As the data center buildout in Texas accelerates, the public conversation has fixated on generation, interconnection queues, and gigawatts. But the firms actually structuring these deals see a different problem entirely: process.
In this episode, Joshua Rhodes speaks with Maura Yates, chief executive of Mothership Energy. Mothership is one of the most active retail electricity providers in ERCOT’s large-load market, managing more than three gigawatts of large load and writing more than 39 distinct data center contract templates to handle the variation across deals.
Yates says the technology to power data centers exists. The bottleneck for data center completion is the time it takes to sign contracts for electricity service and the time it takes to connect them to the grid. As Texas debates SB6 implementation, co-location rules, and demand-side management, that distinction is shaping which projects get built and when.
Joshua and Maura discuss topics including:
Why Mothership has written more than 39 distinct data center contract templates.
The difference between behind-the-meter, front-of-meter, and co-location deals.
What the Crusoe Goodnight data center PUC ruling means for future co-location projects.
Why most data centers are data center companies, not power companies.
HowYates expects ERCOT to integrate large load without pushing costs onto residential customers.
Timestamps
00:00 - Introduction & Maura Yates
01:26 - Building Mothership from Both Sides of the Market
03:27 - The Problem Mothership Was Created to Solve
05:51 - White Label REP Model and Getting DERs into ERCOT
08:37 - Option One vs. Option Two Retail Licenses
10:42 - Selling the Residential Book and Pivoting to Large Loads
13:27 - Entering the Data Center and Bitcoin Mining Space
16:24 - How Mothership Structures Data Center Deals
19:13 - What the Market Needs: Process Over Technology
21:24 - Co-location, Net Metering, and BYOG in ERCOT
26:13 - Do Data Centers Actually Want to Be Power Companies
28:12 - Eclipse: Mothership’s Market Access Platform
32:01 - Forward Curves and Empowering Price Takers
38:07 - The Grid’s Future: Distributed Supply and Data Center Growth
41:03 - Closing
Resources
People & Organizations
Company & Industry News
Related Podcasts by Energy Capital
Transcript
Joshua Rhodes (00:04.366)
Hi, everyone, and welcome to the Energy Capital podcast. I’m really excited to have Mara Yates here to talk about mothership energy in ERCOT. Mara Yates has over 20 years of experience in the power industry, starting in distributed energy resources at Arizona Public Service before moving to governmental affairs at Sun Edison and then becoming VP of sustainability at NP2 Energy, which was bought by Shell in 2017. In 2021, Mara co-founded and is the CEO of Mothership Energy Innovations with her business partner, Caitlin Brammer.
Mothership is the 18th largest retail electricity provider in the nation and is contracted and manages over three gigawatts of large loads in ERCOT. In addition to being a rep, Mothership provides risk management, technology and consulting services to electric cooperatives in Texas and specializes in data center and wholesale procurement. Yates, welcome to the Energy Capital Podcast.
Maura Yates (00:54.712)
Thanks for having me. Super excited to be here chatting with you all.
Joshua Rhodes (00:58.05)
Yeah, and we’re again, super excited to have you. Before we dig into mothership and your current role, like looking back kind of where you’ve come from, your LinkedIn started in a regulated electricity space at Arizona Public Service before kind of going into more corporate with Sun Edison and MP2 and Slush Shell. And now you’re CEO of like a company in the really competitive space. How were those different, those roles that you have in the different spaces where they
Maura Yates (01:26.35)
That’s a really good question and it’s one I enjoy answering because ultimately it’s that background that has allowed us to create incubator mothership innovations to be the shock that it is. So when you look at the places that have spent time, I’ve spent time at the utilities, I spent time at the development side, both a DER and large scale solar development shop, all the way then to a deregulated pseudo rep slash utility in ERCOT. And really just bounce back and forth from utility side.
to developer or industry side, back to a version of the utility, back to the industry side. And now I’m back on quote unquote the utility side or the delivery side and supply side. And where we’ve landed with Mothership is really taking the learnings of participating and being a part of a company on both sides of the transactions, right? So you typically have the utility on one side and the developer on the other side.
And so by jockeying back and forth between these spaces, between our regulated utility all the way into the deregulated market, it’s given us a really well-rounded understanding of how these different parties participate in the market, but also really what they find important and what they find valuable and like, what is it that the utility is trying to get out of transactions? What is it the developer is trying to get out of transaction? And when you better understand both sides of a transaction, you’re able to come up with a better solution for the transaction.
which is why, again, we’ve developed mothership in its latest stage to really be this deal shop and this service provider for both developers and other load serving entities and utility providers. And I should clarify, I’m using utility and like not the traditional ERCOT sense of TND, but really the one providing and servicing power to the end user.
Joshua Rhodes (03:10.094)
Yeah, no, that makes sense. like, can you kind of tease out a little bit like of that? I seeing both sides of kind of how things operate is super valuable. So as you were seeing both sides, can you kind of tease out what was the problem that mothership was originally created to solve? Talked about both sides, but what was that problem there?
Maura Yates (03:27.596)
Yep. So the more time that we spent on both sides of these deals. So for example, I go back all the way to Arizona Public Service when we structured some of the first utility incentive programs and we were very aggressive, like in a very solar friendly way in terms of the incentives we were offering. So we worked really closely with the industry and the solar and battery developers. And our goal there was to better understand what is it that gets you to transact, right?
Like if our goal as the utility was to get solar adopted by homeowners, what is it that the solar homeowners need to adopt? And what is it we as the utility then need to provide to enable this transaction? And we did a good job. Typically the utility or the load serving entities role is to provide some form of incentives and compensation. And we did that and we had success. We deployed a lot of projects, but incentives aren’t the most sustainable mechanism for growing an industry. So you have to continue to find value and way to
extract value as you move away from incentives as the technologies become more competitive. When you start moving down the advancement and the maturity continuum, you start to realize that the challenge in getting these technologies deployed from both sides is really the customer experience. Something has to change in the customer experience. Everybody talks about it, both utility talked about the customer experience and the developer talked about the customer experience.
But they talked about it differently and it’s the same customer. So our point was this customer is facing a very bifurcated customer experience. They would get a bill from their solar developer. They’d get a bill from their retailer. They said that solar developers sold them on one thing, but then when they got this bill, they can’t make sense of it. They’re speaking two different languages. And then all of sudden you have a bad customer experience and the customer’s not happy with their solar. Their expectations weren’t met because they were set the incorrect way.
And so when Winter Storm URI happened in 2021, this thesis really became a more apparent need. We observed that if more distributed energy resources had been in the market, they would have had a massive impact on what happened during URI. And when you take that step back and say, why weren’t more DERs in the market? For us, the answer was the value proposition is too complicated. It’s still too much of a bifurcated experience.
Maura Yates (05:51.566)
for the customer. So our original thesis at Mothership was and is to be a white label retail electricity provider. However, in the beginning, we were focusing on these DERs. Since then, we’ve now migrated to the other end of the spectrum and do very large load. Yeah, kind of getting further from the beginning. But the idea with the original launch of Mothership and our first clients that we launched with were, hey, we need to get more customer-cited, residential-cited assets in the market.
Joshua Rhodes (06:07.202)
We’ll get that.
Maura Yates (06:21.582)
because that’s where we’re going to see reliability. In a market, we’re going to continue to grow to have challenges because that residential load shape is the one that creates that fall. And so as such, how do you do that? How do you get customers more comfortable with adopting? And our thought was we got to turn it into one. The solar company, it all has to be the same customer, the same economics that are all being balanced between one transacting party rather than a solar company and a retailer that both have their own.
value props in their own metrics. And so if you put it together, it starts to become a more seamless customer experience. Now, the challenge is when you’re asking these solar companies and these distributed energy providers who I’d spent my career working with to become a retailer and a market participant, that’s a really different risk profile than what they’re used to. You go from talking kilowatt hours to megawatt hours. You go from talking about, I need to deposit on your PPA.
to I’m getting a margin call from the market, I got to post collateral. It’s a very, very different transaction world. And so when you think about it, if you’re trying to blend these two worlds together, you’ve got a DER company and you’ve got a traditional retailer, who’s going to embrace the other, right? Is it the retailer that becomes the solar company or is it the solar company that becomes the retailer? And it is our thesis that really these solar companies are best suited to become the retailer.
because it’s these solar companies and DER providers that have that long-term touch, that long-term anchor agreement. And more importantly, that’s what’s leading the conversation with the customers. So we need to figure out how to get them enabled. They’re not quite really prepared to be a rep. so, mothership, the origin of mothership is we would white label this to help them get into the market. We’d show them how to be a rep. We’d teach them.
We’d help them understand where risk is, how to mitigate it, how to use their assets for it, and really understand what a P &L looks like, all with the anticipation that once we show you, get it set up for you, we want to send you on your way, you’ve grown up, and go do it on your own. So the original was really designed around a customer experience to figure out how to get more DERs in the market.
Joshua Rhodes (08:37.048)
This actually brought me to something when I was researching a little bit on moral mothership, the concept of an option two rep, which I’ll be honest, I had to look up. I didn’t really know. Like when I think of reps in the Texas space, I guess I’m thinking of option one, that your general like retail electric providers that most people probably are signed up with or using. So can you explain what an option two rep is and kind of how it operates in the space and how it’s different?
Maura Yates (09:00.984)
So Josh, the question around option two is really interesting because at Mothership, we actually operate a bunch of different retail electricity providers, a combination of option ones and option twos. historically, majority of retailers in the market had been option one. Option one is what allows you to serve Rezzy, small com, and everything that’s essentially sub one megawatt or even greater than one megawatt. But it’s really the universal type of license.
And at Mothership, we operate several of those. And I should clarify, under each license, you’re also able to do these things called sub-LSEs and even further segregate and divide. And it’s a key thing that we do at Mothership for risk isolation, but also data. It gets way more clear and pristine data that’s not co-mingled when you do this type of segregation. So that’s the option one piece, but then we also operate some option twos, as you’ve noted. And with an option two,
Historically, it’s been used for really large, more sophisticated industrial loads that have come into the state and didn’t necessarily need to go through a retailer. They had enough load, or a traditional retailer in option one. They had enough load themselves. They perhaps had enough sophistication. You can become your own, with an attestation, a retailer designed for just one megawatt and above loads. So a load will attest that they will be serving and employing power themselves under this setup. And as a result,
it’s less regulatory requirements because you’re just really focusing on that load that has attested that this is their option rep. So from a regulatory standpoint, a lot lighter, but again, really geared towards these larger facilities. So we anticipate seeing a growth of option two retailers in the future.
Joshua Rhodes (10:42.894)
That makes good sense and I think it’s actually a pretty good transition point into kind of where Mothership is. So it looks like in 2025, like Mothership actually sold a large book of retail customers to Atlantic Energy. And can you explain that pivot and kind of like where you were looking to go like with that?
Maura Yates (11:00.268)
Yeah. Great question. I hope this doesn’t offend everybody in the resi space, but those in the resi space know exactly what I’m talking about. Resi is just a different beast to serve. The risk profile is different. In ERCOT, historically, when you serve resi load, you are short the market. You’re always short the market. And the reason you’re always short the market is because it’s what resi load that was driving this volatility. The load shape, the elasticity, resi is super volatile. And it’s really hard.
to reliably serve that and also do it in a cost effective manner for both the retailer and the load. I mean, I can create a retail rate for a Rezzy customer that has me de-risked, but that’s not a rate a Rezzy customer is gonna pay. And so it’s really the question of for the rate the market will bear, how much can you de-risk it? And it’s hard to de-risk it given where the market pricing is. So we chose to exit the Rezzy market for the general challenges of Rezzy as a result of us
receiving a very large influx of requests from very large loads. So as you can imagine, the book that we sold to Atlantic was a rather large book of RCEs and it seems like 100,000 meters. We can serve that same capacity with maybe one or two large meters customers. So for us, it just became an efficiency standpoint. However, I will also say super grateful and appreciative of our time of doing Rezi because
They’re a huge piece of the market. And unless you understand how Rezzy behaves in the market, what that load does to real-time SPPs, what it does to the load forecast, until you understand how Rezzy behaves, you don’t fully understand the market. So we understood it. It’s innate. We felt it. We’ve hedged it. We’ve worn that risk. And it gives us a good idea of how the ERCOT market behaves as a result.
Joshua Rhodes (12:50.21)
Yeah, that’s one of the things I’ve always said and like in a lot of my research and stuff at the University of Texas and other places, like the residential consumer was a swing consumer and it was the one that increased the most whenever the temperatures either got really cold or really hot. I hear you, got to understand kind of residential in Texas just because of our air conditioning load, electrification of heat, all these other types of things. So you sold a large book of retail customers and you mentioned it just a second ago where you can serve the same amount of load with just one or two large meters.
versus hundreds of thousands, potentially, of smaller meters. So can you talk about where that pivot went to in terms of what are you all up to now?
Maura Yates (13:27.532)
Yep, great question. So a lot of the same thesis that caused us to start the business of changing the customer experience, bringing more nimbleness and innovation into the resi space was what allowed us to gain a reputation outside of the resi space at the same time. a lot of really complex conversations started coming to us. Folks knew us from our past lives when we spent time at MP2.
And they were looking for a lot of really complex deals and feedback on how to do something boutique. And they knew that we did that in our past life. So they asked, hey, can you guys start doing it again? And it was a really important time in the market. So post 2021, that’s when you started seeing a lot of the Bitcoin mining and the data or the crypto market moving to town. And at first, I think my initial response when someone said, do you guys want to serve Bitcoin? was like,
We’ll it on the second round, you know? We don’t want to do it right away, but we’ll catch it on the second round. We were still a small shop and money, like that’s a big position to put on. And so we passed on the first couple and we started to educate ourselves just through more dialogue. And through education, we learned about how to de-risk these and the appropriate ways to de-risk these. And we became really disciplined and okay, these are our risk parameters for doing a deal. And if a deal will do it under these parameters, we’ll take it. But we’re not willing to go beyond our risk parameters because
This is a big project and one thing going wrong could be chaos. And so we got really disciplined and being disciplined and also really good at the boutique structures allowed us to get into some deals that were so nuanced that we were the only ones who could possibly structure and put it together. And so as a result, our risk conditions could be met in this complex structure. And we started papering transactions that required, you know, seven, eight different documents to be transacted at the same time.
in order to make the power flow. And so after we did a couple of those, people knew that we had this reputation for being the shop that will kill a lot of brain cells, spend a lot of time, and think really creatively about how to get a solution. Because this all goes back to, OK, the goal is to get our customer a really good customer experience, but not expose ourselves to risk. And so we migrated into this large load space by way of these Bitcoin mining data centers, which arguably felt the riskiest. And on top of that,
Maura Yates (15:49.71)
the market conditions at the time. This is back in 2023 when we had summer 23 pricing and summer 22 gas. And so the risk was just a multiple. And so the tools that we used to come at these creatively to manage risk and meet the needs of the market, which were they didn’t have much cash, right? Like they were cash constrained. So how do we manage risk on a low cash portfolio and not make any compromises that we need from a risk portfolio? And that’s what led us to really
start forging into the space and really get a reputation in the data center space.
Joshua Rhodes (16:24.77)
Right now, if you look at the aircott space, I mean, you get two people like us in the same room or on a panel or a podcast for that matter. We can’t talk about energy without talking about data centers, right? And that’s just like how everything is. Like, can you speak to what the structure of a deal means for a data center? Like, I don’t know if someone came to you now. I mean, they’re probably late in the game now, but like the data center comes to you now. Like, what does a deal look like to them?
Maura Yates (16:50.008)
Good question. Where my head starts going is pick your flavor, right? So as much as we want these data center deals to be standardized, because you’re like, everybody’s just building a data center, right? They’re all just like, it should be really simple. They’re all just, you know, trying to compute. It becomes really nuanced and it becomes really different because who sits behind that developer, not only as tenant for the data center and whose machines will be in there, who sits behind them as a lender comes into play.
So you have a of parties with really nuanced needs. So certain lenders like to see something a certain way in a contract. Sometimes the relationship between the developer and the tenant and the actual occupant of it are really different. And so traditionally what will happen is they’ll come to us, they’ll give us an understanding of who their tenant is. And more importantly, they tell us how they’d like to provide power to that tenant. That’s where we start. What do you guys want to do? We don’t start by telling them this is what we’re going to do.
we want back to the customer experience. Tell us what you guys want to do. Let us figure out how to do what you want to do because that’s going to make you happiest and that’s important to us. So tell us what you want to do. We’ll figure out how to do it within the means that we have and deliver on that. That’s what we’ve really been focusing on. When I think about the number of contract templates we have, so this will blow your mind. The last time I calculated, we had over 39 templates of data center contracts. Because one, there’s an evolution and as you keep learning, you learn little things.
it would do better. But more importantly, some are co-located behind the meter, some are in front of the meter, some share part of the behind the meter. Some are behind a POI and have two data halls. Some have two tents behind a single POI. Some want to curtail, some don’t want to curtail. You know, so everybody’s got these different requests and all these different requests impact different parts of risk. And so the elements that we have to put in to control those are very different across our contract. So when we think about one of the areas that
really distinguish mothership and our retail side of the business, we absolutely call out our data center contracting services as one of the things that distinguish us. We’re fast, we have really good templates, and we have really good templates that work for customers, because we’ve worked with most of y’all. We know what you like to see. We know if your lender wants certain step-in rights, etc. So a really solid set of templates to work from, though none of them are totally the same.
Joshua Rhodes (19:13.986)
When you look at the market structure as it is and your 39 templates and all the things you need to do, how could the system be doing it better? What would make your job easier or make you only need, I don’t know, say 17 templates instead of 39?
Maura Yates (19:26.924)
That’s a fun question to answer because if you ask that they’d say, Mara doesn’t want to simplify this. She loves the chaos. There’s a part of us that loves the chaos, right? There’s ways to figure out how to do things better and constantly doing things better. Now there are things that can be done that don’t need to add to the chaos. When I think about like where innovation needs to happen in the market, we frequently look to technology for innovation.
I think I get asked this question a lot and I think my answer is always the same. I don’t think it’s technology. We have wonderful technology. We have a lot of people working on promoting really good news. You guys had the hardware software side covered. What we need to evolve and innovate on is process. What’s killing these projects is the time to contract and the time to power. If you ask any of these large load developers, what’s killing them is that they’re just waiting to interconnect. There’s no technology that you can put out there that necessarily changes it.
What needs to change is how we process things, how we move through things. It’s very inefficient. Now, I say that, but at the same time, I don’t necessarily have a great solution because we’re in this place where our market’s undergoing a massive shift and they don’t want to knock Urquhart because Urquhart’s working really hard amongst a bunch of moving pieces and it’s really hard to keep organized. So I certainly don’t envy them and I think they’re doing the best that they can with a really large project. But that’s it.
key piece of this is it’s creating, it’s innovating in the space of process, because process is time, and that’s what’s keeping us from deploying these data centers more rapidly.
Joshua Rhodes (20:59.394)
Yeah, as we’re recording this podcast, just released a podcast talking to Pablo Vegas, CEO of Arqat, where he talked about different types of tools and he was making a case. We really need tools, even if they’re not perfect. We’ve got to get some stuff out there and just learn from them. Even if it’s not the greatest tool, you can hit a nail with a wrench. It’s not going to work great. And you can figure out how to make that wrench better and turn it into a hammer to actually hit the nail. He didn’t say that. I just said that. But anyways.
Maura Yates (21:23.534)
Yeah.
Joshua Rhodes (21:24.462)
Yeah, iterate and such. So one of the things like last week I was at a grid lab event in California and one of the things that I learned about ERCOT that I didn’t really understand and you’ve mentioned this kind of before in terms of what you mentioned, similar to something like this. with large loads that want to bring generation as well, whether that’s batteries or various types of behind the meter generation, it’s my understanding that the current way things are structured is you can’t net on your point of interconnection.
And so like if you’re like a 500 megawatt data center, but you have a 300 megawatt battery or 300 generation, you can’t net and say, want just 200 necessarily from the grid, or you can’t operate that away. Is that the case? Am I getting that wrong? Did I misunderstand that? Or like, how does that work with large loads kind of bringing their own generation? Cause a lot of, a lot of folks are talking about that these days.
Maura Yates (22:15.478)
Yep. So a question if you got something wrong, amidst the constantly changing updates, comments, depends on the point in time. Yeah, it’s very hard to keep straight. And by the time that this podcast airs, everything we’re about to say will already be updated because they’re having a workshop tomorrow. So whatever we say now might have changed. The bring your own generation, the net metering, the co-location.
Joshua Rhodes (22:24.462)
That makes me feel better, thank you.
Maura Yates (22:41.908)
This is a really, in my eyes, this is the next big topic that’s being discussed as part of the SB6 proceedings. The conversation that’s sucking the air out of the room right now is the BAT0 process, which is the PRR 145 to change the interconnection process. That’s kind of getting into a spot where we see that the end of the road or a light at the end of the tunnel. Okay. The next thing that the commission and PUC and ERCOT are starting to take up are projects
58, I’m gonna maybe get this wrong, 482. The demand side management. In the last couple of meetings and workshops, they’ve already started signaling, hey, we need to have this conversation. It’s where the curtailable load, the CLR conversations coming in. It’s where we expect conversations around voluntary early curtailment loads to come in. It’s the demand management piece of Senate Bill 6.
That’s just getting kicked off. So we would fully anticipate more conversation around net metering of these resources going forward. Now, that said, there’s a couple pretty important kind of rulings, or I should say one important milestone really just was promulgated and that’s specifically around the Good Night Project with CRUSO out in West Texas. And that’s a co-located facility with thermal and renewables. And.
It’s interesting because it’s perhaps setting a precedent for the conversations in that 5842. And as a result, what came out of that CRUSO filing were some pretty interesting comments. One, it stated that that load, that data center, needs to be able to fully shed and come off. And all of the native generation, the wind and the thermal, need to be available to push to the grid in their full capacity. Now that’s...
a little bit contested because the spirit of SB6 was really, think the understanding is maybe it was more designed towards dispatchable generation, not necessarily renewables. So I think there’s still conversation going on about that and that will likely be picked up in that demand management conversation as well. But that was one of the first things that came out of that net metering docket. And the second most important piece is that they told the loads, you’ve got to be available to curtail in an emergency.
Maura Yates (25:00.232)
and you can’t participate in any demand response or answer service programs. So it seems like it is setting the precedent for the things that will be decided around what the large load requirements are, agnostic of whether or not you’re co-located. But I think at the end of the day, we are going to continue to see a huge drive and push for co-location and behind the meter, simply because the perceived bottlenecks there are less.
Joshua Rhodes (25:26.488)
Got it. think the thing I was trying to say earlier is one of the things I brought up when I was at that meeting was there was a couple of hyperscalers in the room and we were talking about like, well, it’s like, can’t it just be a CLR, a controllable load resource? That’s where it was coming in. It’s like, well, they won’t let us net at the point of interjection, which you’re talking about. But of course that may change, I don’t know, tomorrow. Like as you pointed out.
Maura Yates (25:49.068)
And it gets tricky too, because some of the ways that loads apply in these co-located applications are they apply to be like a load only resource so that they aren’t netted, some want to be netted, some can’t be netted, some are looking for exclusion. So it’s a tricky space, but a pathway to netting, there will still be pathways to netting in several of these interconnections, but there’s also pathways for non-netted resources.
Joshua Rhodes (26:13.912)
The deals you’re looking at in the various contracts, like, do these data centers actually want to be power companies? I get the feeling that they don’t necessarily want to. They’re kind of being forced to just given like maybe how slow the rest of the system is running.
Maura Yates (26:29.23)
So I think the answer would differ based on the hyperscaler you’re talking to. There are some hyperscalers that currently have incredibly built out power desks and power teams. However, I will flag not super heavy in the wholesale market transaction. They’ll sign a lot of bilateral VPPAs per se, but not necessarily papering a lot of the short-term hedges or cash positions and not as active as a true power marketer would be.
they’re set up and I think those intend to take a more aggressive role as kind of like a power marketing participant. However, there are others and sometimes it kind of blows our mind because it’s your number one expense outside of your hardware is electricity. So you guys should have this really dialed and do you guys understand everything that’s going on here? And I think several data center companies are data center people, not power people. And so they kind of take what is given and
take what they hear everybody else has and that’s acceptable, rather than really pressing for ways to refine terms.
Joshua Rhodes (27:34.338)
Yes, one of the conversations I’ve had is, you know, these data center companies, do seem wanting to move so fast and maybe have buckets of money that they’re willing to throw at this. But I do think eventually, where you bring some real value in this, they’re going to get to the point where they’re going to have to be smart about electricity. I mean, right now, speed to power, right, just got to build data centers. But at some point, there’s going to be a competition for how cheap can I make a token for the next word in
when my students are cheating on their homework or whatever it is, know, or like, you know, writing code, they’re going to have to bring those unit costs down. So they’re eventually going to have to get smart on electricity and how they structure these deals, right?
Maura Yates (28:12.942)
They are. And what happens is sometimes it’s something bad that prompts them to get smart. I remember back in summer of 23, there was a really large publicly traded data center company that just got raked across the coals by their shareholders for their price of power. And when the market, they had done a poor hedging strategy. I don’t think they were well informed on what they should have done or what they advocated for in their contract. And as a result, they were not
position to take advantage of the economics in the market. And so in a market where, let’s say, real time was clearing 70 blocks, they put on $120 hedge and that showed up to the street. And so they need to get more sophisticated and the challenge is sometimes they try to get more sophisticated, but not with the right tools. And so what we really focus on, I’ll keep saying it, but back to our original mission, which is customer experience.
We focused originally on Resy, now we focus on our large loads customer experience. And getting them data and teaching them how to use the data is one of our biggest drivers. So as a result of serving these large loads who are sophisticated, we have some that are very sophisticated and some that are very hands off. We treat them all the same. We treat them all at the same level of sophistication. And so for these large loads that are super sophisticated that need to provide very precise sub billing allocation, cost allocation down to
different tenants, getting them data is huge. We launched a new platform internally at Mothership called Eclipse. We’ve now taken externally because it was our market access platform. We wanted to make sure that our data centers saw the market the same way we see it. Because if we see something that they need to take action on, we want them to see it too. For the example you noted, a lot of data centers right now want to run on index. And I get it, you see me roll my eyes on that because it’s
Works until it doesn’t work, right? Index is good until it blows up. There’s always a right price to hedge and there’s always a right way to liquidate hedges. Some get that. Right now, most of them are writing index. But as we’ve experienced before, the market changes. When the market changes, all those guys want to hedge. We give them the visibility to understand when that market is changing every day. Every hour we give them this visibility by logging in. And they’re able to see, okay, now is the time I should take action to hedge because
Maura Yates (30:35.596)
I’m seeing this forward curve just crawl up, which is giving me an indication that real time and my cash position is probably going to have the same increase. And so it gives them the idea of like, now’s a good time to hedge. And more importantly, maybe I’ll just hedge these hours. Maybe I’ll hedge the nighttime. Maybe I’ll hedge the daytime off peak. it gives them the ability to make educated decisions to improve their company economics. beyond that though, this again, all goes back to the customer experience.
If you give them a bunch of data, but they don’t know what to do with it, then that doesn’t matter. And digesting this data is super important. So we’ve taken Eclipse one step further to digest this data and actually make the recommendations and say, hey, hedge now, hedge this price, or hey, wait, this is what this would synthetically look like in your portfolio. And more importantly, let me pull all of your historics, give you all this detailed level clearing information on your historics to say, maybe we should, maybe we shouldn’t.
big thing there is data to give them the sophistication that they lack because they’re not power people, they’re data center people. But they need to be power people, so we need to give them that access. And if you think about that same conversation, that same conversation we had with the DER player saying, you need to be in this space, but you’re really not equipped to be in this space, so let us enable you and let us teach you how to be in this space. And that’s what we do now for our data centers and large loads.
Joshua Rhodes (32:01.516)
Yeah, I’m glad you went there. That’s exactly where I was going to go next is you’ve talked about, you know, delivering data and delivering tools. You built a tool and I believe you told me that like, so you built Eclipse originally for in-house use, right? It was just going to be mothership was going to be using it. But then you were using it so much or sending screenshots to people or something like that. like, we want this. So I just went to the website just a little bit before. It looks like anyone can sign up for free right now if they want to check it out. So there’s like a trial period going on. And so we’ll put a link to it in the show notes.
But yeah, so you talked a little bit about kind of what it does, but can you give me any specific use case? How would someone come to the conclusion that I need to hedge? Like what would they be seen on your tool right now if they looked that would give them an indication of that.
Maura Yates (32:47.096)
Yep, so to level set our tool, like if you picture in your head, there’s a sidebar and on this sidebar, there’s a bunch of different modules and these modules are different ways that mothership operates. So as you mentioned, Josh, we built this not for anybody else. We built it for ourselves. We built it because this is how we needed to see the market, how we needed to ingest data, how we needed to interpret it. And we kept screen sharing it with all of our data center clients saying like, hey, look how your load settled or hey, like look at this.
and showing them what we saw. Showing them, this is the read we got from the TDSP. This doesn’t align with your telemetry. Let’s go see if there was a meter read error. So we start showing them and all of a sudden they’re like, you spend most of our time now showing, we should just get you a login because it’s going to be way more efficient. Let’s teach you to fish rather than keep feeding you. So we decided to turn it into an external facing platform where we started by just putting our data center customers in there. It’s where they could retrieve invoices.
retrieve their interval level report. So we give all of our customers interval level billing data with a massive spreadsheet behind it. Knock on wood, I’m going to get challenged by some other rep when I say this, but we’re pretty sure it’s the most in-depth interval report because it’s really complex math to pull together. But that all sits in this platform and they’re able to take that interval level report and push it up against our other unit, which is our forward curves unit and our settlements unit. And one,
They can shadow settle themselves and say, did they get built everything correctly? Hey, where did I incur this charge? Hey, does this look like what my telemetry looked like? They can do a bunch of validation themselves, which typically that data for validation is totally gate-kept by an REP. That’s REP data. That’s not data that the load typically gets because that’s part of a transaction statement that we get from a TND. But we push it all out there because, okay, you can see your load data. You should know what you consumed. So we put it all out there.
They go look at their historic position. They can shadow settle it. Or more importantly, the question you immediately asked. We’re talking about your open position and we’re trying to figure out should you hedge? And a lot of these large new data centers are coming with this demand for a physical tie to power. So that’s why they’re going to behind the meter. But for those that can’t go behind the meter, they’re wanting asset specific physical hedges to say, I’m buying, I know you exist. You’re selling that to me for 15 years. If I can’t locate next to you.
Maura Yates (35:14.754)
I at least want to make sure that I have somebody whose blades are going to turn. And so as part of that transaction, we will tell them to like, we’ll go in and say, okay, if this is the hedge that we want to do, let’s go price it on the forward curves. So we’ll go into our Eclipse platform. One of the free units that you mentioned right now is our forward curves unit. These are the forward curves we use every day. This is our exact pricing platform for forward curves. It has ancillary services. It has every ancillary. It has basis hub for all the load zones.
It’s got more pricing curves than you would get through any other kind of publicly available curve tool because it’s our curves. It’s literally what we price all of our deals on. So we say, hey, come in here and let’s go price up what this hedge should be. I check my curves every day in market. I know our curves are good. So this is what the curves are showing as the hedge price. Then let’s go shop this and let’s go see if we get this as the hedge price because if somebody comes back two, three, four, $5 higher than this, they’re out of market. Don’t hit that. But it gives you the ability to gauge.
when you should hit, whether you’re getting close to your price target, and if the numbers that you’re getting are good. Because if you think about it historically, this is a really big piece of why we made this public. You guys have, by saying you guys, the data centers and the generator, like the renewable generators, you guys have been price takers. You’ve been price takers because you haven’t had the visibility into the wholesale market to know how to price your asset on these wholesale desks. So I remember back in the day when I was at Sun Edison and we were going into one of
It doesn’t exist anymore, but one of the very large utilities and going into the generation arm and saying, hey, we want to sell you our PPA down in Southwest Texas. And we didn’t say we’re going to sell it to you for this price. We said, what would you pay us? And like how weak to have to go in and say, well, what would you pay us? Right? Like I want to go in and at least know my value and say, I know that I’m worth at least this much. And so that’s rampant across the industry is.
Solar developers, even it’s improved today, but I’d still argue solar dev shops do not have price on real time pricing. Pricing moves a seven by 24 and even an on peak product for a five year term that can move two to $3 overnight. And if you’re a solar developer, you’re not capturing that two to $3 price movement unless you know what’s happened. So our forward curves tool gives developers the ability to have power and some control in these.
Maura Yates (37:40.334)
pricing scenarios and not just be price takers. So it’s a really good way for them to optimize that procurement as well.
Joshua Rhodes (37:46.756)
Sounds like it unless you come to the table with a much stronger bargaining position than you had before,
Maura Yates (37:51.438)
Yeah, and all the tables you’re probably going to are tables that we talk and trade with. So it’s very easy for us to call BS if they show a number that’s out of the market. We’re like, that’s not right. We know you just hit something at this price point. So definitely empower.
Joshua Rhodes (38:07.0)
Totally. I know combined we’re probably at the 40 minute time for the three chunks or whatever, but I do want to ask a final question. You’ve seen this market move and right now it’s in a, we’ll say turmoil, but it’s like, it’s bubbly right now. There’s a whole bunch of things going on. You’ve built a tool that allows you to make a little bit more sense of kind of maybe what’s going on. But as you look into the future, like what technology excites you right now? Like, what do you think when we get beyond this kind of place where we’re at now?
What are you dreaming about for version two of this?
Maura Yates (38:37.452)
Man, it’s so funny. I was telling somebody, it’s hard to turn on my creative brain outside of the creative immediate world that we’re in. When I think about the grid in the future, I absolutely still am of the hard belief that we repeat history and we are moving away from centralized grid to a decentralized grid. And I think when we originally thought and expected this to be happening in higher volume, this was five years ago, what we described as distributed energy were these like tiny 10 megawatt or less like DER projects.
I think we’re talking about it now like distributed at the transmission level too, where it’s bring your own generation. We’re back to a different era of power plants and supply, right? We are a place where we’re have to have way more localized supply. There are parts where it’s probably gonna get more expensive for that need to bring it into more local areas. I think we’re gonna be in a world where I do think ERCOT’s gonna figure out the large load issue and they’re gonna do it in a way that doesn’t add liability concerns. That’s their number one concern.
They’re going to figure that out. They’re not going to put all the other rate payers at risk. So I think we’ll figure that out. I think a lot of vols going to exit the market. So we’re going to get a lot of batteries that are deployed, batteries that are going to come. And when they have to curtail, it’s going to allow them to shift over to this four gigawatt hour battery and not impact the curtailment. Because ERCOT’s going to say you have to curtail. And so data centers are going to be left with no other option than go install your battery. Go install your on-site gen.
So we’ll see a lot more battery. I think it’ll be more strategic. I think the services might still be called ancillary services of some sort, but I think they’ll look a little bit different. I think some of those will be mandatory. And I think maybe we’ll start getting some more back to like smaller data centers, like localized data centers. But here’s the reality. We are a society that uses data and compute everything. So whether the AI bubble busts or not, data, we just keep.
driving more data. So a data center, I don’t care if it’s AI, it’s going to have some occupancy as long as we continue to consume data. So I see the data center world growing as well, perhaps more efficiently, but still growing too. So I don’t know if I avoided your question or if I gave enough of a look at like the imagination. It’s a little more of the same, probably moving more in that distributed, more distributed pathway, especially as we bring the BYOG becomes more of a prominent requirement. So.
Joshua Rhodes (41:03.084)
No, absolutely. mean, I’ve heard people say all kinds of like things like peak oil, maybe even peak gas, but I’ve never heard anyone say peak electricity. Like you said, I think we’re going to continue to use more data. It’s going to be electrified. It’s a great place to stop. Yeah. Mari Yates, thanks for coming on the Energy Capital Podcast.
Maura Yates (41:17.752)
Thanks for having me. Appreciate it. Thank you.
Joshua Rhodes (41:20.269)
Absolutely.
Thanks for listening to the Energy Capital Podcast. If today’s conversation helped you make better sense of how the energy system actually works, share the episode with a colleague and hit follow on your podcast app. You can find us on Apple Podcasts, Spotify, and all the usual platforms. For deeper analysis and context each week, subscribe to the Texas Energy and Power at texasenergyempower.com. That’s where you’ll find every episode, every article, and our latest updates. We’re also on LinkedIn, X, and YouTube.
where we share clips, insights, and ongoing commentary on energy policy, markets, and the grid. Before we go, a quick note. The views expressed on this podcast are my own and do not represent the official positions of the University of Texas, Ideasmiss, Austin Energy, or Columbia University. A big thanks to Nate Peevee, our producer. I’m Joshua Rhodes. Thanks for listening, and we’ll see you next time.
