Like much of the rest of the country, the ERCOT region of Texas is projecting massive growth in electricity demand, driven mostly (though not exclusively*) by new data servers. The servers are part of the tech sector race to provide (power-hungry) A.I. services to the market. The Texas legislature is considering a bill, SB6, that aims to address the mismatch between current grid-based generating capacity and those projections of rapid electricity load growth.
Unfortunately, to the extent that interested audiences both inside and outside of Texas hear about this bill, they will hear about it on social media. That is too bad, because (for reasons explained in chapter 4 of my book) those accounts will be packaged to fit each reader’s partisan and ideological biases. And of course when the legislature makes its decision, none of us will get exactly the policy we want; but we will be encouraged to chalk up that deviation from our ideal policy to the moral or intellectual shortcomings of the other side.
But I watched portions of the legislative hearings on SB6 last Thursday, and it was instantly apparent that the whole truth is much more interesting, nuanced and encouraging than most voters will ever know. I will explain what I mean, but first a few paragraphs of necessary background information.
Background and Basics
First, as with all things Texas energy politics, the best curator of these developments is Doug Lewin. Other energy news services do a good job as well, asked does the Texas Tribune. But if you are interested in really understanding Texas electricity policy news on a deep level, follow Doug’s podcast and newsletter (after you finish reading this post) :)).
Second, the load growth projections facing the Texas grid are indeed startling — less because of their size than their speed. Within the last year, both summer and winter demand peaks topped out at less than 90 gigawatts (GW). Current projections have that number rising to 150 GW by 2030. Those projections also show demand exceeding the grid’s supply capacity as early as next year.
[Ironic side note: I asked Google for data on the total generating capacity on the ERCOT grid. The “AI overview” it provided as the first result was hilariously wrong; click the screenshot of those results (at right or below) to enlarge it.]
This sudden electricity demand growth is what’s driving the legislature’s concern, and SB6 has emerged as the leading bill in response. So the key questions facing Texas, and policy makers in the rest of the country, are these:
- Should these AI servers (and other very large new loads) be required to “bring their own generation” with them before interconnecting? Or should they have the same rights to connect to the grid that you and I do?
- When new large loads make deals to co-locate with power plants, are they “cream skimming“ power that would otherwise be available to serve the other customers on the grid?
- What sort of new generation can best satisfy increases in electricity demand? More renewable energy of the kind that has been growing rapidly on the Texas grid? Gas fired power plants and other so-called “dispatchable“ generation? Batteries? Demand response and other distributed energy resources?
- Can we solve some of the supply problem by building new transmission lines so that underused power plants can sell their power in parts of the grid where generation is (or will soon be) scarce?
The Fun House Mirror
These are not easy questions to answer because there is no optimal solution for everyone concerned. They involve choosing between different policy values. And since today’s Balkanized information environment delivers to each of us “news” that is distorted in different ways by different fun house mirrors, we will each see a version framed to fit our respective ideological or partisan predispositions. Worse, we will be encouraged to ascribe policy choices we don’t like to ignorance, corruption, or some other nefarious factors.
Arch conservatives will process these developments in ways that firm up their suspicions that climate alarmism prevents green energy advocates from appreciating the need for reliable, affordable electricity (when the sun isn’t shining and the wind isn’t blowing). Arch progressives will process this news in ways that firm up their beliefs that Texas Republicans are either lackeys for the fossil fuel industry or MAGA performance artists more interested in fomenting outrage than good governance.
And importantly, many of these distorted framings will be motivated by the sincere desire to tell readers and listeners what the writer/speaker believes that they most need to know. Many writers and podcasters have firm opinions about how these questions ought to be answered. They proselytize because they think that their beliefs are the true beliefs, and that spreading them helps everyone.
For example, in one of my social media communities I can expect that most community members will challenge the accuracy of ERCOT’s load growth projections and the need for more grid-based generation at all; some will likely suggest that Texas policymakers who disagree with them are the willing dupes of industry because they can’t imagine any other plausible explanation for believing in those load growth projections.
On another online platform I expect that my information feed will include plenty of people who push hard for more-renewables-plus-more-batteries, a combination that has performed well during recent grid stress events. They will have trouble viewing legislators who seek to add more “dispatchable” generation (read: less weather-dependent than wind and solar) as reasonable people.
And in the news ecosystems of ideological conservatives, any failure to add more dispatchable power will be portrayed as naive disregard for the overarching imperative to keep the lights on. And so on.
The Undistorted Picture
But almost no one will see what I saw last Thursday morning watching the SB6 hearings: namely, (a) legislators from both parties asking grid managers and experts probing, intelligent questions aimed at solving a real world problem, and (b) questions framed respectfully and cautiously, and (c) grid managers trying to be careful, circumspect and honest in their answers. (Don’t believe me? Look up the recording of the hearing.)
For example, Republican Senator (and emergency room physician) Donna Campbell asked a series of open-ended questions that seemed aimed at finding the best solution. She asked about whether the 150GW projection was a reliable one. She asked if there are incentives in place for new large loads to bring their own generation to serve some or all of their needs. She seemed to trying to determine if there is a set of policies that will work for all consumers, and her exchanges with the testifying experts were substantive — the kind of questions one expects from someone who is sincerely trying to understand a complex problem.
A caveat: I did not see the entire hearing. Maybe there was some grandstanding. Maybe there were questioners whose minds were already made up, and who used the hearing to lobby their fellow legislators. But if so, those moments were not illustrative of the larger event. The portions that I watched reflected people of both parties engaging experts in search of solutions to a difficult problem.
Perhaps that part of the reason for all this civility and problem-solving is that the problem itself doesn’t neatly map onto the partisan divide. The examples of the cross-cutting politics of this problem are many:
- Ideological conservatives who favor free markets want to reduce regulatory barriers to entry for investors in new generation. If so, that will mean a massive influx of new solar power projects that are already in the interconnection queue, awaiting approval. Solar power is cheap, and there is pent up demand for capital investment in that technology.
- For their part, west Texas oil producers and bitcoin miners want to use cheap renewable electricity in their operations. In many places those customers face a glut of nighttime wind generation, which keeps spot power prices low over night, but also a shortage of power (and higher prices) during the day. Adding more solar generation in those regions would depress daytime prices as well, to their benefit.
- The tech companies that will own these new server farms have long professed a preference for using clean power. But above all they want reliable power, and they have the cash on hand to pay above market prices for it. If clean power advocates insist on policies that don’t satisfy servers’ need for 24/7 power, they will turn to dirtier power supplies. If Texas policymakers try to deny those companies access to a cleaner, reliable supply, they may locate their servers elsewhere.
- And supporters of all forms of existing generation can find reasons to support new investment in transmission. There are isolated pockets in the Texas power market: places where there is a chronic supply-demand imbalance and insufficient transmission capacity to import or export power. Better transmission connections would enable owners of all kinds of power plants — wind, solar, natural gas, batteries — to make more money.
The cross-cutting ideological and partisan nature of this issue may be incentivizing less nakedly partisan and performative behavior by Texas legislators. Or maybe it’s that state legislators receive much less press coverage than members of Congress do.
Certainly the most ambitious, higher-profile Texas politicians do attract more press attention, and they do all they can to show contempt for those on the ideological left: attacking cultural progressives, attacking the way Texas universities are governed, demonizing undocumented immigrants and those who try to help them, etc. And because our senior senator (John Cornyn) faces a potential primary challenge in 2026 from a particularly virulent MAGA populist, he will feel the need to join the contempt chorus in order to keep his job — for all the reasons explained in Climate of Contempt.
But away from the limelight it may be that many Texas politicians of both parties are willing to actually try to solve a complex problem. So let’s take that little win. And when those politicians make choices that differ from the ones we would make, let’s not to take that as a sign of their ignorance or corruption, but rather of their differing notion of what the public interest requires. — David Spence
—————
*Population growth, economic growth, cryptocurrency mining, and the electrification of oil and gas wellhead operations also contribute significantly to these so-called “load growth” projections.
