, and there is another $11 billion worth of transmission upgrades, either planned or under construction.  

The debate after the 2021 blackouts was so vicious because Texas is seen as the future of the wider American electricity grid system, as so much of the new capacity has been renewable. In particular, its massive 40GW wind power capacity and the speed at which it was built are often highlighted to illustrate how other countries can achieve greater renewable generation.

However, wind and solar push down electricity prices when operating at their maximum capacity in favorable conditions. Given ERCOT's market structure, generators were compensated solely for the electricity they contribute to the grid. If wind and other renewables consistently offer energy at a lower cost to the market, gas plants struggle to operate at a profit enough of the time, but when wind and solar cannot contribute to the grid, there has to be backup capacity found somewhere for it to run, or the grid will go offline and Texas, the grid is isolated from other power sources in the US.

Energy projects are often pitched in cost terms using something called the levelized cost of electricity (LCOE). It serves as a metric that calculates the average net present cost of electricity generation throughout a generator's operational lifespan, but it was initially used only for dispatchable power generation, which can run for much longer (solar doesn’t work at night) and more predictable periods (the wind stops blowing sometimes). This results in costly direct financing for standby backup power ready for immediate use. These costs vary monthly, but ERCOT consumers paid billions of dollars in 2022 to keep CCGT plants online. Under recent changes, these costs are likely to increase for consumers. ERCOT also finds that substantial payments are required to disincentivize some energy usage, resulting in ERCOT paying Bitcoin miners tens of millions of dollars to not mine during the 2022 heatwave.

Energy is needed when it is needed, not when generation facilities are capable of providing it. Renewable energy advocates love to stress that wind/solar are the “cheapest forms of electricity,” but they are making this claim on LOCE grounds, which vary around the world but in Texas are estimated at $40MWh for onshore wind and $36MWh for solar, compared to $82MWh for nuclear and $38MWh for gas.

However, if both the (marginal) generation costs (the LCOE) and the (marginal) integration costs are included, the costs look rather different. Based on independent analysis from Rice University, wind could cost, on average, as much as $291MWh and solar a cool $413MWh on a Levelized Full System Costs of Electricity (LFSOE) in Texas. So much renewable capacity has not just been added because of LOCE estimates being presented to investors, but because of generous state and federal subsidies. A 2018 paper estimated that, on average, while Texas’ dispatchable generation received $0-$2/MWh, wind power received $16-$30/MWh. Although state subsidies have been reduced, federal subsidies for renewable energy will continue and have been expanded under the Inflation Reduction Act. Headline renewable energy may appear cheaper in terms of generation, but Texans bear the costs of unreliability and large amounts of capacity being offline.

What are the implications? 

Both inside and outside the United States, Texas is held up as an exemplar when it comes to transitioning to renewable energy. However, it should actually be held up as an example of how difficult it is to move a grid to renewables, even with the advantages of the fabulous wealth (in 2023, the State of Texas had a $33 billion revenue surplus) and space Texas enjoys. The problems with the grid will not be unique to Texas but are common to all grids that rely on large amounts of intermittent power.

If Texas chooses to do nothing to change how it funds energy infrastructure, it will be costly and risk grid failure through persistent underinvestment, but at least it will do so from a position of robust financial health. At least in the near term, it can arguably afford its dysfunctional grid. European countries planning further intermittent energy capacity to replace dispatchable sources do not have anywhere near such healthy finances.

Texas has an advantage in that its wind capacity is entirely shore-based, but Britain and Germany, among others, plan to host substantial amounts of their generation capacity offshore, making maintenance and replacement more difficult and expensive. Britain is currently planning on 50GW of offshore generation by 2030, while Germany is planning 115GW to meet an 80% renewables target in the same year. Given that the levelized full system electricity costs are even greater in Germany than in Texas due to how much is based offshore (wind at sea is estimated at $504MWh), this will be staggeringly expensive for any country attempting to base their electricity on offshore wind.

Even these sky-high prices are perhaps even an underestimate. Existing problems in turbines do not consider the real-life effects harsh conditions at sea will have on offshore wind turbines, but they look likely to have shorter lifespans than their onshore counterparts, requiring replacements sooner and more often. Shifting seabeds can also damage connections to the mainland, costing hundreds of millions of dollars to repair.

Even though much of the US has a far warmer and sunnier climate than Northwestern Europe, making the prospect of widespread solar energy more viable, these hopes are still dependent on massive takeup and installation of battery technology, which currently cannot meet the required tasks. Part of what makes Northwestern Europe more viable for wind energy is the weather, which also makes solar less viable there. Despite some proposals to power some European countries via connections to vast solar farms in North Africa, the energy still has to be moved thousands of miles via some of the harshest environments on earth. Experience from Texas shows this will not be cheap. In the European case, it would also increase energy dependence on countries that are not democracies. Solar is likely not the answer for wet, windy, cold countries.

The grid balancing issues that Texas faces will also be present in other grids, which are aiming to be significantly intermittent. As Samsung found out at its semiconductor factory, this causes chaos for industrial production, even withstanding the increased cost of electricity. In Britain, the industrial price of energy tripled between 2004–21, undermining the prospect of significant economic growth in a country that has not seen any real wage increases in nearly two decades. It is theoretically possible some companies might be able to justify investing in their own power generation; it is far easier, especially with Britain’s dysfunctional planning system, to close down operations. 

While European nations do hold an edge over Texas due to continent-wide interconnectors, enabling them to purchase electricity from neighboring regions (albeit at elevated costs) when necessary, an increase in the number of intermittent energy sources across Europe will lead to reduced power availability during unfavorable conditions for renewable energy generation. If everyone needs power at the same time, some will have to go without or pay nosebleed-inducing prices. Sometimes, the wind does stop blowing and can be for as long as weeks at a time. 

Given the desire of European countries to decarbonize their economies, it would perhaps be wiser to consider expanding nuclear energy despite its current very high upfront capital costs. Current nuclear reactors are over-regulated, and European designs are almost certainly over-engineered, adding considerable costs.

Conclusion 

Energy policy, planning, and forecasting are incredibly complex and have decades-long implications. Getting it right requires learning from others and adjusting accordingly. Unfortunately, politicians in countries planning to go further than Texas in the number of renewables the grid relies upon are either unaware or unconcerned about the issues affecting the Texan grid, perhaps mistakenly thinking other countries' grids will continue to bail them out when domestic intermittent sources fall short. The issues that grids face regarding increasing amounts of renewable energy generation require intense management and resources. It is unclear whether the significant financial resources required to manage these problems are worth the risks. If energy becomes so unreliable or expensive that it becomes impossible to operate advanced industrial processes, states that rely on intermittent energy will be significantly weaker and vulnerable to economic and military pressure. If a blackout took out a grid in a northwestern European country in the depths of winter, it could very easily kill hundreds or thousands of people. This has not only happened in Texas once before but is at risk again this winter. States looking to increase renewable, intermittent energy should learn from Texas before it is too late.

Thank you for reading this edition of the Kitstack. My friend Aria Babu has also launched a substack that focuses on birth rates, fertility, and pronatalism, and I wholeheartedly recommend it. A year out from the 2024 US Presidential elections, my forecasting colleague Robert de Neufville has given his thoughts on the chances of President Biden remaining in power. Subscribe here. Ed West generously gave the Kitstack a recommendation, and I’m happy to reciprocate - he is always well worth reading. Last but not least, I always find Sam Atis’s Stuff He Finds Interesting interesting and valuable. 

I’ve included some additional factual content on the energy grid in Texas below.  

Notes - The Current Texan Grid 

In 2022, Texas continued its 17-year streak as the leading contributor to U.S. wind-generated electricity. Notably, wind power surpassed Texas's nuclear generation in 2014 and exceeded coal-fired generation for the first time in 2020. The current Texas energy mix is still dominated by combined cycle gas turbine units (CCGT), which account for 37% of generation in 2023. Non-CCGT gas then produces a further 9%, totaling 46%. The newest gas-fired station is Bacliff Power Plant, with a capacity of 324MW that was brought online in 2018, but plans are underway across the state to either build or convert GWs of new capacity. (LFSCOE - $40MWh)

The second largest contributor to the grid is wind at 24%. Total installed capacity is over 40GW, and ERCOT is forecasting another 37,268MW of generation to be added by 2027. The largest wind farm in Texas (and second largest in the US), Los Vientos, is a four-stage 912MW facility close to the Mexican border, owned by Duke Energy. There are currently no offshore wind farms connected to the Texan grid, and although the Federal government has proposed developing offshore facilities in the Gulf of Mexico, no private firms have bid to build them and connect them to the grid in Texas, instead preferring a potential development in Louisiana. (LFSCOE - $291MWh)  

Coal, although significantly reduced from being the second largest contributor until 2018, when it contributed 25% of generation, is the third largest contributor at 14%. 13 coal-fired power plants contribute to the grid. Martin Lake Power Plant has a capacity of 2,250MW and is the largest coal-fired power plant in the state. Since 2012, seven coal plants have been fully decommissioned; of the 13 still running, four are supposed to close by 2030. However, of these four, some may be converted into gas-fired plants, given the strong financial incentives unveiled this year by the Texas Legislature. (LFSCOE - $90MWh)

Nuclear power generates 9% of Texas’s energy. The South Texas Project Electric Generating Station and Comanche Peak Nuclear Power Plant were completed in 1989 and 1993, respectively. Although expensive in upfront capital costs, nuclear power is clean and reliable. This is partly why Dow, the chemicals giant, is exploring an experimental small modular reactor to power operations at its Seadrift manufacturing site. Aside from bespoke projects such as Dows, there are no current plans to construct any new nuclear power stations in the Lone Star state, and the last proposal to expand Comanche Peak, in a partnership with Mitsubishi Heavy Industries, was put on hold in 2013. (LFSCOE - $122MWh)

Solar has seen the fastest growth in recent years, currently contributing 7% to the grid, up from 2% in 2020. A favorable climate, large amounts of available land, and generous subsidies have been responsible for increased solar generation capacity. The largest plant in Texas, the Roadrunner solar power plant, is the third largest of its kind in the US, with 497MW of generating capacity and 52MW of battery storage. Enel Green Power, the owner, did not disclose how much it cost to build the plant. (LFSCOE - $413MWh). ERCOT is expecting to see 15,659 MW of solar capacity added to the grid by 2027.

Despite Texas being a significant oil producer, producing 5.49 million barrels per day (MBD) in May of this year (for context, Saudia Arabia produced around 10 MDB before imposing cuts to its production in July), Texas does not incorporate oil into its energy mix. Oil-fired power plants were heavily discouraged in the US by the 1978 National Energy Act passed after the Arab oil embargo.

1. Integration costs can be subdivided into balancing, grid, and profile costs. Balancing costs encompass expenses borne by the operator to address the uncertainty of intermittent generation, ensuring a continuous balance between supply and demand. Grid costs pertain to adjustments in the grid necessary to support the renewable system. Profile costs encompass all expenses associated with aligning supply with demand under the assumption of perfect market condition forecasts, distinct from balancing costs.