As Summer Heat Hits, How Is the Texas Grid Faring?

Amid new projections of surging demand, energy analyst Doug Lewin explains what worries him–and what doesn’t.


In the over three years since Winter Storm Uri, there’s been far more attention paid to and media coverage of Texas’ oft-precarious electric grid. Highly contentious debates have raged around how to regulate power generators, address soaring demand, and, most basically, keep the lights, ACs, and furnaces on.

Perhaps no other Texan (save maybe a handful of state bureaucrats) has followed the slow-moving and seriously wonky minutiae of these matters more closely than Doug Lewin, an Austin energy consultant and policy analyst. In the wake of Uri, he developed a large online following by tirelessly live-tweeting marathon legislative hearings about how to fix the grid. 

Lewin’s carried that on in the months and years that followed as officials at the state’s Public Utility Commission (PUC) and grid operator Electric Reliability Council of Texas (ERCOT) have gone about implementing an overhaul of the power market. Nowadays, Lewin’s smart analysis can be found at his Texas Energy and Power Newsletter. The Texas Observer spoke with him about what he’s watching for as the summer heat hits, huge projected increases in grid demand, and how lawmakers and regulators are handling it all. 

TO: At a state Senate committee hearing earlier this month, ERCOT CEO Pablo Vegas said that the grid operator is projecting Texas will need to nearly double the current capacity of 85 gigawatts to 150 gigawatts by 2030. This huge uptick in projected demand, he said, is driven in large part by rapid expansion of new Bitcoin mining facilities and AI data centers that suck up tons of electricity. Was this huge uptick in projected demand surprising or should it have been expected? 

This Senate hearing was not the first time the 152 gigawatt (GW) number came out. It actually came out of the regional planning group [an obscure ERCOT committee], where they deal with long-term planning issues, at a meeting in early April. I wrote about it in my newsletter. But then, I guess the Senate hadn’t really been paying attention to it. And Dan Patrick tweeted [on June 12] saying it was a shocking bombshell or whatever. 

The other thing that is important here is we don’t know whether or not this demand is actually going to materialize. I think it is absolutely true that there will be rising demand, but I don’t think you have to predict the future for that. We’ve seen that in the last two years. We set a peak demand record in 2019 of 75 GW. We did not exceed that in 2020 or 2021, and then in 2022 we hit 80 GW and in 2023 we hit 85 GW in the summertime. That is a very large jump. Obviously 150 GW in the next six years would be an even bigger jump.

But I think a lot of this [projected] load is speculative. I think the 152 is probably higher than what we’re actually going to see. I do think a steady rise of 5 gigs a year and maybe even higher than that is possible. But I think we need to take some of those projections with a little bit of a grain of salt.  

What’d you make of Patrick’s response to ERCOT’s projections? Is it concerning that the Lieutenant Governor, who has been very outspoken and influential on electric power policy since Uri, was apparently surprised by this massive projected increase in demand? 

I always try to keep in mind that all these legislators are dealing with all the issues. I specialize in and focus on this one issue area. I’m never surprised when they don’t know something. I would say what was a bit surprising to me was for Patrick to lump in AI data centers with Bitcoin in his tweet when he said something like “They don’t add a lot of value. They don’t create a lot of jobs. Maybe we don’t want them.” That’s a pretty huge change in tone for that high of a level of state leadership to basically be turning away economic activity. 

I don’t really have a strong opinion on Bitcoin one way or the other, as it seems everybody else in the world does. I’m much more interested in the other large flexible loads that can come on to the grid. EV fleets. Direct air capture machines for carbon capture. Green hydrogen from electrolyzers. 

And something like AI data centers, it seems to me, would be something you would want in your state or your region or country. It just seems that it goes very much against the kind of Texas ethos that has been around for a while, for better or worse—“open for business,” right? It’s like, well, open for business, but not for AI data centers. That really struck me as a pretty major change in tone. That was a little bit surprising to me; the fact that he didn’t know that these loads were coming was less surprising.

Patrick also called for the Legislature to look closely next session at the impact and burden of all these Bitcoin miners and data centers coming online in Texas. Many Texas Republicans have been major promoters of turning the state into an oasis for the Bitcoin industry and claimed they would bring major benefits to grid stability. Is this a sign state leaders may be singing a new tune now?  

Last session, the Senate unanimously passed a bill that would have reined in [grid incentives for] Bitcoin. [The bill died in the House.] So I don’t think it’s totally new for Dan Patrick. Ted Cruz and Greg Abbott and some other high-level elected officials have been much more solicitous of Bitcoin miners coming to the state. 

I think it’s going to become more and more important for the state for a variety of different large energy users to have some conditions on them being able to connect [to the grid]. I think the 150 GW [projection] is probably too high. But let’s say growth was half that amount in the next six years. We’re talking about going to 120 GW—over 85 last summer. That is outrageously fast. You can’t build out a grid for a peak of 120 gigawatts in the next five years. That is not going to happen. It just bumps up against physical limitations.

Now if you look at the stuff that is already in the PUC’s interconnection queue [for potential new ERCOT power generation projects], there is about 330 GW. A lot of that, most of that is solar and storage. So you bring lots of solar and storage into the system. And then there are various large loads that are coming that have some ability to reduce their use during the [most critical hours for the grid].

An ad with the text: When Texas is at its worst, the Texas Observer must be at its best. We need your support to do it. A button reads: JOIN NOW

And when I say, “We don’t have enough,” or, “We can’t build out a system fast enough,” what I’m talking about to be clear is a few hundred hours. There’s 8,760 hours in a year. Well, what we are really talking about is a hundred, 200, maybe in a really extreme year 400 or 600 hours [that the grid may fall short]. 

We’re heading into the heart of summer. So far, it hasn’t been as hot as the past two summers where we had record-breaking heat across the state and for unbearably long periods of time. There were numerous close calls with the power supply kind of teetering on the edge of meeting demand. How has the grid been doing so far and are you particularly concerned about getting through this summer? 

I think the main concern is around the limitations that ERCOT is placing on transmission. It’s what’s called a generic transmission constraint, which basically just means that ERCOT’s placing a constraint on it because they’re worried about that transmission line getting overloaded. 

This was the situation on September 6 last year where we had very robust wind in South Texas and other resources in South Texas that were curtailed. They were not allowed to put their power onto the power line because ERCOT was worried about the power line becoming overloaded. And they kind of precipitated a near-crisis. That was the night they went to an Energy Emergency Alert Level 2. They skipped Level 1. They went right past that and just went straight to 2, because they could actually get some demand-response resources at that level that they desperately needed. And they were able to stabilize that situation and avoid rolling blackouts, but they were within a few hundred megawatts away from rolling outages on September 6. 

It wasn’t a situation where we didn’t have enough capacity. We did. We couldn’t move it where it needed to go, or ERCOT decided not to move it where it needed to go. So that is a risk. I don’t think it’s a risk that’s being talked about enough. 

To me, the far bigger problem is still winter. And, while we’re in summer, and that’s what everyone’s thinking about, it’d be hard to overstate how much better our summer situation is because we have nearly 25 GW of solar. And on the days we’re not getting much solar, well, guess what? When the clouds come over the sun, it’s not as hot. So the solar power is so well correlated with these very hot, very high-demand summer days.

In terms of all of the new laws, policies, rulemakings that state leaders have charged the PUC and ERCOT with implementing in response to Uri, how would you grade the power regulator and grid operator? 

I think it’s really important to point out that the Railroad Commission is involved in this, too, and they’re pretty opaque. So it’s hard to tell what they’re doing. But I don’t have a lot of confidence that the natural gas supply system is any better off than it was. It’s probably a little better off, just in the sense that at least natural gas infrastructure is now designated as critical and so won’t be shut off in rolling outages. Some things around the edges like that could have a big impact. So I think we’re better off than during Uri. But as far as actually weatherizing the natural gas supply infrastructure, I think precious little of that has gone on. 

There, I think the PUC and ERCOT have done a pretty decent job, and you have evidence to support that during Winter Storm Heather [in January] compared to Winter Storm Elliot in December 2022. Those were similar events as far as the temperature. And we had pretty poor performance of thermal power plants during Elliott and much better during Heather.

I’d say where they’re not doing as well: There has been no increase to energy efficiency whatsoever. And that is even though [federal regulators], just like in 2011, gave us recommendations and many of them were ignored. And we ended up with Winter Storm Uri. 

They gave us recommendations after Uri. One of those was to implement energy efficiency programs. That has not been done. There have been a lot of good meetings, a lot of good discussion. But talk is not action. And it’s been three years and it’s really past time. 

One other thing: At the Senate committee hearing, it was very interesting hearing PUC Chairman Thomas Gleason saying, “After Uri, the focus was really on reliability. We’re trying to do a better job of balancing that now with affordability.” And these last couple of years have been extremely expensive, and I think that is the whole conservative operating posture [adopted after Uri], and it’s still there. And I think that this current regime with Chair Gleason and Pablo Vegas, they just really need to get a handle on that.

It’s interesting because the only way they’re ever going to be able to drive down costs is to do more renewables and storage and energy efficiency and demand-response. But state political leadership, certainly some state political leaders, seem to be very against renewables. And renewables are one of the only things putting a downward pressure on prices. So it makes things endlessly interesting.

Editor’s Note: This interview has been edited for length and clarity.