Here’s something surprising: in states like North Dakota and Texas, the surge of new industrial and data center load has actually moderated electricity prices.
The very thing many people blame for higher power bills has, in some cases, had the opposite effect. According to a new report from Lawrence Berkeley National Lab, load growth has slightly lowered retail electricity prices on average over the past five years.
So what’s really driving them up? The answer isn’t renewables or AI. The study finds that generation costs are down 35% since 2005, but transmission costs have tripled and distribution costs have more than doubled. Billions are now being spent to upgrade the grid and harden it against extreme weather.
This week, we’re joined by guest co-host Caroline Golin to unpack the new data. We’ll discuss what’s driving those infrastructure costs, why utility spending remains so opaque, and what could happen over the next five years as large loads multiply.
Later in the show, we’ll talk about a new proposal from Energy Secretary Chris Wright to accelerate the interconnection of large loads with onsite generation or flexibility capabilities. The proposal could speed up data center projects, but also risks triggering a new clash between federal and state regulators over reliability, costs, and control.
Credits: Co-hosted by Stephen Lacey, Jigar Shah, and Katherine Hamilton. Produced and edited by Stephen Lacey. Original music and engineering by Sean Marquand.
Open Circuit is brought to you by Natural Power. Natural Power specializes in renewable energy consulting and engineering, supporting wind, solar, and battery storage projects from concept through financing. Discover how we’re creating a world powered by renewable energy at naturalpower.com.
With resilience now a leading driver of grid investments, Latitude Media and The Ad Hoc Group are hosting the Power Resilience Forum in Houston, Texas on January 21-23, 2026. Utilities, regulators, innovators, and investors will all be in the room — talking about how to keep the grid running in this new era of heatwaves, wildfires, and storms. Register today here!
Transcript
Stephen Lacey: You’re in Boston on Halloween week. You know what that means, right? You got to go to Salem.
Caroline Golin: Well, I’ve been to Salem on Halloween week actually already. I’ve done that before. I’ve done everything. My husband believes that probably in a previous life I was like a Scottish witch or something.
Jigar Shah: So you know what the story of witches are that I just learned, is that witches… So women were the original brewers of beer and of spirits. And so the cauldron, all of that stuff, came from their brewing experience.
Caroline Golin: We’ve been ahead of our species for some time.
Stephen Lacey: Yes.
Jigar Shah: I mean, that goes without saying.
Stephen Lacey: But the Salem Witch Museum is so bad. Have you actually been?
Caroline Golin: It’s so bad. I was preparing myself, my sister. So we actually, my dad’s birthday, and he passed away, we came up here and did his memorial for it. And we were so excited. My sister and I were like, “Yes, we’ll finally do it.” We walk in, we’re like, “What the…”
Stephen Lacey: And it hasn’t changed in 30 years. Like, I went when I was a little kid.
Caroline Golin: That seems like a project for you.
Stephen Lacey: Absolutely. Yes.
Jigar Shah: You need to adopt a museum, Stephen.
Caroline Golin: I mean, podcasting may work out, but I think animatron witches are the future.
Jigar Shah: Salem Witch Museums are forever, Stephen.
Stephen Lacey: Are you guys dressing up this week?
Caroline Golin: Oh, yeah. Our family is big time.
Stephen Lacey: My kid asked to go as a vampire, and I told her that’s redundant. We’re already getting drained by the power company.
Caroline Golin: Oh, Stephen.
Stephen Lacey: From Latitude Media, this is Open Circuit. Pop quiz. Which one of these drove up electricity prices more over the last five years? Renewable energy mandates or data centers? If you picked either one of them, you’re wrong. And you’re in good company because that’s what so many politicians and pundits believe too. Lawrence Berkeley lab just released a detailed study of what’s actually driving electricity prices in every state. And surprise, surprise, the results don’t map to any clean partisan view of the world. States with the most renewables often have the cheapest power, data center growth actually lowered prices. Meanwhile, the real culprits, infrastructure replacement, extreme weather, and natural gas volatility barely come up in the conversation. So this week we’re digging into the numbers and most importantly, with demand on the upswing, do they tell us anything about what’s ahead? Welcome to the show. I am executive editor of Latitude Media, Stephen Lacey, and I am here with Jigar Shah who’s an investor and the former director of DOE’s Loan Programs office. Hello, Jigar.
Jigar Shah: Hello.
Stephen Lacey: I realize it’s been like nine months since you probably stepped foot in your office at the Department of Energy. What do you miss most?
Jigar Shah: Oh, nothing. I mean, I miss all the people. I miss all of the extraordinary people and all the work that we were doing. But I definitely don’t miss Forrestal. I mean, it’s called brutalist architecture for a reason.
Stephen Lacey: I go back and forth between loving it and hating it.
Jigar Shah: No, no, no. I did a lot of work to put a survey together. And it was like 91% of the people surveyed said that they would like the building to be torn down and started over. And so, of the people who worked at DOE, like there was only a very few, group of nostalgic folks, who want to keep it.
Stephen Lacey: Katherine is out this week and in her place is Dr. Caroline Golin, the CEO of Envision Energy Advisors, and up until this summer was the global head of energy market development at Google. Caroline, welcome.
Caroline Golin: Thank you for having me. I’m so excited.
Stephen Lacey: Good to see you. Speaking of missing offices, do you miss the nap pods and gourmet snacks at Google’s offices?
Caroline Golin: Oh, for sure. Yeah. I mean, it’s amazing how many of my kids’ lunches were subsidized by the Google mini bars. It’s great.
Stephen Lacey: What have you been focusing on the last few months?
Caroline Golin: Oh, a lot of things. Working with some suppliers to help them transition into what is going to be a new era of powered land over renewable PPAs and working with some utilities to think through planning processes and some developers and startups. It’s a really interesting space. And I am really enjoying being on this side of the aisle. When you’re at Google, everything comes to you somewhat buttoned up, with this great vision of how they’re going to solve all your problems. And now I’m on the side of people scratching their heads going, “We don’t even know what the problems are yet. Can you help us define them?” So it’s been really exciting and I’ve been enjoying it. Yeah.
Stephen Lacey: So you all might remember Caroline from our live show at Transition-AI. And she’s here to help us understand the dynamics in the electricity market. So if you remember a few weeks ago, we also had Charles Hua, the CEO of PowerLines, on the show to talk about the impact on politics and regulation on the rising cost of electricity. And prices are up, utilities are filing record rate cases. Everyone’s looking for someone to blame. And of course everyone wants a simple answer, but as we explained, it’s right now at least, largely transmission and distribution costs that are rising, not the fault of renewables or AI data centers.
And so we’ve got this new report now from the Lawrence Berkeley National Lab that goes deeper. And it looks at what caused power prices to rise over the last five years. And it concludes that it’s volatile gas prices, the soaring cost of poles and wires, and billions in spending to harden the grid against extreme weather. Those are the main culprits. It also shows that net metering in renewables built to satisfy state targets also played a modest role, and we’ll talk about that. But with the help of market-based renewables, generation costs are actually down 35% since 2005. But transmission costs have tripled, distribution costs have more than doubled, and utilities are pouring a ton of money into wildfire mitigation, storm recovery, and replacing aging equipment. So first I want to get your reaction. I just gave my interpretation. Caroline, what was your read?
Caroline Golin: I mean, generally I think the story isn’t new, right? I think that 10, 15 years ago, utilities started recognizing that load growth was flat. They also started recognizing that there were these creeping disruptive technologies like rooftop solar and electric vehicles coming into bear. And they needed to invest in what was a naturally aging out system. And you had the confluence of extreme weather. So you had both a proactive approach, which was saying, “Hey, our system needs to be upgraded.” And you also had this reactive sort of fear-based approach, which was that we have no way to spend capital anymore. And therefore we have no way to earn a return. And those two collided at the same time. And so for the past 10 years, you’ve seen multiple, what we call grid modernization dockets. You talked about this in the podcast previously, New York REV, we had it throughout… With Massachusetts, down in North Carolina. And they were billion, billion, billion, multi-billion dollar cases.
But the problem was the cost-benefit analysis was just a few points increase in reliability. Which generally resulted in some homes having power for an hour or two more each year in exchange for their costs going up by 20%. And they were approved. And they were approved because largely the interveners across the country didn’t have a framework for how to assess this and push back on it. I think that’s evolved over the last couple of years. But they were approved and then they went into rate base and that’s what you saw happen. And I think it’s going to continue to happen until regulators and intervening parties, environmentalists, customer advocates, get a framework to push back on it. On the renewable side, I think we forget that 80% of all renewables deployed, wholesale markets, were deployed by corporates. So they’re out there trying to hedge against gas prices.
They’re out there trying to get the least cost [inaudible 00:08:12] price they possibly can. That’s a very different model than utilities and states that are just required to fill a certain volume. And so you see an incentive structure. So I think it all pencils to me, it all makes sense to me. I think the other part of this that we may talk about, we haven’t talked about before, is that a lot of that cost got distributed across rising data center load. So the states that came in that had massive data center build out, they all of a sudden had new load to distribute these billion dollar grid modernization costs. The states that didn’t have that, didn’t have that. And that’s where you saw the prices go up. Whether that’s the case moving forward, I think it’s going to depend on how we decide to build out the grid and what we look at in terms of meeting load growth.
But I think the story has always been there. I don’t think it’s a new story. It’s been the story for the past 10 years. It just hasn’t been the one that’s nice to focus on because it doesn’t involve a ribbon cutting or such. But it’s been going on for years. I mean, I did these grid modernization cases 10 years ago. I sat there yesterday with Mark Christie on a panel and we joked about the fact that the first time I sat in front of him in Virginia was to beat up against a six billion Dominion grid modernization docket. And he didn’t give Dominion the money because they didn’t do the cost benefit analysis they needed. They ended up doing it and getting the money and then data centers came in and took up all that cost increase. And that’s how the story went. But that’s sort of what’s happening and I think it has been happening for quite some time.
Jigar Shah: Yeah. I mean, I thought your summary was good. Which is that there was no storyline. That ultimately the only storyline which I think LBL failed to capture is that the utilities continue to be a place where innovation goes to die. And so when you look at Tom Siebel and his work with C3, I mean, Exelon put $20 million into that company. And the whole point of it was you have a certain number of maintenance people and you have a certain number of things. How about we use data to make them maintain the stuff that needs maintenance as opposed to just randomly checking on stuff that’s fine. And after six years of failing miserably, Tom Siebel was like, “Why am I doing this? I should just sell to the oil and gas industry.” And so he did. And the same thing’s true with Microsoft with their AI stuff.
They’re like, “Why are we continuing to try to sell to the utilities when they don’t want to use AI and they don’t want to use next-generation technology? They want to do other stuff.” They’re like, “We should just help the oil and gas industry get more oil out of the ground.” And so when you look at batteries or dynamic line ratings or alternative conductors or whatever else, magically they’ve been fully deployed in the UK and Belgium and Brazil and India. And we’re like, “Are they ready for prime time? Maybe we should do a 20-year deployment of dynamic line ratings in this country.” And EPRI is like, “I don’t know, we might need to do 12 more tests before we’re ready to deploy this stuff.” And so, yes, Caroline’s completely right that we, basically we’re completely behind on investing in our infrastructure in the ’80s and ’90s. The utilities were like, “Oh, we should make up for lost time.”
And so this time when natural gas prices went down in 2010 is a good time to not give those savings to ratepayers, but instead raise our investment in the wires. So that rates will go up only a little bit, right? Fine. But then they failed to use any new technology when they were doing it. So they were like, “We invented all sorts of stuff.” That was CleanTech 1.0 or 2.0. Those companies all went bankrupt and went to Europe and then all that technology went there or to China. And we’re still in the same mess.
Even with this huge power issue, one of the things that the report highlights is one of the reasons why rates have gone up so much is that everything’s gotten expensive. Transformers, this, that, whatever. Why? Because it’s all traditional utility stuff. So the stuff that the utilities like to deploy, they have bought too much of. And instead of saying, “Huh, is there something that we could possibly do that’s 90% cheaper to do the same thing? That would put less stress on the supply chain?” They’re like, “But we don’t like that. We’re where good dreams go to die.” And so as a result, there still is no movement on the utility side to be like, “We should deploy innovative technologies that we innovated 20 years ago and piloted for 20 years at scale.”
Caroline Golin: And also I would add to that, the regulators are using two different yardsticks to measure investment. So if you’re measuring investment on the generation side, your yardstick is going to be, does it meet load growth? Does it meet peak demand? And if you’re measuring investments on the distribution or transmission side, it’s going to be, does it hit some level of SAIDI or SAIFI scores. Some level of reliability. But the two paths shall never cross. And if you treat it differently, you’re going to invest differently and you’re going to over-invest, right? And underdeliver. Which is what’s happening.
Jigar Shah: And the other piece of exactly what you just said, Caroline, is that on net metering, for instance, we have had study after study that shows that when you reach 1% penetration of net metering, then you’re no longer guaranteed that these systems actually reduce costs for the rest of your neighbors. And that as you go to 5%, it depends on the circuit, et cetera. Once you go beyond 5% penetration, then you are like most certainly costing your neighbors money. The only way to not cost your neighbors money is to force every one of those systems to put in batteries and then to have the utilities flex those batteries. And so now the question is whose fault is it that net metering raised costs in California? California has hit 10% of households with net metering. We all predicted that it would have an impact once you went above 1% and most certainly when you went up against 5%.
That’s clearly the regulators’ fault. Right? They should have mandated that the utilities integrate net metering systems into distribution circuit planning. Have they done that? No. And in fact, even today California is like, “We don’t want to use your behind-the-meter batteries to manage these distribution circuits.” And so NEM 3.0 requires you to put in a battery, because you won’t get any benefit out of it otherwise, but you don’t use it. And so part of my problem with this stuff is that I think LBNL did a fantastic job of summarizing all of the data, but I don’t think they actually did a great job of revealing any insights and actually providing those insights into the executive summary, right? I mean, like-
Stephen Lacey: That’s what podcasters are for.
Jigar Shah: Sure. I guess so. We still have a job until AI replaces us too.
Caroline Golin: AI will never replace me. I’m far too particular.
Jigar Shah: It might replace me though, I think [inaudible 00:15:01].
Stephen Lacey: I mean, it’s really hard to know… I just want to talk about transparency for a minute. It can be really hard to know exactly what utilities are spending, what encompasses these T&D investments. Why is it so opaque to understand what exactly utilities are spending on, categorized as transmission and distribution investments?
Caroline Golin: Well, I think part of the problem generally is that the process that we’ve developed at the state level, so all distribution investments, whatever type of system you’re in, it’s going to happen at the state level. It’s going to happen with your public service commission. And what you see throughout the country is the parties that get involved at the public service commission get involved for largely two reasons. Three. One, you’re a large industrial group and you just want to keep your economic development benefits and you want to keep your cost of power low. You do not care how you keep your cost of power low. You just need to keep your cost of power low. Two, you’re an environmental group and you’ve been trained to argue to promote renewables at best and that’s it. Or three, you’re a customer advocate group and you’re in the same bucket as the industrial group with like, “I just want to keep costs low.”
Very few places across the country, and I think people like Ric O’Connell have been trying to change this for a very long time, have actually interjected the technical expertise of what grid modeling should look like. So there’s a process problem. I think the other problem is, I don’t think anyone understands how arcane the majority of the distribution system models are in this country. The vast majority of utilities don’t actually even know where all the homes and their houses are. So the idea that they’re going to model them with some optimization around reducing cost to customers is somewhat idealistic. And it’s not as if the regulators have that technical expertise. So you’re just sort of chasing your own tail in this situation. No one’s set up to do it. I don’t think the transparency is a product of hand waving. I don’t actually think the utilities are going into this space creating a black box saying, “You can’t see the cost of poles and wires. You can’t see this.”
I don’t actually see that happening. I think it’s a product of the utilities not actually knowing what the scenario analysis should be or how to place it correctly. So they just do the bare minimum and they run it sort of in circuits. And when you do it like that, I mean, we know who half the intervening parties are in this country. They’re not PhDs in electrical engineering that can sit through this work and truly understand circle optimization. It’s different than being able to go out and say, “I know what the cost of a megawatt hour is for solar and that’s cheaper than the cost of a megawatt hour is for gas. So therefore we should ram this through the integrated resource plan.” It’s just a much more technical exercise to which both the utilities and the intervening parties are not reaching the bar of what can and should be done.
Jigar Shah: Well, and the other thing that I find frustrating is that there’s no interest in fixing it. When you look at where we’ve come since 2005, remember, I think part of what Caroline just said is that the utilities by definition had no idea what was happening downstream of their distribution substation. Their data ends at the distribution substation. So they didn’t know how many homes were there. They didn’t really care. They didn’t know about any of this stuff. So then we said, “Well, why don’t we add smart meters?” And then we had the ARRA stimulus bill and all this other stuff. And we’re like, “We’ll pay for 50% of the cost and you pay for 50% of the cost.” And of course none of that data ever was collected and/or integrated into the modeling software that we’re talking about here or the system approach. So we paid for nothing.
So the utilities got all the benefits, which was they don’t have to send somebody out to read your meter manually anymore, and they got all those benefits for shareholders. I never got an app for my phone that shows me how I use the power, even though I was promised that power. And if I did get that app, well, now that data could be useful, right? In Texas, they did force the unlocking of that data, but around the country emission data has shown that the utilities have systematically made sure that that data never saw the light of day. And so there’s that. Then the question becomes what is the dynamic between the utility and the regulator? In the past, the dynamic was the utility spent money, the regulator looked over their shoulder, and they could say that these things that you did were denied access to rate basing because it was not prudent.
Today, that’s not really true. Today what happens is utilities deliberately, or not deliberately, depending on your point of view, overwhelm the regulator with all of these projects. They don’t have enough staff to be able to look at it. I talked to Google Tapestry a long time ago and said, “Why don’t you just offer that service for free just to analyze all of that stuff for the regulator?” So the regulator instead says, “Well, let me just take two of these projects and take a deep look at two of them because I can’t get to all 50 with the staff that I have and to figure out prudency.” And so now the dynamic between the utility and the regulator is basically the regulator pre-approves everything the utility put into that filing. And now the utility just sort of does stuff. And then when our friend Marissa tried to solve this problem in Connecticut, Avangrid went after her personally, to the point where she had to quit her job in Connecticut.
And then, now the utilities in New Jersey are going to the regulators at the BPU and threatening them. And saying, “I know that the governor has promised that we’re going to freeze rates, but if she does this, I’m going to go after you personally. And I’m going to do exactly to you what we did to Marissa.” So I don’t actually think that we’re in a healthy place right now. I think the utilities have gotten drunk on the fact that their stock price is way up. Higher than some of the tech companies. Stock prices have gone up and they like this situation. And they’re like, “We don’t want you to regulate us. We don’t want you to stop us. We don’t want to be forced to provide 20% discounts to electricity bills by 2030. We want to be able to do things exactly the way that we’re doing them now and not use any innovation.”
Caroline Golin: I actually think there are a number of utilities who recognize that the golden box that they’ve been given can be taken away very quickly in a highly populous country. And I do think that there are utilities out there saying, and not all of them, but I do think there are utilities out there saying, “If we were supported in this transition so that our stock price doesn’t go down, so that pension funds don’t suffer…” Which is another layer of this. If they were supported, they would support it. They would transition into a place of providing solutions. Because the technologies exist there and they have capital partners now.
The reality is there weren’t capital partners that were willing to see the grid transition as a long-term stable asset for investment. They exist now. The problem is I don’t think the regulators have any clue on how to make that transition. And it actually shouldn’t be the utilities determining how they receive increased basis points moving forward or decreased basis points moving forward based on the way that their goods and services are provided. And I’m not sure anyone’s given a model for how to do that correctly. So I think the utility culture is a bit split right now. I agree with Jigar that there are a number of utilities out there that are just happy to just lap this up for as long as they can, but the ones who see the need to transition are struggling to actually find a model that will simultaneously be accepted by the financial industry and by the political [inaudible 00:23:02].
Jigar Shah: But aren’t you just highlighting the problem here?
Caroline Golin: Yeah.
Jigar Shah: None of the utilities in Canada are for-profit utilities. They’re all basically versions of large municipal utilities. I mean, aren’t we just describing the fact that for-profit utilities are actually incapable of doing their job? The reason you give them all this money and all this freedom and $26 million pay packages that they do not deserve, and everyone knows they don’t deserve it, because they do the exact same job as the person who runs CPS Energy, who probably makes $500,000 a year, but whatever, is that you should be able to deploy technology at scale. And if you can’t deploy technology in scale and lead the entire industry and do all of that stuff, and you’re actually even worse than rural electric co-ops and munis at getting this stuff done, well, then the whole model is broken. The incentives are broken, all of it’s broken. That you really need to go to sort of a municipal-utility type function.
Caroline Golin: Well, I haven’t seen a ton of municipal utilities doing it much better. But I think that’s just because they’re risk averse.
Jigar Shah: Well, their rates are way lower.
Stephen Lacey: And is that also because, Jigar, that like the IOUs are disproportionately dealing with wildfire impacts and-
Jigar Shah: No. Of course not. Are you kidding me? There’s a lot of munis that are dealing with wildfire impacts and other things too. I mean, it’s a state thing. It’s not like an IOU thing. I just think that in general, the profit motive… And I’m not accusing anybody of wrongdoing. I’m not suggesting that they’re doing illegal things or anything like that. But I’m just saying that if you’re in New Jersey and you just got hit by Hurricane Sandy, right? You’re like, “This is an opportune time for us to run up the scoreboard and figure out a way to do all these things.” And then to Caroline’s point, a lot of the regulators don’t really have the knowledge and expertise necessary to know which exact line items were required and which weren’t required. And then they just mix it in.
California did the same thing. I mean, PG&E has been so far behind on the vast majority of its distribution circuits. They just reclassified a lot of their projects from distribution circuit upgrades to wildfire upgrades. And people were like, “Okay. That makes sense. This is a wildfire moment. Let’s do this, right?” Do I fault PG&E for that? Maybe. But they’re just doing what their incentives tell them to do.
Caroline Golin: I mean, my favorite was when tree trimming was considered grid modernization. That one was big for me.
Jigar Shah: The arborists.
Caroline Golin: The arborists. I was like-
Stephen Lacey: Well, that’s what I mean about a lot of this stuff is lumped together and it can be hard to… Yeah.
Caroline Golin: It’s just lumped. It’s just lumped and you can’t piece it out. And the other thing is, there are very few things you’re going to see in these packages that are a waste of time or a waste of money. They aren’t. They actually aren’t. They’re just not optimizing for where we want the grid to go. Is it important that we trim trees? Absolutely. If you’ve ever… I live in Atlanta.
Stephen Lacey: Indeed.
Caroline Golin: Trees everywhere. You have to trim trees. Is it important that we underground wires? Yeah, it is. So I think the conversation needs to move from, is this good or bad? Which is to say what are we trying to build to? If we’re trying to build the exact same grid we’ve had for the last 20 years, I think we’re missing what’s happening in society.
Stephen Lacey: I think that’s a really helpful dive into the complexity and opacity of the T&D investments. I want to go to two other storylines that you alluded to, Caroline. The impact of renewables and then the impact of large loads. So let’s just go to renewables first. The research showed that market-based renewables lowered prices while renewables built under state renewable energy mandates raised them slightly. Jigar, help people understand that dynamic.
Jigar Shah: Well, there’s two pieces to this. One is that I think we all know that we would never have had market-based renewables in states that didn’t have mandates unless someone went first and trained all the workers and figured out how to get the cost down and all those other things. So thank goodness for some of the states going first. And second, I think that the ongoing renewable portfolio standards are probably causing for-profit companies in the renewable energy space to take advantage of the mandates. And so we probably should phase them out.
Stephen Lacey: How would they take advantage?
Jigar Shah: For instance, if you look at the state of Colorado, they had these mandates on renewable energy. They brought in wind power that was $17 a megawatt hour from I think the Dakotas or Wyoming. And then Xcel was like, “Well, we can’t figure out how to integrate that into our transmission system.” And so then they had to rate base $3 billion worth of transmission upgrades for that wind power. In a market-based mechanism, they wouldn’t have gone forward with that project. Or they would’ve figured out a much cheaper way to integrate that wind power into their system by forcing those wind systems to put in batteries, probably at the site, and to regulate how that power went through.
But because it was a mandate, the governor was like, “We’re getting $17 per megawatt hour wind.” And Xcel was like, “Woo-hoo. We get to rate base $3 billion of unnecessary transmission upgrades.” And so everybody was like, “Oh, this is great. We’re meeting the RPS and we’re enriching our shareholders.” And I don’t think Xcel was a bad company to do that, but I just think that it’s probably not the way in which a market-based mechanism would’ve been integrated into the grid just because at some point they just would’ve not moved forward with the project if it wasn’t done intelligently.
Caroline Golin: And I wonder actually, how much of this is the cost of balancing? So I look at states with a 100% RPS. I look at a state like Nevada. And they’re running after the deployment of solar. They don’t have a regional transmission system where they can actually wheel that across multiple states and optimize it. And so they’re forced to balance it. And they balance it, maybe not with storage or flexibility or all the things we want to see on the grid, but they balance it with new natural gas. I mean, that’s what we see happen in SPP. We’ve seen that happen to some degree in MISO, but in all these other states that had these forced mandates. The mandates didn’t say, “And thou shall not balance it with fossil fuel.” And so they did.
They invested in balancing. And that balancing drives up costs as well. So the devil is in the details in terms of how you planned for that and how you really met that. You take other states, like my state, Georgia, no RPS. But they just said, “If you can beat avoided cost, you can get in the system.” And so you actually see solar substantially drive down the costs in that state.
Jigar Shah: Well, and even in Texas, right? I mean, because Texas has said, connect and manage, right? So we’ll connect to you, but we’re not giving you a guaranteed ability to get transmission capacity. And so they’ve curtailed a lot and magically the solar and wind folks have added four hours of battery storage because they don’t want to be curtailed. And so part of this is actually just putting the responsibility back onto the renewable generator and saying, “You have to play nice within the system, otherwise you’re raising costs for other people.” And in the renewable portfolio standard states, I would say that planning happens two steps later. It doesn’t happen at the beginning.
Stephen Lacey: Let’s talk a little bit about large loads. LBNL’s national average showed that load growth shaved off about 0.6 cents per kilowatt hour since 2019. And obviously as part of the conversation now, everyone is thinking about large loads as inherently driving up electricity prices. And that is not what we saw over the last five years. So Caroline, can you just explain what you started to explain at the beginning? Why do large loads potentially help lower the costs across the system, I guess assuming that infrastructure isn’t constrained?
Caroline Golin: Yeah. Yeah. So what I was saying earlier is that 2010 we’re over capacity in this country. And we’re also… Load growth is flat. 2015, you start seeing this wave of utilities really investing in the distribution system for a number of reasons, which we’ve gone through. 2017, you see the rise in data center deployment. So the gift that kept on giving, the utilities, was that they could continue to invest in the distribution system and they could recoup the costs of that over the entire… All customer classes. Which included these large loads interconnecting. And for the data centers, these were marginal costs. And they were actually fine to pay for them and they were concentrated. So if you look at Dominion, the timing is almost perfect, right? You see, I think it was maybe 2012, 2013, $6 billion grid modernization package. And you also see a 100% renewable energy standard coming out around that time as well.
We have to fact check on the exact dates. They’re hit at the exact same time with this massive growth in data centers. That data center growth actually kept the cost of both of those packages down for the average residential customer because it assumed in a cost per kilowatt hour… Well, you would’ve had to peanut butter across the entire system, right? Moving forward, that may be a different problem. And we can get into that. But in the states where you didn’t have these big large loads coming in, because it’s all about the numerator versus the denominator, those costs had to be spread across what was a smaller denominator. The smaller denominator means increased costs at the residential level and at the commercial level. Because things like grid hardening, grid modernization, all of these things, these are peanut buttered across all customer classes. They’re not ring-fenced, they’re not attributed based on distribution system versus transmission system. They’re paid for by everyone. So the more everyones you have, the lower the costs. If you didn’t get more everyones between 2018 and 2025, the cost went up.
Stephen Lacey: My favorite technical term is peanut buttering.
Caroline Golin: Yeah. Is that not something we use on this podcast?
Stephen Lacey: No, no, it is.
Jigar Shah: I love it. I love it.
Stephen Lacey: Yes. In fact, I think I’ve heard you use it in the past.
Caroline Golin: More often. Yeah.
Stephen Lacey: And it’s one of my favorites.
Caroline Golin: Well, yeah, I learned that with my PhD. It’s like just, you just peanut butter it.
Jigar Shah: Oh, my God. I have peanut buttering. But I do love eating peanut butter though.
Stephen Lacey: Yes.
Caroline Golin: I know. That’s actually how my husband and I met. Me eating peanut butter out of a jar. It’s a great story.
Stephen Lacey: That’s lovely. Wow. What a cross section of stories.
Caroline Golin: Yes.
Stephen Lacey: Okay. So, Jigar, what determines then whether that trend continues and what constraints on the system… What are the factors that make costs go up?
Jigar Shah: So first, let me just tell you something that just occurred to me. Which is, in 2012, I was living in New York City and was helping Richard Kauffman with the REV. And there was this woman who reached out to me whose name is escaping me. And she was like, “I hate energy efficiency.” And I was like, “What is going on? First of all, who are you and how’d you get my number?”
Caroline Golin: I gave it to her.
Jigar Shah: And she was like, “Mothers against high electricity bills in Connecticut.” Or something. Right? And the utility had just written her a letter saying that the reason your rates are so high is because you guys don’t use enough electricity. You guys… Basically Connecticut had reduced electricity consumption that year by 1%. So they had to raise rates like 5% to cover for that. And you’re just like… So that’s when it occurred to me that load growth is probably a good thing.
Then I remember Connecticut regulators were like, “We need more load. Everyone needs to work together to get more load.” The whole thing was just crazy. It felt like an Arnold Schwarzenegger pump you up thing. But anyway, it was nuts. But, yeah, look, I think that we’ve been doing this for decades and decades and decades. So this is not new stuff. Basically the way this works is you get a big steel mill, you get a big aluminum plant, you get this in, you get that in, whatever it is. And you decide, “Hey, they’re going to get an economic development rate because why not.” Rich people are going to get rich. And then basically you say, “Hey, when you hit a peak, it’d be great for you to shut down for five days so that we can manage the peak and not have to upgrade the system and do all that stuff.” And it was all done by telephone. I think even last week it was still done by telephone. And fine, right?
Caroline Golin: Pigeons actually I think in the early days. Pigeon carriers.
Jigar Shah: So it’s fine. And so then what happened with the data center companies is that George W. Bush decided that he was going to ban incandescent light bulbs in 2007. And we started moving into… Everyone going to LEDs and CFLs and all this other stuff. And so energy efficiency was really a thing. And so the data center companies were like, “Oh, there’s all these gaps in the places in the country and they’re desperate for us to actually fill those gaps with load.” And so they weren’t treated like the folks in the steel industry and other people were 40 years ago who were disruptive actors within the grid. They were treated with a red carpet. And in some states they even passed laws that said, “We’re going to exempt you from taxes and this and that. And please come on in. And we will give you an economic development rate.”
So now everyone is saying, “Look, in order to reduce the impact that you’re having on the grid, we would like for you to be flexible. We’d like for you to do this stuff, et cetera.” Which is basically the same thing we did the 1980s. And they’re like, “Ugh, that’s not how you treated us in 2012 when we came around. What? Are we going through a rough patch in our marriage?” And I’m like, “No, we just didn’t need to impose all of these requirements on you to reduce the impact that you had on costs for everybody else that was unfortunately allergic to peanuts. And so now we’re eating SunButter.”
And so we’re in this weird place where we still need load growth because we have historic underutilization of our existing assets. Most of the time we’re at 50% of our assets being used. And so it’s only the 3% to 4% of the year that we’re actually hitting a peak. And then at that moment, folks have to be flexible. And then that reduces bills for everybody. Because they don’t do all those upgrades. But we just didn’t impose that on folks in 2012. And so we’re in this weird spot where everyone’s like, “But you’re changing the rules of the game.”
Stephen Lacey: So Caroline, you’ve been on the other side of that table at Google. Does that feel consistent with your experience? What are those requirements and what requirements seem reasonable?
Caroline Golin: So I think that every industrial manufacturing customer has gotten an economic development rate so long as I’ve been in this industry. That wasn’t new. I think what was new was that our load shape looked incredibly different from everybody else. So you think about a steel mill, right? Incredibly peaky. And if you know how to place a steel mill correctly on your grid, you can make money off a steel mill, right? The smarter utilities did.
But data centers, always on, 24/7. Always on. If you know how to place a data center on your grid correctly, that can also be a real benefit. That can be a load stabilizer on your grid. If you choose not to do that and you choose to let developers decide where it’s going to go, that can become a real headache. What I think the reason… Where I see the disconnect is actually, eight years ago when the data center industry started ramping up, if you were going to ask a manufacturer or an industrial to be flexible, you had to open the can of worms across all those customer classes. And for years in this country, we have disincentivized demand response and flexibility at the industrial level for multiple different reasons.
But to the point where we have mandated state opt-outs. To the point where we have said, “You may not ask industrial customers or manufacturing customers to be flexible.” And that created a precedence that of course the data center industry it was going to utilize and why wouldn’t they? We’re no different than any other economic development customer. So don’t treat us differently. And this is the precedence you set for everybody else. And at the same time, you could argue from a technical perspective that data centers do function differently. If they aren’t on, there are goods and services in this country that rely on them always being on. And they can’t shut.
Jigar Shah: Those six, seven videos.
Caroline Golin: The KPop Demon Hunter, YouTube wannabe.
Stephen Lacey: Yes. Exactly.
Jigar Shah: You won’t be able to watch KPop Demon Hunters for the seventh time.
Caroline Golin: The world will come to a crashing end.
Jigar Shah: As everyone realized with the AWS shutdown last week. They’re like, “Where’s my Netflix?”
Caroline Golin: Where is it? Where is it?
Stephen Lacey: I want to unpack one thing there, Caroline. You used the word where. And I just want to understand what you mean by where we put a data center. Does that literally mean where it connects to the grid or how it interacts with the grid?
Caroline Golin: Yeah. Nodally where do you connect it?
Stephen Lacey: Yes.
Caroline Golin: Where do you have, and I’ll use the market term, where do you have potential negative pricing where actually load sync will be beneficial to the grid. Or where you have congestion. If you think… But no one thought this was going to be what it was. And at this time, seven years ago, you have two people at most at every single utility responsible for large load interconnection. And you’re getting five, six, seven, maybe 10 requests a year. So it is a process error as much as it is anything.
Stephen Lacey: Let’s talk about what could happen over the next five years. So if we have seen in states like Oklahoma or North Dakota, where data centers have helped moderate rates, how do we borrow that experience and carry it forward for other states when we are seeing this truly unique data center boom? And how much will data centers actually start to impact rates?
Caroline Golin: I think their impact is going to be directly proportional to the solutions they’re allowed to bring to bear when they interconnect. And if we maintain a posture that says, the only way that a large load can interconnect is through traditional upgrades to network infrastructure and through building new power plants, then they will raise rates. If we widen the aperture on what the solutions can be, I think that they will not raise rates. And I think potentially they could prime the system for continual cost stabilization when the next waves of electric transport, industrial electrification hit. I’m not confident that will be the path we take, but to me that’s the fork of the road.
Stephen Lacey: And when you widen that aperture, what kind of solutions are you now seeing? Are you talking about grid-enhancing technologies and distributed capacity? What does that aperture allow you to bring in?
Caroline Golin: Yeah, absolutely. All those things and then more. I mean, we can just go storage for example. The use of storage, whether it’s behind the meter or front of the meter, and utilized in a joint-contract structure, whether you look to sort of reduce peak demand on the distribution system through VPPs, efficiency distributed, and commoditize that, aggregate that up and allow for large loads to purchase that. Whether you look at dynamic line rating, whether you look at Volt-VAR. It is all of these renovations that allow us to not build a brand new room.
And so I think that a lot of the data centers, particularly… I mean, the hyperscalers have been wise to this for quite some time. But even the general data center industry is getting wise to the understanding that overbuild is a real risk. And if we overbuild, someone’s going to end up with those stranded assets and the utilities are going to make sure it’s not them. And so they’re stuck between the place of, are we going to just pay for the traditional system through capital allocation that may or may not speed up our ability to interconnect, or are we going to have to work and bring to bear new solutions that may be low capital costs, but high speed to electrons. And the issue right there is like, and I’ll go on this tangent for a second, is that’s actually the problem. The problem is that we incentivize capital deployment. We don’t incentivize electron deployment.
Jigar Shah: Wait, I think we covered this. This is the business model problem that we covered.
Caroline Golin: Yeah.
Jigar Shah: I mean, this is what I’m saying. It’s toxic.
Caroline Golin: But what I’m saying is that I think the data center industry is finally getting that.
Jigar Shah: Well, it’s about time. I mean-
Caroline Golin: Well, yeah, it’s about time, but the renewable industry isn’t completely there yet either. And they’ve been around a lot longer.
Jigar Shah: If they listen to the Energy Gang and now Open Circuit then they would.
Caroline Golin: If they listen to you, Jigar.
Jigar Shah: But I can’t force a horse to water.
Caroline Golin: But, and I said this yesterday, is like if you are a passive load waiting to interconnect, you can get it in 10 years. But you have got to bring solutions to bear. And I think the ones that figure out how to do that and can work with the utility on that, I think we could see some real changes. My concern though is that the utilities still, and I know we’re going back to this, are not incentivized towards that.
Stephen Lacey: Yes.
Caroline Golin: Right? Yeah.
Stephen Lacey: The conversation continually comes back to that set of constraints. Jigar, if you look at the next five years, just look at the large load piece. Do you think the story will be different than it has been in the last five years?
Jigar Shah: It damn well better be. I mean, I didn’t work in government and wrote the damn VPP liftoff report and the grid modernization liftoff report and all that crap for no one to do it. And so in the next five years, I think we’re in a political moment right now, where you’ve got governors making promises that they don’t know how to keep. And the only way that they can keep those promises is to make the changes that we’re recommending. And even if you decided that you woke up on the wrong side of the bed and you really wanted to solve all of these deals with modified jet engines that you’re putting behind the meter at your data center, there are not that many modified gas turbines to be bought. We’re the only solution that could reduce people’s electricity bills and meet speed to power for the data centers in this moment. Whether it’s behind-the-meter batteries, whether it’s a more broad-based set of demand flexibility solutions on the balancing area, or whether it’s something else.
I just think that we are hurting our own ability to extract the maximum amount of economic development that this economy can create because people are just not getting organized. And I know what that looks like because when I started my career, I helped facilitate this thing for the Clean Energy Council or Consumer Energy Council of America, and it was basically with Jim Rogers. I think at the time it was at Cinergy. And it was all about deregulation. And so when you think about the fact that McKinsey and Accenture and all those people were making $300 million a year advising the utilities on how to spin out their power trading groups and all that stuff, that’s what it looks like for people to take change seriously.
All of the work that people were doing, working groups and this and that, around deregulation, was real. The stuff that’s happening now, I mean, it’s like a small conference here, an EPRI conference there, this thing or that thing. People are not taking this moment as seriously as they need to. Now, maybe they are on the data center side finally. Thank goodness. But the governors are not getting educated. I don’t even think that the capital markets, that Wall Street’s really getting educated on what this looks like.
Stephen Lacey: Okay. So we talked a lot about speed and flexibility here. I want to bring this into another piece of news. So last week, the federal government stepped into new territory, maybe potentially a turf war. Energy Secretary Chris Wright has asked the Federal Energy Regulatory Commission to fast track interconnections for large loads like data centers and industrial campuses, giving them a 60 day, what I’ve heard called, the shot clock to connect the grid. Under the proposal, these large loads that agree to be flexible, curtailable, could get priority treatment. So I want to get your reactions to this development. I know some critics warn it could potentially strain reliability, shift costs onto other customers, encourage large loads before there’s enough generation or grid capacity to meet them. What do you think about this move? Caroline?
Caroline Golin: I actually think there’s some good policy in this. I’m concerned the process may diminish the returns on some of the policy that’s in there. It’s a very short timeline. There’s no real clarity on how comments will be assessed. There’s no real clarity on what DOE wants for it to do if utilities don’t comply with this. So there’s a lot of process questions I think that are up in the air and I’m concerned that process could get in the way of some good policy. It’s clear here though, if you read through it, that the administration is saying, let the hyperscalers, let the data centers pay for it. They’re going to pay a 100% of their fees coming in. It’s clear that they’re saying, let them co-locate. Let them bring the natural gas or storage, I think this could be a big win for storage, to the table. I think it’s also clear that they’re saying that you have to start looking at load and the load queue in the same way that you would look at a demand queue. Or a demand queue in the same way you would look at load queue.
And so I think that there’s a lot of good theoretical posturing. My concern is two things. One, I go back to this idea of the aperture is not large enough. It’s very clear that the only thing that can happen to solve this problem is co-location, data center flexibility. Those are two tools in the toolkit, but they are not the full toolkit. And I would love to see comments and engagement around widening the aperture around where flexibility happens. Does it have to be at the data center level? Can it be on the distribution system? Can it be broadly within the market? How can you package that up? I’d also like to see more conversation around avoiding network upgrades. Right now it just says that the data centers are supposed to pay for a 100% of the network upgrade. In a traditional approach that could not be the best and most efficient solution.
And so I’d like to see more conversation about how we look at a broader set of solutions in terms of those traditional network upgrades. The last thing I’ll say is that if you don’t marry this with sort of state legislation that incentivizes the utility to go after all those other wider aperture solutions, you’re going to have grid lock, right? Because I don’t think you can put a stick without a carrot right now, to oversimplify it, and see this go quickly. And so if I’m a governor, I’m thinking about, “Okay. What is the basis point change that I need to incentivize in order to make sure that the utilities feel like they can invest in the AI solutions, that actually understand their grid topology better, hire the workforce to do this more quickly, do the analysis and look through the alternative mechanisms that make this go fast?” If that doesn’t happen, I think you’re going to be, pun intended, a bit gridlocked.
Jigar Shah: No, I think as usual, I’m in violent agreement with Caroline. Look, I think the thing that I find fascinating-
Caroline Golin: I’m putting that on my car.
Jigar Shah: Yeah. Exactly. I always let Caroline talk first. I think, first thing first, this is Travis Fisher’s idea. And so this is Cato Institute.
Stephen Lacey: Yeah. Travis Fisher from the Cato Institute.
Jigar Shah: Right? And so this is what he passed in New Hampshire. He’s like, “People should just be able to do whatever the hell they want.” Totally. I love it. I’m here for it. Basically the administration is going to war with the electric utility sector and I am here for it. The fact that the Trump administration decided to go to war with the electric utility sector, I love it. Let’s do this. And then the other piece of this is the reason this is happening is because in the mind of MAGA, this is all behind-the-meter natural gas. Right?
Stephen Lacey: Yes. Right.
Jigar Shah: When the rest of us know, “Hell, no. That’s not going to work. This is going to be batteries.” But I don’t need to talk about it that much, right? And so I have popped so much popcorn, I put some masala on there. I’m like, I’m eating it. This is going to be awesome. The other thing I would say is, we just had DERVOS. I had the top 60 CEOs from all of the demand flexibility companies there. We had this big conversation. And Lord Almighty, is this a gift, right? Because getting those 60 people to all row in the same direction is hard. But now they’re all rowing in the same direction. Why? Because Caroline just told everybody that they’re part of this broader aperture. But they’re not part of the broader aperture unless they put their comments into this thing, which is due on April. And so I don’t think a lot of people are going to put comments into this ANOPR just because it’s very clear it’s due by November 15th. And then we’re going to go into a NOPR for sure.
So they’re going to save their best comments for the NOPR. But I think this is going to be amazing. And then in terms of the state legislature piece, this is again where the CleanTech folks and the environmental folks got to coordinate a little better. And so we want the state legislatures to pass a law that says that the utilities are required to get better asset utilization out of their existing assets. That’s it. Right? It shouldn’t be banning natural gas. It shouldn’t be banning this, banning that. It should just say, “You’re at 40% asset utilization. You need to put together a plan to get to 60% asset utilization.” Right now the regulator is lined up with a state law to regulate the utility around asset utilization. The utility knows that. Tesla made this argument in 2013 during the REV. So this is not-
Caroline Golin: I made that argument, Jigar. I was [inaudible 00:55:20].
Jigar Shah: Caroline Golin made that argument in 2013. So we are so ready for this war that Chris Wright wants to have against the electric utilities. And I am here for it. And the last thing I would say, it’s very obvious that this letter was directed to the PJM. It is not like all ISOs, all transmission operators, even though it clearly is, but it’s really directed to the PJM. And again, I’m here for it.
Caroline Golin: Well, and I think if you are not in PJM, you’re vertically integrated. You’re looking at this as license to operate to create a load queue and placement of where you want these data centers on your system. And that is a gift to a lot of these utilities.
Jigar Shah: Totally.
Caroline Golin: Because for a very long time, they have been regulated into a place of almost paralysis. And I don’t think we talk about it enough, but managing a load queue is just in operations and maintenance costs. And many regulators across the country are not letting these utilities hire and staff up and create the tools to actually do this effectively. I don’t exactly understand why. It seems asinine to me, but that’s what’s going on. And this is a gift that says, “Oh, no. No, no, no. We want us to do this and keep the lights on? You want us to do this reliably? Now we tell these large loads where they go.”
Which is what has needed to happen for such a long time. And I think that if I’m a utility looking at this outside of PJM, I’m also putting in my comments. Because this goes one of two ways. And it doesn’t have to be a war on the electric utility. If they look at this and scaffold it in the way that’s going to help them manage their system as opposed to being this reactive tool across what is a never ending, never ending load growth story. This is not going away.
Jigar Shah: Bless your heart.
Caroline Golin: It’s true.
Jigar Shah: I absolutely am here for it. I would love for the utilities to be like, “This is not a war on us. This is just a war on our business model.” Or, “This is just a war on our calcified thinking. We should just become a modern utility. I should just earn my $26 million a year salary instead of actually acting like I just run a water utility and I make $500,000 a year.” I don’t care what argument they tell themselves, but if we get innovation that we invented 20 years ago to now get accelerated at this exact moment when it can reduce everyone’s electricity bills by 20% by 2030, I’m here for it.
Stephen Lacey: Indeed. And I’m so glad that Dr. Caroline Golin could be here for it as well. Caroline, thank you so much.
Caroline Golin: Thank you, Stephen. This was great.
Stephen Lacey: That’s going to do it for this week’s Open Circuit. Jigar, great to see you. Good conversation this week.
Jigar Shah: Always.
Stephen Lacey: Open Circuit is produced by Latitude Media. The show is produced and edited by me, Stephen Lacey. Anne Bailey is our senior podcast editor. Sean Marquand is our technical director. He wrote the music to the show. And for more in-depth reporting on all of this stuff go to Latitude Media. You can sign up for our newsletters, including our AI-Energy Nexus newsletter, where we cover a lot of this. And of course, you can find Open Circuit anywhere you get your podcasts. Please give us a rating and review, a five-star review. That would be really helpful and we’ll see you next week. Thanks everybody.
