ERCOT’s current rulemaking process will shape the Texas grid for decades, driving infrastructure investments that last 30 to 50 years and cost billions of dollars.
During this year’s South by Southwest festival in Austin, ERCOT Chief Executive Pablo Vegas sat down with Energy Capital Podcast hosts Josh Rhodes and Matt Boms to explain the grid operator’s approach.
ERCOT has added more than 60 gigawatts of new supply since the devastating Winter Storm Uri blackouts in 2021, and battery storage resources have grown from a few hundred megawatts to more than 16 gigawatts. Vegas said the grid is more reliable today than it was three years ago. The challenge now is to plan for what’s coming. Last month, ERCOT announced that load interconnection requests now exceed 410 gigawatts. For comparison, existing load has peaked at around 85 gigawatts in recent years. New data centers drive much of that growth.
In this episode, Vegas described how ERCOT determines which projects in the interconnection pipeline are likely to be built. Even a fraction of those projects could reshape the system, especially if data centers arrive in the concentrations that some projections suggest.
Vegas also walked through ERCOT’s proposed batch study process for reviewing large load interconnection requests, and why the current one-review-at-a-time approach is inadequate given growing load projectionsAnd he discussed residential demand response — and why it may be a faster path to reliability than building new generation or transmission.
The ERCOT grid is growing like never before — yet demand is growing even faster. ERCOT’s response to this challenge will shape our grid and our economy for generations. Check out this week’s episode to learn more about what that response will look like.
Timestamps
00:00 - Introduction and Pablo Vegas
03:40 - Lessons from Winter Storm Uri
05:42 - How 60 GW of New Supply Changed the Grid
09:18 - Filtering the 230 GW Load Forecast
13:28 - Why Data Center Load Broke the Old Process
19:15 - How ERCOT’s Batch Study Process Works
22:02 - DERs and the Distribution Grid
26:57 - Real-Time Co-optimization and RTC+B
30:11 - Battery Duration vs. Flexibility
33:26 - Residential Demand Response
36:41 - How ERCOT Is Using AI
40:35 - What Texas Should Learn and Export
Resources
People & Organizations
Company & Industry News
ERCOT Goes Live with Real-Time Co-optimization Plus Batteries
Texas Task Force Aims to Tear Down Barriers to Virtual Power Plant Pilot
Programs & Processes Discussed
Related Podcasts by Energy Capital
Transcript
Joshua Rhodes (01:37.714)
the most in the first few years leading our cup?
Pablo Vegas (01:40.878)
I guess what’s been really interesting learning is, and when you work in the utilities space, and I worked in the utility space in Arcox, back in Albuquerque, Texas, but I’ve worked in the utility space, as you said, for quite a while, for over about 20 years.
Pablo Vegas (01:56.908)
I was surprised how different grid operations is from what it looks like inside of the utility. So the issues that the grid operator deals with in contrast to what a utility company deals with is pretty stark. And while there’s a lot of kind of overlap on elements of it, of course, but kind of the focus of what we’re looking at, which is, you know, looking at all of the issues across all the different components and players in the system, it’s a lot more complex trying to consider all of the
Pablo Vegas (02:26.67)
needs of the different stakeholders that are part of the process. When you’re a utility company, you’re always laser focused on your customers, you know, directly, and you’re serving them and you’re making sure that you are working constructively with, you know, your regulators and policymakers to serve your customers. At the grid operator level, the customers are still a very critical part of the conversation, but you’re doing that through one layer removed in helping to oversee the processes.
Pablo Vegas (02:52.354)
that govern the utilities and the power producers and the retail electric providers and the brokers and traders, the large industrials and the small consumer, everything in between. And so you’re really thinking about kind of the policy making that affects each one of those segments in a different way. And that’s what surprised me a little bit in terms of just how complex that can be, how in many cases, how political it can be. We have a very unique environment in Texas where, you know, we govern, legislate all the policy for what happens in the wholesale power markets.
Pablo Vegas (03:21.624)
which is different than what happens in other states. And that’s something that adds to the complexity inside of Texas, but it also gives us benefits, gives us the agility, the ability to make change. And we’re seeing that realized today at the pace of change that’s happening.
Matt Boms (03:34.9)
And I think you came in at such an interesting time
Matt Boms (03:37.373)
after Winter Storm, Yuri. We just had the five year anniversary of Yuri recently. Do you still feel like that’s hanging over our heads or do you feel like we’re now entering this new chapter? We’re going to talk about low growth and data centers and all that. But you came into the job during that kind tumultuous period. Do you feel like that’s behind us now or do you still feel kind of the shadow of Winter Storm, Yuri hanging over us?
Pablo Vegas (03:58.754)
Bye.
Pablo Vegas (03:59.48)
firmly believe that the memory of Yuri is going to remain with those who experienced it for years and years to come. I don’t think it’s something that is gonna be forgotten and put into the past and nor should it because it has taught us lessons across so many different facets of the work that we do, the importance of great communications. That’s something that was, I think, a very early lesson learned in the process that there needs to be so much better transparency
Pablo Vegas (04:29.424)
in terms of the role that our cop plays, what we do and how we do it, why we do it, who we work with in delivering the general service that everybody experiences. It was clear that it was not well understood as we went through the whole URI experience. we’re still working on that facet of it. But beyond that, the lessons learned in terms of just how critically important and dependent.
Pablo Vegas (04:52.738)
we all are on electric reliability is something that we can’t ever forget or should put in our past. And that’s something that should remain at the forefront as we think about policies, as we think about protocol changes, as we think about how we’re going to manage the growth that’s ahead of us, how we’re going to manage the transition to the different supply resources and everything in between. So it’s something that I want to continuously be helping people understand the things that we’re doing.
Pablo Vegas (05:20.428)
to change the circumstances that allowed Yuri to happen and we have made a lot of progress on those fronts. But it’s never gonna be gone from the memories of those who experienced it and we’re always gonna keep the lessons learned in our forefront because those lessons have to.
Joshua Rhodes (05:35.214)
Yeah, so since Yuri there’s been quite an addition of capacity. It’s been a bit of a different type of resource mix that we have. think during winter storm Yuri there was just a couple hundred megawatts of solar. Now there’s tens of gigawatts of solar and like an energy storage. And can you speak to like how the system has kind of changed since then, Yvette?
Pablo Vegas (05:58.338)
Yeah, it’s changed quite a bit. Over 60 gigawatts of new supply added to the Earth Hydrate since winter storm Yuri. That’s a huge number. Last year, we helped connect a record amount of generation supply to the Earth Hydrate, over 16 gigawatts in one year. And we’re on track to have a record first quarter, this quarter in terms of interconnecting new supply. So it has been changing rapidly. As you pointed out, the majority of the supply coming online
Pablo Vegas (06:28.178)
is solar resources and energy storage resources. That makes up the large bulk of the supply that’s connected in SINCE-URATE. And what that’s doing is it’s helping grid reliability through several lenses. So, you know, the battery storage has such a unique flexible characteristic where it is able to be responsive instantaneously when it’s needed on both being able to discharge as a supply source or for charge.
Pablo Vegas (06:56.438)
in terms of balancing for load need. And it can do that instantly at any point in time. And you’re right, back during Yuri, we had in the hundreds of, low hundreds of megawatts of energy storage. And today we have over 16 gigawatts installed, well over that in terms of gigawatt hours, because most of our storage is between an hour and a half and then two hours now. So that’s a fundamental change. And what batteries can do, if you remember,
Pablo Vegas (07:23.182)
during the summer periods of 2022 and 2023, those are some fairly hot summers. 2023 was, I think, the second hottest summer in Texas history. And that’s when we reached our peak. We hit the peak of a little over 85 gigawatts that summer. And we were experiencing conservation calls. We were pretty concerned about the levels of consumption in contrast to the supply we had. A short two years later,
Pablo Vegas (07:49.198)
with the additions of the solar and battery storage that we’ve seen in the last two years, we’ve had similar levels of demand and have not seen any kind of scarcity issues related to that. And so it’s completely changed the tenor and the risk profile of summers in Erickaht. That’s been a tremendous benefit. The batteries also have been a tremendous help during the winter in terms of managing the transitions of the early morning hours and in the solar sunsets in the evening as well. And so it’s really helping with those
Pablo Vegas (08:17.72)
periods that are now emerging as those peaks, it has changed the reliability picture. If you look at where the grid is from a probabilistic perspective today, we are more reliable in terms of the risk of emergencies today than we were five years ago or three years ago. Now, all of that is dependent upon how much load we have today and what the mix is. So when you look into the future, that can change. And so we can’t sit back today and say, we’re good.
Pablo Vegas (08:45.538)
because all these additions have made us more reliable. It’s really our job to look ahead three years, five years, and 10 years ago, is the trajectory gonna continue to remain improving or is it gonna change? Is it gonna become less reliable? And those are the things that we are focused on right now because the growth story ahead of us is significant and sitting back and just kind of looking at what’s happened over the last five years and thinking that will serve us for the next five, I don’t think works.
Matt Boms (09:13.102)
I think you teed that up perfectly. We want to ask you about load forecast. We want to ask you about data centers, load growth. And I think you hit the nail on the head. There’s 49 other states that love to have the grid we have in Texas right now as far as the diversity of our resource mix, the reliability. When you look at costs as well, we have some of the lowest costs in the country when you look at one hour. But when we think about five, 10 years from now, the load forecast that we’re seeing, I guess the first question is, how do you look at that load
Matt Boms (09:43.116)
forecast, do you take any it with a grain of salt? How do you figure out what’s real and what’s not real? And then when you have that number, how do you plan a grid of the future given all of the, not just data centers, right? The population explosion, industrial loads that are coming Texas, how do you plan for all of
Pablo Vegas (10:00.485)
Yeah,
Pablo Vegas (10:01.146)
and if you look at the population, just in the last five years, we’ve seen population grow from 25, 26 million in the state to over 31 million. Tremendous growth. The economic growth from a GDP perspective has far exceeded the U.S. economic growth for the same period of time. So it’s across all the measures. It’s people, it’s the poor economy, and it’s the new economies, the digital economy that are coming, are driving some of the fastest growth measures anywhere in the world right here in Texas.
Pablo Vegas (10:28.654)
As we look at that, the forecast that we received last year, if you all recall, we got a load forecast that was 230 gigawatts for about the 2030 time period. today we peaked at, we have around an 85 gigawatt peak. So it’s almost tripping the size of the aircraft grid if we believe that that’s gonna be the trajectory. I do take that with a grain of salt. We recognize that with these interconnection processes we have today, that
Pablo Vegas (10:56.854)
really it’s very simple to start a process to think about connecting a large load into the grid. It literally takes a computer and a relatively straightforward submission and you can be in a large load queue. Of all of that load that’s out there, about 200 and plus gigawatts of load that’s there today, the vast majority of it, over half of it, hasn’t begun any process at all with formally interconnecting into the grid, other than just submitting an application.
Pablo Vegas (11:25.216)
half of it you take off right away. Then you’ve got like another 70 gigawatts that’s out there that has some start process started but has not advanced far enough to have what’s considered planning studies ready to go. So you’re very quickly down to a level that’s I think a lot more reasonable where those projects that are serious that have backing behind it are moving forward. And that’s what we’re really trying to focus on today. There’s policy discussions going on around how to...
Pablo Vegas (11:52.972)
define the requirements to be inclusive in this forecast in our cod. There’s rulemaking going on, more so actually more than policy rulemaking. And that’s going to be a really important outcome that I think everybody who’s interested in this space has to watch closely because the requirements around that are going to drive everything else. It’s going to be inclusion requirements for being studied in this new batch process that we’re talking about. Sure, we’ll talk a little bit about that. It’s going to drive the inclusion in the
Pablo Vegas (12:21.294)
planning forecast, which is used for transmission planning. And that’s a super critical element. If you think about that, we need to get that number right because transmission investments are extremely long dated. They last for 30, 40, 50 years. They are very expensive. They can cost billions of dollars. And so you have to be building a system that is going to be rationalized and reasonable for the load and the consumers that are out there. And that’s an important part of that forecast.
Pablo Vegas (12:49.23)
But I also just want to add, you this forecast is also used to schedule and plan outages on the system for generators. We use the load forecast to create the space for generators to take outages in the future. And they need that in order to maintain their plans. We use the load forecast to set our budget at ERCOP, what everybody pays for ERCOP to do its job. So it’s such a critical part of all of these facets of what we do. And so...
Pablo Vegas (13:15.97)
Getting that right is critical and this rulemaking that’s happening right now around it is going to be very consequential.
Joshua Rhodes (13:22.894)
Yeah, I was talking to one transmission service provider and they said a few years ago, like three years ago, they had.
Pablo Vegas (13:29.504)
one
Pablo Vegas (13:29.828)
large load.
Joshua Rhodes (13:30.254)
study
Joshua Rhodes (13:30.754)
in their system. you know, last month they had 100 current large load systems like large load studies going on. And this is one of the really only big four big ones in the state, but that’s one of the big ones. And these loads, to my understanding, are quite different, right? It’s different than having diffuse load, like just population growing and overall, because these loads are also points of like, they exist in a particular location in a particular place. And that has particular implications for all the infrastructure and how to get power in and out.
Joshua Rhodes (14:00.12)
to them. And then plus like when you do like a large load study on one, it can impact what the large load study would look like for another one across the system. Because like we can’t scale the infrastructure infinitely. So can you just talk a little bit about what this batch process is? There are cuts like that you’re having rulemakings and proceedings and like meetings and all this kind of stuff. Like what will that do that’ll be different than how it’s done before?
Pablo Vegas (14:21.762)
Yeah, let me back up a little bit and talk about kind of how we used to manage some of the grid stability issues prior to this whole scenario evolving. So we would have what’s called our largest single contingency defined for the system where we plan for the eventuality of a very significant loss of supply in this case. And it would be around one of the nuclear units that operate in the state of Texas. We have large nuclear facilities that are over a gigabyte
Pablo Vegas (14:51.766)
watt in size at a unit level. And what we need to be able to do is there’s the possibility that if something goes wrong in the system, you could see a nuclear unit drop offline. so you have to be able to have the reserves to be able to manage the stability of the grid and the eventuality that something like that happens. And so that was always the way we talked about you manage the eventuality and risk associated with very large changes in a supplier.
Pablo Vegas (15:19.374)
Well, then now comes in the data centers. And data centers, like you said, are concentrated loads in a location. Today, a large data center might be 60, 80, or 100 megawatts, a very large data center under current definitions. In the future, what we’re looking at is sites that are going to have one gig to potentially two, three, or four gigs in kind of a concentrated area. And so that takes that whole paradigm of
Pablo Vegas (15:45.998)
planning for stability and single contingency issues to a whole different degree. Because now it’s not just a potential supply unit coming offline. You now have loads that have the same risk where they could trip offline and they’re larger than your otherwise previously defined single contingency. So now we have to start thinking about planning the grid differently with reserves in order to be able to manage that eventually. that just to add that dimension into the planning, that’s one thing that is
Pablo Vegas (16:15.83)
really fundamentally changed and why it’s so significant, the fact that these very large loads at a single site add a lot of operational complexity. So that’s something I think that’s important to understand. The batch process that we’re moving towards is necessary because today when a load comes, like you said, historically we’ve had, know, a transmission company would have one, two or three large loads that they’d be working on over the course of a year. And today there’s dozens of them and sometimes hundreds of them depending on.
Pablo Vegas (16:44.622)
company we’re talking about. And what’s happening is that our process effectively works closely with the transmission operator. And we do these planning and stability studies to ensure that the grid and the transmission system can stably connect a large load into the grid and can serve it reliably across 8,760 hours. We analyze literally every hour of the year to make sure that that site can receive energy reliably and stably throughout the
Pablo Vegas (17:13.678)
entire year’s period. And so when we do these studies with a transmission company, we go back and forth over a period of time. It could be six, 12 months worth of work to do that. And then they will get a planning study approved. And that’s kind of a stage in the interconnection process. The planning studies are approved, ERCOT has signed off on it, the transmission company has signed off on it. And effectively then what the customer does is then they build their site. And so then they build a site that could take one, two or three years to build a site. And then right before they come online, they go through one more check.
Pablo Vegas (17:43.278)
and they go through us all energy stabilization check essentially. And so we just make sure that if nothing has changed, you know, since the couple of years before when the planning studies were approved, now we have the as built so we know it’s actually there. And so we do a last table of stabilization study and give them a green light to energize. That’s the way the process has worked for the last 20 years. Well, imagine what’s happening now with data center that gets in a planning study approved, they start constructing a site.
Pablo Vegas (18:11.886)
they’re building their site starting at, let’s say, 120 megawatts in the second year, and then they’re ramping up to, let’s say, 800 megawatts over the next five years. While they’re doing that, 10 other projects within the same geography are doing the exact same thing. So the planning study that was approved for that company two years prior, all of sudden the topology of the grid looks vastly different than it did when the study was approved. And we’re sending them into this never-ending re-study loop.
Pablo Vegas (18:40.718)
where we have to go back and say, the study, the planning study you just had approved has changed materially because the entire transmission system in your geography has changed based on all of these other projects that have come in. And so we recognized pretty quickly that this was the bottleneck we were heading towards and the uncertainty that that was creating for loads was really very, very destructive to the stability that you need for massive investment. So these data centers are huge investments.
Pablo Vegas (19:10.626)
The batch studies is intended to solve that issue. So what it does is it takes, and basically we’re working towards what we call batch zero, which is to establish a baseline where all the projects that are getting ready to come online in the next couple of few years are going to go through this batch study and establish essentially the requirements and demands on the transmission system in order to serve all of these loads. It will then create an allocation of transmission that’ll be held for those projects.
Pablo Vegas (19:40.37)
Because historically we’ve never held transmission. Nobody in the US holds transmission for a project. You do a study and then you build and you assume the transmission capacity will be there when it’s ready, but there’s no holding of an allocation of transmission. Well, in this world we have to do that because with this Batch Zero we’re going to have a bunch of projects that will get approved, the transmission will be allocated for them to build over the next three to five years, and then every year we’re going to have subsequently more data centers coming in looking to interconnect and looking to grow.
Pablo Vegas (20:09.71)
or manufacturers or other industrial customers or any large load. And so we have to do that with the knowledge of what’s been committed historically, so we don’t derail those projects. Then we have to be able to support the new ones coming in and then define the transmission investments needed to serve all of them. So that’s what this batch study is gonna do is effectively do all of them together on a cycle of somewhere between six and 12 months and then each year rinse and repeat.
Pablo Vegas (20:36.814)
come up with the study, evaluate what’s available on the system, determine what needs to be built to serve, and then continue to remove that cycle during the period.
Joshua Rhodes (20:45.89)
That’s follow
Joshua Rhodes (20:46.91)
up. When I teach like electricity markets and electricity systems at UT,
Pablo Vegas (20:52.086)
I use
Joshua Rhodes (20:52.174)
just to skip, really kind of skip over the part of like, if there’s too much generation and not enough load, like that never happens. That’s not a big deal. That data center trip, gigawatt data center trip offline and PJM and like creating havoc, right? And then you’re right to bring that up. That is a new operational paradigm, right? It’s like, it’s not just like losing supply. It’s like losing demand.
Pablo Vegas (21:13.59)
It has the same electrical effect. What’s unique is that the way large electronic loads often behave is that when they see a fault on the system, what they’re designed to do is to protect the assets behind the meter, right? Because you’ve got super expensive GPUs and equipment behind the meter. And so they’ve got UPS systems that they can flip to instantly when they see something like that to protect those assets and then a couple cycles later reconnect to the grid. So we need to make sure that we understand when they’re going to do that and what are going to be the performance requirements
Pablo Vegas (21:43.545)
so that then we can plan around that. Because like you said, once we start to see a lot of these gigawatt sized facilities spread throughout the state or clustered in areas on the state, that becomes a very unique risk.
Matt Boms (21:57.282)
think the next logical question here, Bob, is we’re talking a lot about the transmission grid. I think Josh and I have spoken about this at length on the podcast and how.
Matt Boms (22:06.508)
We have this really successful experiment in Texas that’s called deregulation. And we’re coming up to 25 years, almost, of deregulation in Texas that I think most of us would agree is a resounding success given all the generation that we’ve brought on. However, on the distribution grid, that’s where I want to ask you a couple of questions because it feels like as a side effect of that experiment, we have utilities that are responsible.
Pablo Vegas (22:33.112)
for
Pablo Vegas (22:33.71)
capital expenditure.
Matt Boms (22:34.734)
And
Matt Boms (22:35.214)
we’re telling them not to dip their toes generation, right? They can’t, the T.D.’s in Texas are not allowed to generate. And what that means is, and I think you deserve a ton of credit for this because you’ve come out and spoken about the importance of demand response. You’ve spoken about the importance of DERs, behind the meter technologies, energy efficiency, all of this. And I think that’s probably even more important now given all the low growth that’s coming. So not feel so hard to crack given the regulatory structure we have in place.
Matt Boms (23:03.266)
The fact that we know that substations will be overloaded in the future because of all the load growth that’s coming. In your mind, is there the magic wand solution to this problem or where do you see this going over the next few years as far as DERs, demand response, energy efficiency, all that?
Pablo Vegas (23:19.084)
You started the question with kind of the recognition of how important the competitive market has been to the success Texas has experienced economically. And it is because it’s driven low cost energy that has helped the economic growth in this state for the last 20 plus years. And it’s continuing to do that. So that’s something that has been an incredible benefit, I think, to all Texans because of that construct. I believe that construct can be.
Pablo Vegas (23:45.096)
extrapolated deeper into the grid at the distribution level to leverage competitive forces on solutions and services that are now becoming more viable than they ever have been in history. Those types of solutions and services are distributed resources that are going to be at premises, and that could be batteries at a home, can be smart thermostats and smart coal pumps.
Pablo Vegas (24:14.218)
It can be rooftop solar. It can be electric vehicles. All of these components that are behind a meter at a residence or at a commercial site can provide a functional resource service to the grid at large, especially when it’s aggregated in total. And that’s something that we have proven already with our ADER program, which is our Aggregated Distributed Energy Resource program, that I think has been highly successful. We just recently
Pablo Vegas (24:43.842)
doubled the cap on that program and we’re looking to increase it even further because we’re starting to really see momentum pick up on these types of resources proliferating throughout the Texas grid. And when you think about that, I mean, it’s almost like the business model of an Uber or an Airbnb where you’ve got distributed capital. So consumers and business owners are investing in energy resources.
Pablo Vegas (25:11.682)
to help them manage their reliability and manage their costs. And those same assets that are providing those benefits to the consumers can be double leveraged by the grid operator to provide services to the grid and provide another income stream or economic incentive for those same assets. So they’re making the capital investment in putting rooftop solar on or buying an electric vehicle, but with the proper integration and aggregation, they can become a meaningful part
Pablo Vegas (25:40.75)
of a supply or demand response mix. And that’s really unique and it’s growing. We’re seeing the evidence of it working. We actually have these aggregated resources inside of our energy only market providing energy. We also have them in our ancillary services. So if you think about that, those are our reserve services that we are really protective of because we lean on those in order to keep the grid reliable. Yet we’ve seen that the reliability and effectiveness of these aggregated resources
Pablo Vegas (26:10.862)
can perform under the ancillary services requirements too, in the ECRS and in the non-STIN services. So it’s really a tremendous, when you think about the potential, we’re just getting started with that. And we’re just starting to see companies that are in this space start to really take hold and grow. So I think we’re in the very earliest innings of this trajectory. And I see a future where the grid operator is really moving out to the edge of the grid and working with
Pablo Vegas (26:39.756)
consumers and the retailers and the aggregators and those that are interested in bringing that competitive creativity to this space and finding ways to actually be very economically successful with them.
Joshua Rhodes (26:52.046)
We’ve
Joshua Rhodes (26:52.176)
talked about low growth changing. We’ve talked about, you talked about ancillary services. We’ve talked about how the supply mix has been changing. And then on top of that, we also recently changed a pretty big fundamental market design, like a little while ago, in terms of doing this real-time co-optimization plus batteries, and kind of how we actually are dispatching resources and how we’re procuring those ancillary services we’ve talked about. Can you talk a little bit about what RTC plus B, real-time co-optimization plus batteries,
Joshua Rhodes (27:22.0)
what that is, how that’s a bit different and like, well frankly, how’s it going? It’s going for a couple...
Pablo Vegas (27:25.806)
It’s been months
Pablo Vegas (27:27.026)
now. It has. Yeah, we launched that went live on December 5th of 2025 and it’s been going really well. We saw during winter storm Fern the operational benefits of this that we hope to see because what effectively we’re now doing is rather than taking the commitments for the ancillary services the day ahead and effectively holding them in those positions throughout the day in real time.
Pablo Vegas (27:53.858)
What we’re doing is in real time evaluating at every interval of the market, whether it’s more efficient from a cost perspective to hold one of the reserves and ancillary service with this resource or that resource. And we take the lowest cost resource, move it into the energy service space so that the dispatch comes down in its cost. And then the more expensive resources we could actually hold in the ancillary services in the case that we need those.
Pablo Vegas (28:20.312)
We’re not losing anything in terms of the amount of ancillary service and reserves we have, but we’re really optimizing the cost effectiveness of that. We are forecasting to be able to save at the wholesale power market level more than $3 billion a year just by doing this type of real-time optimization between what’s serving energy and what’s being held for reserves. We’re seeing that financial trajectory realized. We’re seeing it during the periods of
Pablo Vegas (28:48.13)
more scarcity, which Winter Storm firm a few weeks back gave us an opportunity to see that from an economic point of view. We also, so I mentioned the operational benefit, what it does is it will take a look at where the constraints on the system are, our dispatch system will, and it will move around between energy and ancillary services, what is providing energy in order to solve those constraints on the system. So it’s gonna have the opportunity to help bring down congestion.
Pablo Vegas (29:17.074)
in the system as well. And we saw that when we were seeing some constraints coming up into the San Antonio area on that weekend of winter storm firm, there were some constraints on the transmission line feeding the San Antonio area from the valley. And we saw the RTC plus B optimization activate some of the energy resources north of that constraint to alleviate that transmission congestion. So it’s looking at the most economic way to dispatch as well as the most reliable way to dispatch in order to manage the overall congestion of system too.
Pablo Vegas (29:46.71)
And so it’s working extremely well. And then with the plus B, we built it from the bottoms up, thinking about the role that batteries play and optimizing how to dispatch batteries in order to, again, optimize the cost of serving energy and ancillaries. And batteries continue to be such a core part of that operational flexibility.
Joshua Rhodes (30:06.766)
Can I follow up with one question? There has been some disagreement in some of the players around which of the ancillary services certain resources like duration limited and storage like duration limited resources like batteries can play. Can you just speak to ERCOT’s stance on kind how that’s being implemented and like kind of how why y’all see it that way?
Pablo Vegas (30:28.046)
So to understand your question, we have always wanted to try to understand what the battery operator is going to do because we don’t control the dispatch, we don’t control the state of charge for batteries. What we do is we set a set of requirements out for what the various ancillary services need to be able to perform and how we’ll plan around them. So non-spending, you have to be able to come online in 30 minutes, you have to be able to run for four hours in order to serve that block.
Pablo Vegas (30:56.142)
ECRS, you’ve got to be able to come online in 10 minutes. And you have to be able to run through these different wheel-throughs. And we’ve changed some of those definitions to actually shorten the requirements. So at ECRS, now it’s really a one-hour product with the latest iteration. So what we were talking about with batteries before was, if there’s a two-hour or four-hour obligation, we need to make sure that you’ve got that in case you’re called upon to provide that service. A lot of that has been resolved with the way RTC works. And so with real-time optimization, it
Pablo Vegas (31:25.216)
knows what is available in the state of charge because it has to get updated and reported by those resources at every interval. And then it will make the decision based on what’s in the tank on what to put into the next, you know, the next dispatch stack. And so I think we’ve become more flexible with the batteries and it’s become less of a concern around some of these requirements. And frankly, we really like what we see in the batteries here in Texas. There’s a lot of talk about, you know, how do we get more duration?
Pablo Vegas (31:51.97)
We would love to see more duration, but we don’t want to lose the flexibility that we have. A one-hour battery that has one hour of, let’s say a one megawatt battery with one megawatt hour of energy means that it can take its entire charge and give it to the grid inside of an hour. A one megawatt battery with four megawatt hours of energy means that it can provide that one megawatt hour over four hours, right?
Pablo Vegas (32:20.97)
Sounds good and it is good. Duration is good. But if you wanted to put those four megawatt hours on the grid in one hour, you can’t do that. And so that’s why the Ercob Market with our very flexible rules, we’ve been attracting those shorter duration batteries because they have the most flexibility to perform when the grid needs them most. You know, during those solar sunsets where we’ve got an hour and a half to two hours of really, you know, peak need, the entire battery community can put all their energy on the grid at once.
Pablo Vegas (32:50.712)
Whereas if they were under the California model, they have to all have four hours of energy behind them and they can’t dump all that energy in a one or two hour period as needed. So you potentially are holding a bunch of economic energy behind a limitation on the inverter that can’t put it all out of the system. So we really like the flexibility that we have with the batteries today. It’s not saying we’re not looking for more duration. Duration is always good. Right now, the name of the game is really flexibility and the way we’re structured gives us the flexibility.
Matt Boms (33:20.878)
We’re coming up on time almost, and I wanted to make sure we asked you about residential demand response. But look, it’s worked its way through different subcommittees. And again, I think your leadership on this issue has been really important. And when folks ask, like, residential customers in Texas why the industrials have access to demand response programs, but for the most part, residential folks don’t, how do we solve that problem? And where is the challenge as far as just getting all the stakeholders on the same page and getting a new program out there that we can also
Matt Boms (33:50.882)
And it’s maybe that the solution is a Texas solution, which is like, it might not be perfect, but let’s fix it on the fly.
Pablo Vegas (33:56.334)
Pablo Vegas (33:56.794)
I couldn’t agree more with you with that last statement that you just said. Every time we introduce something new, there’s always the potential that it could be improved upon. And that should never prevent us from taking the step forward and learning through that experience. And I think that’s right where we are right now at the precipice of considering some ways to do residential demand response, where let’s start by clarifying, residential demand response is available broadly across Eurot. There are many municipalities that have
Pablo Vegas (34:26.388)
successful programs and competitive retailers are offering demand response incentives as well. But we believe that there’s a lot more that we haven’t tapped into yet and that’s where I think the opportunity lies. And so what we’ve been talking about is how can ERCOT step into that space, provide an economic incentive on top of what the competitive retailers and the municipalities are already doing in order to grow.
Pablo Vegas (34:52.002)
the share of residential demand response that’s available to the grid. And that’s the opportunity set. It’s not about replacing or substituting what’s already happening. It’s about augmenting it to make it more valuable and to bring more people into the programs. And so the issues with moving this forward are the issues that happen with almost any change in our market. There’s always very different stakeholder needs across the spectrum that are looking at, what’s the impact going to be on wholesale prices? What’s going to be the impact on
Pablo Vegas (35:21.314)
potential economic revenues down the road. And so I think we have to think through balancing all those different interests. But in the short term, what I would say is we know we’re going to be faced with challenges with this low growth that’s coming our way. We know demand response is a relatively low cost and absolutely quicker timetable of a product that could be developed than building new supply, as an example of building new transmission.
Pablo Vegas (35:48.418)
We spend a lot of time talking about building new supply and transmission, but we could create a demand response program in a matter of months and get it up and running and get the potential reliability benefit of it sooner. And that’s the approach I think we should come into this looking at, saying let’s put a tool in the toolbox. Let’s try it. Maybe it won’t be perfect. Maybe it needs to be improved significantly after we start it. Let’s put the tool in the toolbox and start learning from it. And then if we realize we need this and we need more of it,
Pablo Vegas (36:18.03)
and it needs to be better, then we can work on it and change it. But we’re right at that junction now where we have to kind of get over that hump in order to move it forward. And it’s really about a a stakeholder policy discussion happening across the stakeholders in the state. And we need to get to that decision.
Joshua Rhodes (36:36.844)
That’s it. Before we get to closing remarks, I kind of want to...
Joshua Rhodes (36:40.072)
ask a little bit of a different question because you get folks like us on the panel and we’re talking about energy and AI. We’re talking about, how are these data centers? But I kind of want to talk about like, how can the electricity sector use AI to maybe do its operations better? I think the implementation of AI and big systems is a little bit of a mixed bag right now. We’ve seen lately we’ve seen Amazon having to come to Jesus moments with their engineers saying you allowed AI to cost 13 hours of interruption. Now, if I don’t get my phone case for next to 13 hours, not a big deal.
Joshua Rhodes (37:09.936)
from Amazon, but if I don’t have power for 13 hours, that is a big deal. But then PJM has recently, another electricity market on the East Coast, has recently implemented more extensive dynamic line ratings. So they’re able to change like the ratings of their transmission network given ambient conditions. Like if the wind’s blowing and the lines are cooled down, you can move more power over them than you automatically would. And some of the underlying machine learning techniques and all those other kinds of things are similar across like these large language models, neural networks versus.
Joshua Rhodes (37:38.902)
other types of models. So I’m just curious to the extent you can, like how or can speak about it? Is ERCOT using AI? How can ERCOT use AI? Are you having talks about it? Like what is it in the use of AI of electricity you’re seeing?
Pablo Vegas (37:51.234)
Yeah, I think the really great irony of all of it is that the big challenge for talking about how to manage the growth and the power requirements, I think can help get solved by the underlying technology that is driving those demands. So in our world, we talk about AI through the lens of people-led AI. That’s the first thing is that this has to be people-led. We’re not going to turn over the systems and the controls to machine learning or to AI, but we’re going to keep it.
Pablo Vegas (38:18.156)
you know, inside of the central function where the people with experience are going to leverage it to help them do what they do better. We are looking at AI in a lot of different areas. So from the obvious load forecasting. using AI to take dozens of different weather forecasts and different load forecasts and be able to look at historical experience and then look at, you know, the probabilistic changes in terms of weather and load and be able to run scenarios.
Pablo Vegas (38:47.326)
instantly getting you a spectrum of likelihoods of what could occur over the next five minutes. We’re starting to introduce AI into that process so that then as we experience what’s going on in real time, the learning occurs and it improves and we’re constantly getting better with kind of the process of being able to load and weather type forecasting to affect that part of our operations. In another area around queue management, there’s process of running studies, building cases,
Pablo Vegas (39:16.15)
and running scenarios, AI is very well suited to help facilitate and improve that because we take data from lots of different sources to build these cases, and then we run them through these stability or dynamic models. so AI can help with doing all of that data gathering, running these processes, analyzing the results, looking for anomalies, and giving the engineers, here’s the places you should look at in terms of these stability results or these dynamic results.
Pablo Vegas (39:41.91)
So we’re starting to introduce AI into the core planning and modeling of what we’re doing. So it’s starting to come into so many different facets of our business and it’s going to continue to grow its capabilities. There were some releases of the AI large language models in the last two months that have had meaningful step forwards in the autonomous capabilities in terms of how long they can actually run without a human intervention. And we’re starting to see the benefits of that in our coding, our development.
Pablo Vegas (40:10.542)
Because if you look at Erkop, we’re effectively a big tech company, right? We’re running a tech platform. The biggest part of our workforce are technology engineers. So it’s becoming a really important part of how we’re thinking about real-time operations and planning. And we’re going to continue to look at, to leverage that to get better on every level of reliability as well as possible.
Joshua Rhodes (40:14.659)
your
Matt Boms (40:30.606)
Well, I think we’re pretty much at top level. And Josh, I just want to thank you so much again. We know you’re busy and the fact that you took a few minutes out of your day to speak with us really means a lot. And anything you wanted to add at the end, I’m feeling optimistic. I feel like this is a huge challenge that we have here as a state, but probably no state is better positioned than Texas to meet this.
Pablo Vegas (40:49.582)
I agree.
Pablo Vegas (40:50.303)
I’d love to just get a quick perspective from either one of you. So you guys work inside of Texas, but you also have the opportunity to kind of look outside beyond our four walls. Are there things that you see outside of our four walls that you think are opportunities for us to learn from? Or are there things that, you know, perhaps we should be looking to export more of in order to try to help this endeavor? Because a lot of what we do is oriented to Texans.
Pablo Vegas (41:16.418)
We really are part of the national economy, we’re core part of it. I think we should always be looking broadly in terms of how we can help others in the industry be successful.
Joshua Rhodes (41:25.088)
I felt like I spent most of my time doing the opposite, getting called in from other places to talk about how we’re doing things in Texas, to be honest with you. Just given how fast we’re able to move and like, you everything from our connect and manage approach and just like this RTC plus B, how fast we’re able to move around. I’m going to kick it over to Matt, see if he has anything while I try to think on my feet really quickly.
Matt Boms (41:47.128)
Well, the one that comes to mind is there are states that do really good job on DERs through TND and deferral.
Matt Boms (41:54.284)
which looks different in Texas because of the problem we were talking about earlier, right, where we have our utilities responsible for just infrastructure, right, just hard transmission and distribution infrastructure. And I think Josh and I agree, like on the transmission side, we’re doing pretty well because you’ve got really good price signals there on the transmission grid, but you don’t have things on the distribution grid because they’re essentially regulated.
Pablo Vegas (42:17.71)
You
Pablo Vegas (42:18.429)
have to stay on that space.
Matt Boms (42:19.576)
Exactly. So I think that’s where we have some room for improvement. Like a study just came out showing, I think almost $2,000 per rate payer over the next 10 years, if you’re able to crack that T and D deferral piece. So I think that’s what I’m looking towards moving forward here.
Joshua Rhodes (42:35.212)
I guess the only thing is that I would say is.
Joshua Rhodes (42:38.178)
We talked about congestion and we talked about the price signals that we do see. Maybe the large loads and everything and all the transmission that we’re going to build will kind of alleviate this for a bit. I do think we can do a better job of leveraging the economic test for new transmission than we do. Because essentially, if we build a new transmission line, please correct me if I’m wrong on this, as I’m speaking to the CEO of Vircoq. We build a new transmission line on the economic test. It essentially has to pay for itself in one year’s worth of savings.
Joshua Rhodes (43:08.058)
mentioned earlier, transmission is a multi-decadal asset. so anytime that I’m talking about everything that Texas does great, I usually am like, well, there’s this one thing I really think we can do better. And that’s what I would say is like, if there’s any way the economic test for new transmission could be more, at least on the order of magnitude of the length or the lifetime of the asset that it’s paying back, like, I guess that would be my only feedback.
Pablo Vegas (43:13.078)
about
Pablo Vegas (43:31.63)
I think that’s really good point. And in fact, I’m glad you brought that up because it is something that we are working on right now. And there is another region, other regions use this concept of an MVP, a multi-value kind of a proposition for evaluating the effectiveness of transmission investment. And what you can do is you can take benefits across reliability, benefits across economics, and benefits across resiliency and look at the three components and say, okay, well, you may not hit 100 % on
Pablo Vegas (44:00.776)
one or the other, the combination, the multi-value combined creates enough of the hurdle, overcomes the hurdle to pass the requirement to build that transmission. We should be looking at transmission to that same ones because there is real value in resiliency, there’s real value in economics, there’s real value in reliability and combining those metrics to come up with a combined value proposition for a project is a much more effective way than trying to just fully clear one hurdle or fully clear another.
Joshua Rhodes (44:29.816)
Pablo Vegas, thank you for coming on the Energy Capital Podcast.
Pablo Vegas (44:32.13)
Thanks,
Pablo Vegas (44:32.422)
Josh. Appreciate it. Thanks, man.
