Transcript
INTERVIEW
**Jigar Shah**
Katherine, welcome to Energy Empire. I was looking around, and it's not clear to me that you've done a lot of podcasts.
**Katherine Blunt**
I think I've done a fair number of them, but I haven't in a while, so I'm super happy to be here.
**Jigar Shah**
Well, you certainly did a lot when your book came out, California Burning, which to this day is sort of the book that all other books on the electric utility sector are compared to. It came out with just extraordinary fanfare. I think Patti Poppe said that she made it required reading, I think, when she became CEO.
**Katherine Blunt**
Yeah was really gratifying, right? It's just like if it even helps like one person's decision-making or understanding of the challenges, that feels really good.
**Jigar Shah**
Well, now there's like 500,000 people who I think are cosplaying electric utility experts in the AI industry. And so I think there's a lot more people who need to like to read your book, so hopefully sales spike here. I wanted to start with part of what you talk about in your book is that PG&E really focused more on shareholders over their ratepayers for a long time, and that really caused a lot of their challenges. Today the electric utility industry seems to be doing the same. I mean, I think they were at $20 billion a year of CapEx in 2003. They were at $178 billion of CapEx last year. They're gonna be at 215 next year. The Edison Electric Institute said that from 2025 to 2029 they were gonna spend a trillion dollars. Yeah. They just updated that number to $1.4 trillion from 2026 to 2030. You know governors are like apoplectic, right? They're just sort of like, "Why are we running in a race where we have to talk about electricity costs?" I'm wondering whether anyone has learned a damn thing from your book.
**Katherine Blunt**
Yeah. So it's a fascinating, extraordinarily challenging time, and I think when you think about CapEx spending, right, I mean, some of this needs to be done. The grid's old, stuff needs to be replaced or upgraded or whatever. But yeah, I mean, what an opportunity to generate returns for shareholders, right? And sort of justifying massive spend because of this AI boom and AI race and trying to make it happen for these companies that are saying like, "We absolutely need this. We needed it yesterday. How fast can you deliver?" And I mean, it's a fascinating convergence, two of these industries, one that is notoriously slow and backward-looking and just really can't move quickly, risk-averse versus the AI giants that are not risk-averse, trying to move mountains in the span of days.
**Jigar Shah**
Move fast and break things. Let's hope it's not the grid. Although the thing this time is the grid. That's not something you wanna break.
**Katherine Blunt**
Yeah, you don't wanna break that, and I think that, yeah, the challenge is that I've been very interested in what's going on in PJM these days. This seems to be ground zero for some of the challenges you'll begin to see elsewhere as the data center build-out continues.
**Jigar Shah**
And I'm curious on the politics side, it feels like when PG&E went bankrupt, Gavin Newsom sort of made it his mission in life to get them out of bankruptcy and back to investment-grade status. Today, it doesn't feel like the 36 governors running for election, Republican or Democrat, want to protect the electric utility. And now with all of this anti-data-center sentiment, it's not even clear whether they think it's right for them to protect the AI data centers, right? Hmm. And so I'm curious in this moment where the governors are. They need to get reelected. So where does this sit politically in terms of whether folks really want to see all this spending?
**Katherine Blunt**
Yeah. So it's a tough question to answer because, I mean, I think it kind of varies by region. But I mean, certainly within PJM, the incredible spike in capacity costs has made it so that people are thinking about electricity prices, PJM's role in that, what role the market plays. Three years ago, five years ago, no one outside this business had any reason to even know what PJM was, right? Much less what it was. And even if they try, they probably don't understand it. But yeah, you saw it in Virginia, you saw it in New Jersey, right? this rallying cry to bring down electricity costs, which of course is not an easy thing to do by any means. And PJM has a role in this, but I mean, state policy has a role in this. I mean, so many things in this space, it's just the convergence of so many challenges over many years, and then you layer on top of that what's supposed to be extraordinary demand, possibly inflated projections. Who knows, right?
**Jigar Shah**
Well, most certainly inflated projections, right? Yeah. I mean, isn't this the story of the 1990s when MCI WorldCom basically said, "We need to build 12 coal plants a month"?
**Katherine Blunt**
Yeah. I mean, very well could be. But I sometimes say, and this is not a popular thing to say, it's like I think we've moved into an era in which electricity necessarily has to cost more just because the grid needs to be upgraded for all different kinds of reasons, and just stuff is old, right?
**Jigar Shah**
And is that your conclusion? That like, out of all of that reporting you did, that the next generation of technologies, virtual power plants, grid-enhancing technologies, grid utilization, et cetera, just isn't enough.
**Katherine Blunt**
I don't think it'll completely offset. I think it could meaningfully mitigate the need for some massive CapEx spending, I think. What do you think about that?
**Jigar Shah**
I mean, I obviously served in the US Department of Energy for four years. Yeah. I think we showed, through the Liftoff Reports, et cetera, that you actually could get far more utilization, return back to the 1970s level of utilization, right through flexing demand with the same level of dexterity that we currently only flex supply. And you could get to much cheaper rates for everybody, right? But one of the big challenges I see is that requires a level of coordination between the utility, between the data center, between the governor/public service commission that just rarely seems to happen.
**Katherine Blunt**
That's exactly right, I think. I've been very interested too in various proposals to have, to require flexibility of data centers. Of course, the AI companies say like, "Oh, we're happy to flex, but we would like that to be a voluntary requirement." Right?
**Jigar Shah**
Well, now Google has hired Tyler Norris, and so he now works there inside.
**Katherine Blunt**
Yeah, I was talking to Jesse Jenkins the other day, and he was like, "Oh, I think, most of my friends work at Google at this point."
**Jigar Shah**
When's Jesse moving over? He's great. Although he's got that company that he works with now, right, Fermata.
**Katherine Blunt**
Yeah, he's got that going on.
**Jigar Shah**
Yeah, I mean, so you've moved beats from PG&E to Alphabet Google. There's a lot of overlap here. There is. But talk a little bit about that move and why you made the move from PG&E to Alphabet Google. I mean, obviously they're a $5 trillion company but you know.
**Katherine Blunt**
Just about. Just about. You know. Yeah, it was about a year ago that, after seven-ish years covering utilities, I just thought it would be interesting to take on a new challenge to try to do a different kind of corporate reporting. But I was also. And I made clear in making the jump that I really wanted to continue to keep an eye on the AI power nexus, I mean Google and everybody else, and what it, and the push to build out these massive facilities and what it means to even have half the projected demand growth for these utilities. And it's been, it's been very interesting. I mean, we talked about PJM, but you may be seeing it elsewhere. I mean, gosh, I think that there's two projects in Wisconsin right now, one for OpenAI and one for, I think for Microsoft, and I think it's gonna double the overall demand on the system for that utility just those two centers, which is like, it's remarkable, right? And it has been really interesting, too, to try to understand and track the ways that utilities are trying to design new tariffs to protect other customer classes. But as with everything the devil's in the details. How is this all gonna work, especially in allocating transmission spending?
**Jigar Shah**
Yeah, I can't even imagine being a reporter in this moment, just because I feel like these are esoteric concepts that matter, right? I mean Mm-hmm. It is what's driving inflation. It's what's driving the GDP growth of our entire country. It's certainly booing the stock market, so it matters, right? Mm-hmm. But when you think about the patience that people have for reading about these things, I mean, when you think about the structure that Google created in Xcel Minnesota's territory Mm-hmm and then Alliance, Allete's territory, where they really went to the utilities and said, "Using Google Tapestry, we're gonna teach you how to do this better and in a way Mm-hmm. That actually would reduce rates for everybody in Minnesota by," I think it was $1.7 billion or something, is what they filed with their ESAs, their energy savings agreements, I think. And but, like Like, I don't even think there's ever been any reporting about the anatomy of how that deal came together and sort of what that
**Katherine Blunt**
Yeah, I got the wheels turning. I should
**Jigar Shah**
Right. But I mean and what that means, right? For the fact that Meta did nothing at all in the negotiations with Entergy down in Louisiana. Yeah. Or, and what does that mean for Meta and Microsoft? Are they just lazy? Does it mean that they actually should do the same amount of work as Google? 'Cause I think everyone sort of suggests that data centers are good load growth, right? But it feels like it's only good load growth if people coordinate.
**Katherine Blunt**
Yeah, I would say this. It has the potential to be good load growth eventually, right? It's just like, how do you manage the upfront costs of inter integrating them into the system? And it's- You make a great point about Google. I think that, and I've written this, Google has assembled one of the smartest teams, I think, in the space on energy, right? Yeah. They've been paying close attention to this for a long time, and I think their understanding of the system really outshines most other competitors, and I think it'll allow them, most likely, to get stuff moving more quickly because they actually have answers to some of the hurdles, right? Whereas I think the competitors, its competitors have been slower to try to come or to reach the same sort of understanding and certainly don't have the same bench of talent.
**Jigar Shah**
Yeah, the other part that I think is so confusing and for everyone, is what's true and what's not true. Right. Like, when I try really hard to figure out why do we need one-plus gigawatt data centers, right? Mm-hmm. Why can't we have a network of 100 megawatt data centers, which you and I both know Right. Would be a lot easier to fit into the grid. Sure. You know, people say to me, "Well for the final training runs, we need the chips to be right next to each other, otherwise they can't work in synchrony and and get everything done," right? Fine. So then I'm like, "Great, how much of that do we need?" And it turns out a lot less than I thought, right? The current models sort of use 200 megawatts worth of power to run the models. I mean, they think that by 2030 they're gonna need maybe four gigawatts of Mm. Capacity to be able to run one of these final training models. And then when I dig in deeper, I hear, "Well, the R&D runs don't need to all be with chips right next to each other." The R&D runs that they do are batchable, and they can sort of like farm them out Mm. To 100 megawatt data centers. So I'm like, "Great, so then, so then what's the answer?" Like, how many, how many one gigawatt plus data centers do we need for training, right? And everyone's like I don't know, but I'm lazy. I'd rather all of them just be one gigawatt plus data centers, even for inference or whatever. And I was like, "But that would break the grid." They're like, "Not our problem. We don't care." And I- It's just like, I'm curious how transparent these companies are about these big questions. How many mega data centers do we need? And like, 'cause Jensen Huang has been talking about inference and how he's got five megawatt data centers with Prologis, and now he just did a test with Span, right, and Pulte Homes. And then he's like, "No, this stuff can be spread everywhere. Inference can be done easily." I like I just- If the electric utilities are gonna say, "We need to raise rates on all these people to like upgrade all of our facilities," they're by definition sort of assuming that these inflexible, very large loads are the way that we have to serve AI data centers, right? And it doesn't seem like there's any answers coming from anybody that I can cite.
**Katherine Blunt**
Yeah, it's a great question. I think I don't even know if anybody really knows the answer to that or has given it a lot of thought. I think right now there's just such a scramble. There's just been this movement building bigger and bigger data centers. There are other ways to do it, as you were just discussing or alluding to. I am not smart enough to tell you exactly what loads could be split. I've read.
**Jigar Shah**
Your reporting for years, so I know that you're smart enough.
**Katherine Blunt**
Well, th I appreciate that. I wish I had better answers as to like what is the actual need for a super-sized data center versus something that's more small workloads aggregated, right? I mean, it makes sense, and
**Jigar Shah**
Well, I, it's not that you have to have all the answers. I'm the, my question is really who's supposed to have all the answers? With all of these smart people working at Google, should they be revealing these answers to you? Like, what about Epoch AI Research, which seems to be this nonprofit that publishes everything? Right. Or is it the US Department of Energy or EPRI? Like, it just feels like someone should be on top of all of this stuff with a PhD telling me what the answer is.
**Katherine Blunt**
Totally. Gosh, when you think of who has any actual oversight over any of this, I mean, is it FERC? Is it at the state level when the utility comes forward with being like, "We have to serve this one gigawatt data center"? And then should the regulators ask does it need to be all situated on that plot of land, or is it possible that it could be broken up into 16 pieces and aggregated, right? I don't know. It's like, but it's imperative that we begin, we, I mean, not me, well, me as a reporter, sure, but also as someone at the regulatory level or within the administration or something, right? Basically being some sort of stress test or some sort of analysis as to is this, is this the most prudent way to do things? You gotta ask that question because it's untenable what's being proposed, right? It can't, it really can't.
**Jigar Shah**
Happen. I mean, I think so, but I feel like the utilities are just pushing their way through. You know, one of the organizations I'm a part of, called Deploy Action paid for some polling, and you saw that Josh Shapiro wrote that letter Yeah and and basically got a utility to just withdraw their entire rate increase. Right. And so it does seem like the polling is suggesting massive majorities of voters in California are saying PG&E shouldn't get any rate increases at all until they've demonstrated increased grid utilization, right, and using the grid that we've already paid for more cost-effectively. I mean, it just feels like there's a trust breakdown now where electric utilities are looked at and saying Do- Is your stuff old? Does it need to be replaced? You said that five years ago. You said that two years ago. Now we're paying 44 cents a kilowatt hour. We went bankrupt. We had wildfires. Like Yeah I don't know that I could trust anything you say. Like, I'm just trying to figure out how you think about operating in this environment where These utilities are coming to the regulator asking for these rate increases. There's really no transparency as to whether the AI data center companies really need one plus gigawatt data centers, and grid utilization is at historic lows. The ability for utility companies to use technology is heroically like something that they're not good at. Uh how do you view this trust breakdown?
**Katherine Blunt**
Well, it's been happening for a very long time, and then with the, with the AI build-out, it just comes at exactly the wrong moment. Right? It's like it's a really bad time to try to make this happen. I mean, sometimes when the utility comes forward with a rate increase or spending, sometimes the answer is yes, some of that spending is prudent and necessary. I mean, we all know though that sometimes it's not, and it's very hard for overworked regulatory staffers to like parse every single project and figure out whether it's gold plating a substation or actually making an upgrade that's gonna be beneficial for the entire system, right? And so with PG&E in particular, I mean, they've just had such an awful run of it over tw 20 years, capped by the awful fires in 2017 and 2018 and the bankruptcy, and it is so eye-wateringly expensive to live in California as it is. I mean, gas right now is at seven bucks. I think I paid $7 for premium gas the other day, right? And it's just like no one wants to
**Jigar Shah**
Yeah, you live in California now. I do. Like, I mean, you're eating the dog food here.
**Katherine Blunt**
Yeah. Totally. And I'm lucky living in San Francisco that like I can mostly keep the windows open and I've got a I don't really need to run the electricity all that much. But for folks in the Central Valley, I mean, we're heading into summer, right? It's like it's, those are some really significant power bills. So PG&E lost the trust of customers a long time ago just because of like certainly the extent of the disasters. Other utilities have had major disasters as well. I mean, we're, we're also living in an era of changing climate and different stress on the grid, right? And this is a backward-looking industry that ha like there's a
**Jigar Shah**
Well, look at Georgia. Huge range. Right? I mean Georgia with the Vogtle nuclear plant, I mean, two of the Oh, God. Longstanding Republican public service commissions got voted out by like 20 points, like Yeah. In a, in a arguably red state, right? And so Right. You know, that's crazy. I mean, but what do you think is the right way to build, rebuild trust? I mean, should regulators force the utilities to demonstrate and to prove that they're using the stuff they've already paid for more efficiently before they ask for more poles and wires? Like, I mean you've covered that dynamic, right?
**Katherine Blunt**
Yeah. I think that should be, that should be part of the solution, and there's, there's a lot of companies right now that are trying to work on that problem, right? To figure out how to find change under the couch cushion, so to speak, right? To find the ways in which you can slot Stuff in here and there and what's not being from a transmission standpoint, what's actually not being used at this time. Like, all kinds of ways to, like you say just to make better use of what's already in the ground. It would be great, I think, to see greater interest in that as part of the suite of solutions. And from a regulatory standpoint, I think it does make sense to ask the question, for sure. What do you think?
**Jigar Shah**
Well, I mean, I've just been shocked by, in the polling, the voters are not mad in a vague way, right? Like, they actually support super specific things, like saying we should tie executive pay to some of these metrics, right? I mean, I guess what I'm trying to figure out is with With your reporting, what is required to rebuild trust, right? Because I just think when you look at Georgia and the Vogtle nuclear plant, when you look at Duke Energy and the fact that they got credibly accused of really dumping coal fly ash into the waterways Mm-hmm. During hurricanes, and now they're asking for 15% rate increases. When you look at Right PSE&G in New Jersey, who credibly got dinged for gold-plating all of their, infrastructure post-Sandy. You know, now that investor-owned utilities are at 30% above the cost of rural electric co-ops and municipal utilities Right what is sensible, right? What can voters expect, or what do ratepayers expect to achieve accountability? I mean, should we be putting this into executive pay?
**Katherine Blunt**
I don't know the answer to that. And I don't frankly know the answer as to how some of these companies rebuild trust after everything that's happened. I think that rightfully so, ratepayers, voters are wary of massive projects or spending or really anything that has the potential to go south, become stranded asset be cost overruns a la the nuclear plant in Georgia. I mean, that was massive, right? Or like what are you doing with all this CapEx and what are you doing with O&M, right? That's the, that's the PG&E story, which is that they were re they were focused on anything that could add value to the system and not paying as much attention to doing the day-to-day inspections and maintenance that would've made their pipeline safer, and San Bruno happened, made their transmission line safer, the Camp Fire happened. The thing though, and I think that people are becoming more aware of this, interestingly. I mean, gosh, the number of regular people I see intervening in rate cases these days is really pretty remarkable, right? And they didn't all find their way there by themselves. Like, there's certainly groups that point people to Sure ways to make comments and things like that. But utilities, and I sometimes say this somewhat jokingly, but it's not a joke, I mean, they do actually write everything down, seriously. Like, I mean most things can be found in various regulatory repositories if where to look. And so it's just this question of oversight and asking more creative and probing questions about the necessity of any of this and what the alternatives might be is imperative. And like, I mean, as the system as it stands, that means that has to come from the probably state regulatory level, and FERC probably has a role to play too. Are we gonna see that meaningfully change? I don't know.
**Jigar Shah**
So if you were to write a story on PG&E now, given that you're a PG&E customer, is the story basically that You know, people hate their utility? Or is the story that the old utility model is breaking down and that it is going to have to be sort of replaced in order to gain voters' trust?
**Katherine Blunt**
I mean, I think it's very fair to say that the old way of doing things doesn't work in this new environment. When I try to think about it pragmatically and how difficult it would be to change, to do some sort of wholesale change, right? I mean, after PG&E's bankruptcy, there was talk about trying to make it into a co-op, or would the state take it over, and I don't think that meaningfully solves the problem. Honestly, I don't, and I think that there's. The most pragmatic way to think about it is I mean, how do we identify the points that need to change the most within the way that it already works? Like, how do we, how do we make the current system work better, ask better questions, have better oversight? You know, maybe, like re like, reduce ROE, right? Right. Maybe re Like, that's probably an obvious one that I think should prob should definitely be explored in this environment if they're proposing to spend however much you said, 1.4 trillion through, right? I mean Yeah.
**Jigar Shah**
All of the EEI utilities across the country.
**Katherine Blunt**
Yeah. Does it have to be 12%? I would say no. Like, and often, I mean, they're able to exceed that sometimes. Should that be allowed? Probably not. What additional risk will shareholders take on, right? That's a, that's a difficult question. We saw PG&E, I mean they were supposed to, in a perfect world, shoulder the liability costs of the, of the fires, but that didn't happen, right? They all cut and run. And how do you redu like, reduce costs and reduce risk to the residential customer class as you have these massive commercial interests coming in and wanting all kinds of things built on their behalf? I mean, that's, that's the, that's the big question. How do you do it in a traditional way? How do you do it in a novel way? And these are questions that I hope people are asking, and I'll, I can ask, but I have no, I have no power over whether it happens, right?
**Jigar Shah**
No. I mean, I'm asking just mostly because you covered them for so long and your book was so comprehensive. I think moving to the next sort of section, there's been this whole conversation around cost allocation, right, and who pays, right, between the utility and the AI data centers, right? I mean, PJM just announced the largest upgrade to their transmission system in a very Yeah. Long time of $12 billion, half of which was peanut buttered over all of the rate classes, right? And so part of what I'm trying to understand is on the one hand you've got the data center companies who've made pretty big sort of promises, right, not to shift cost to consumers. But then on the other hand, you're like You just like shifted $6 billion worth of transmission costs to all the ratepayers and had them all pay their fair share. Like, how do the AI data center companies do this at a media and comms perspective? Like, how do they convince voters that they've actually succeeded in meeting the pledge.
**Katherine Blunt**
Yeah. Yeah, I think frankly all of the companies are trying to staff up in this regard and trying to tell that story differently just based on what I can tell. It's been interesting. So I've written about this a few different times now, and I would say even a year ago when you looked in the, various filings related to creating these lar the quote, unquote, "large load tariffs," right? Everybody, Microsoft, Google, Amazon, were like, "Really? Do we actually have to pay more? I don't know. Like, that doesn't seem very fair. Like, we're, we're. All these upgrades are gonna be beneficial for everybody." And like, that fell.
**Jigar Shah**
Away We're, we're used to being on the economic development rate.
**Katherine Blunt**
Right. Right. And so there was a lot of. Not too long ago, there was quite a bit of resistance to some of these new measures that were meant to guard against the risk of stranded costs, guard against the risk of massive cost shifting. I think they got wise pretty fast and realized that's not really a tenable position to hold, and also as they began printing more of their own money, so to speak, they're like, "Whatever." Like, "We'll just. Whatever it costs, we'll pay." But to your point, transmission spending in particular is gonna be the toughest, the toughest thing to address as it relates to cost allocation. PJM is a good example, but there's other, massive transmission spending plans that are gonna be largely in service of the AI companies. But you can make some sort of argument that it's gonna be an overall system upgrade. It's just like, but to what extent is that actually necessary? That's a very difficult question to answer.
**Jigar Shah**
Yeah. And part of this actually then contrasts with the electric utilities and their conservatism, right? I mean, because they don't necessarily want to take advantage of demand flexibility. Like, not all the utilities have actually required data centers to sign interruptible rates, right? So that they can mandate that in period of most stress they have to turn down. I mean, it got, it got so bad in Texas that the legislature passed SB6, right, and said, "No, we're passing a law that requires you to sign-" Yeah."interruptible rates on these data centers."
**Katherine Blunt**
And correct me if I'm wrong, but I think that's the only place right now in which it's a requirement. I'm pretty sure it'
**Jigar Shah**
Think that's the only place where it's a requirement. In California Oh, agreed. Obviously PG&E's got the FlexConnect for distribution and then T-Flex for. And so there's, there's lots of utilities who are moving that direction, but it's not mandatory.
**Katherine Blunt**
They're trying to make it an incentive, right? They're be like, "You can have a faster interconnect if you agree to interruptible service at least for a period of time." And
**Jigar Shah**
Yeah. I think that's right. I mean, and it's basically because When you look at the way in which utilities have adopted the NERC best practices, right, the National Electric Reliability Council Mm-hmm. These, uh Yeah. These acronyms get me too. There's a lot of places on the grid, right, that Grid Care and Google Tapestry and Camu and ThinkLab and others have identified that can accommodate these load growth opportunities, on the existing grid during normal situations. But then what the normal utility standard is that if this line goes down, what happens? If this generator goes down, what happens? If this happens, if that happens, this sort of N minus one minus one sort of standard, right? Mm-hmm. And so then what they find is they have to magically spend $8 billion worth of upgrades to be able to meet that standard, right? And so if they can get data centers to sign interruptible tariffs, then they can say, "Well, we'll make best. We'll do the best we can," right? But ultimately if Worse comes to worse, then we're gonna have to get you to shut down because we have to save the grid. But in the meantime, I think when you look at PGE's T-FLEX standard Mm-hmm. The way that they're implementing it, you would only have to turn down your data center once in 10 years. Right? Mm-hmm. So it's not like it's gonna happen for 200 hours a year. It's, if you back test it rarely happens, but that really simplifies the way in which the utilities can do the measurements, but then it reduces their rate base, right? It reduces the amount of money they can spend on their system.
**Katherine Blunt**
Yeah, it does. And I mean, it'll be interesting to see what happens in California. I think that for the most part, I think tech companies would mostly prefer to go somewhere else just because electricity costs are so high.
**Jigar Shah**
Here. But sometimes they can't. I mean, California does need compute, right?
**Katherine Blunt**
Sometimes they can't, yeah. I mean, a huge thing that, complicates all of this is the fact that real expertise on the way the grid actually works is a very hard one, and as people get more up in arms about it, they're paying closer attention, but they don't necessarily understand it better, right? That's something that motivates me in my reporting and trying to at le at least just help people understand why the thing isn't working. You know, I wrote a story about PJM back in, a long one, back in January, and just it wasn't this, like colorful yarn. It was just a long story about why PJM is pretty close to being in shambles, right? And it was like, it was like the most.
**Jigar Shah**
Frustrating I think they are officially in shambles. They put out a white paper I think they are officially in shambles. I think they put out a white paper last week that basically said
**Katherine Blunt**
I read it. I read it. "
**Jigar Shah**
This doesn't work anymore." Like
**Katherine Blunt**
That was a fascinating, that was a fascinating white paper. It was, it did start off saying "This doesn't work anymore. Here's three ideas that we might explore or combine going forward." And the first one and the third option are each sort of somewhat confusing and convoluted, but the second one is very easy to understand, which is that we might move into an era in which reliability is not guaranteed for anybody, and how do we figure out who loses power when Which is conditions warrant, which is so politically unpalatable, right?
**Jigar Shah**
Well, but it's a, it's just a glass half empty way of saying what I've been saying that's glass half full. The way that I say it is, hey, for people who wanna opt in to demand flexibility and to allow their thermostat to be varied by two degrees or to let their water heater be controlled by the utility or let their Tesla power wall in their garage be flexed by the utility, they should get paid something for it, right? Yeah. They should get paid that thing, right? That's just the other way of saying what PJM said, which is PJM is basically saying that we are no longer going to honor that consumers can do whatever they damn well please, and we have to meet that requirement, right? Yeah. We are gonna set standards such that consumers can't just do whatever they damn well please, and that's how we achieve a more cost-effective grid.
**Katherine Blunt**
I mean, if you put it strictly in the terms of like DR, like then that's fine, right? It's like, I think though, I mean, I, it's been a while since I really dug into this, but it's been interesting how challenging it's been to like get good demand response programs on the residential side integrated in different market structures, right? And like We.
**Jigar Shah**
Had 75 case studies in the VPP Lift Off report that we published in January 2025. I think they have them. I think the problem is that it's culturally like unacceptable right now. Like, I mean, what Tyler Norris said in his work, right, is that for I think it was like 74 hours of grid flexibility with no event being more than two hours, right, you would capture like 74 gigawatts of additional capacity on the grid. Right?
**Katherine Blunt**
That's a, that's pretty staggering.
**Jigar Shah**
Yeah. But I'm just saying like right now the standard is no hour left behind, and so like we will spend whatever it takes to not have 74 hours of demand flexibility. I mean, that's basically what PJM I think is saying in that white paper is they're saying Maybe we should allow for 74 hours of demand flexibility, and then we'll allow each state to set their own standard. If Ohio wants to spend ungodly amounts of money and let the utilities rate base it to not have those 74 hours of demand flexibility, they can do that. If Virginia does want to do that, then they can do that. If certain states want to add a bunch of batteries, they can do that. If certain states want to add this, they can do that. Like, I think, I mean, I think that's what PJM was saying, is that we'll provide you a certain la standard of reliability, which is not 100% 8760, and you at the state level can decide whether you want to fill in that gap or not through demand flexibility or whether you want to overbuild your system.
**Katherine Blunt**
Yeah, I mean, I think the question is how bad might it get, right? Does it get to the, to a point in which there is some sort of actual need to, I don't know, make it more dire than that, right? More so than
**Jigar Shah**
Meaning do we still need generation and transmission upgrades?
**Katherine Blunt**
I just mean in terms of what the prospect is for the need for outages in different regions. Could it get worse than what you're saying? And it is it all addressable through some combination of investment and demand response?
**Jigar Shah**
Well, I think it's very obvious that we need investment, right? We've got 250 gigawatts of solar, wind, and battery storage that have been safe harbored, right? So that stuff's gonna get built by 2030 'cause no one put all that money aside to safe harbor to not build those projects. Right. So that's happening. You've got natural gas power plants, obviously, that are happening 'cause they're sold out through 2031, so those projects Yeah. Are gonna get built, right? You've got $12 billion worth of transmission upgrades that PJM pushed through, so those projects are getting built. But I think when you look at reliability you look at Snowcrete that occurred in, what was it, January, February? Like, I've got PTSD, so I can't remember. The PJM did fine through Snowcrete, right? Like, it went. Like, we didn't have any power outages. And so my sense is that what modern technology allows us to do is to ride through this stuff, right? And I mean, the DOE at that time said we should be able to run backup diesel generators and backup natural gas generators, right, to be able to ride through Snowcrete. And in the end, they didn't need them except for maybe Maryland, and then Maryland didn't end up needing them, needing them either. So I think there's actually a lot of near-term data that shows that PJM did just fine through one of the worst crises that it could have had, right, in the polar vortex where people needed heat for 10 or 12 days in a row.
**Katherine Blunt**
Yeah. Well, we'll see what happens this summer.
**Jigar Shah**
I like it. I like it. So last question I had was, I mean, what should the Trump administration do, right? I mean so they passed this You know, like sort of where they announced this ratepayer protection pledge, right? Right. But the DOE has the right through the 2005 and 2007 Energy Policy Acts to mandate certain critical transmission lines get built. They have the right to actually unlock all those backup utility generators, right, through EPA and other things in a formal capacity. They have the ability to do all of that modeling work we just talked about, right? They could just give Gridcare or Google Tapestry or Camus Energy or ThinkLab $50 million of cold hard cash and just say, Mm run this analysis for free for all governors in the country so that they are not sitting in the dark, right? Like, what is it that the American people or the Wall Street Journal reader should expect that the US government should do with all of its national laboratories and expertise?
**Katherine Blunt**
Yeah, so what do we have going on right now? So the, with the ratepayer protection pledge and also the, gosh, I can't remember the formal name of the edict that they came out with in January trying to push PJM to do an emergency backstop auction. But there was a cost allocation piece in that as well. As with everything, I mean, this is, this is so complicated, right? It's like the government can come in and, but with the cost allocation pieces relates to the backstop auction. I mean, making sure that's overseen appropriately is at each state level within PJM, right? And like it's not, that's not within PJM's purview. Is this sort of like out of market action necessary and appropriate? I'm, I don't know. It's never been done before, but it's interesting and like the new CEO of PJM certainly has a big challenge in figuring out how to implement that within a matter of months, right? There's supposed to be no. Sorry, you can hear my dog barking.
**Jigar Shah**
I love the fact that your dog is into our conversation.
**Katherine Blunt**
So into it. PJM, yeah. But it's a great point in terms of like what should the government be supporting right now? I mean, I think that Gridcare is a fascinating company. I think Tapestry is super interesting. I think they're all interesting, and I think like trying to figure out how to expand or incentivize some of these more novel solutions to unlocking capacity in different places, I mean, makes sense to me.
**Jigar Shah**
All right, I'll put you on the spot for one last question. So with all of your extensive reporting on PG&E, and now your extensive reporting on Alphabet Google, what is the one thing that you think could be mandated, right, not just volunteered or whatever, that could make this entire situation around AI load growth easier, more palatable, more tolerable.
**Katherine Blunt**
Yeah, I mean, I don't think it's gonna be interruptible service. I don't think it's even gonna be bring your own generation. Like, I don't know. I. But maybe it's having the analysis that we've been talking about this entire time, right? Does it have to be one gigawatt? Could it be one gigawatt aggregated? Could it be 500 megawatts here, 500 megawatts there? Does it have to be five nines reliability? I don't know. What does it have to be the most extreme thing that everyone seems to be proposing? And I think certainly in at least some cases, the answer is no. So could we sort of mandate that as being even part of the interconnection study maybe? What do you think?
**Jigar Shah**
Well, I mean, I'm all in on team interruptible tariff, 'cause I think that as soon as you just mandate an interruptible tariff for everybody, then they start thinking, "Okay, what do we do to not get interrupted?" Right. "Do we do everything. Do we do the analysis differently? Do we be more honest with you about our form factors? Do we put behind-the-meter gas in, right? Do we build long-duration energy storage like Google did with Form Energy in Minnesota," right? Like, what do we do to not be interrupted, right? And you're like, great, that's on you now. Like, it, Yeah. Because you've signed an interruptible tariff. Like, I I just think that with all of the resources of the data center companies, it's shocking to me how they just shrug their shoulders and go, "I don't know. Woe is me. I don't know how to do any of this stuff."
**Katherine Blunt**
Yes. I hear you. Let's see what happens. Let's see, let's see who's the next to mandate it outside of Texas.
**Jigar Shah**
Well, with that, Katherine, it's always been a pleasure to talk with you and I know your family has a place up in Deep Creek, Maryland. My family has a place up there, so maybe I'll see you there this summer.
**Katherine Blunt**
Yeah, I can't wait. It's always great to talk to you. Hope to see you soon.
**Jigar Shah**
Take care.
**Katherine Blunt**
Bye.
ASK JIGAR
**Jigar Shah**
Welcome to Ask Jigar, a weekly segment on Energy Empire where I answer your energy questions. If you want to submit a question, the link is in the show notes. All right, let's get into it. Our first question is from Diane Boss: "You are a proponent of nuclear energy development, but I never hear you talk about nuclear waste. I'd like to hear your thoughts about that." It's an important question. For many years, we've talked about Yucca Mountain as the repository that obviously, never came to pass. Today, a lot of our nuclear waste from our, government nuclear waste, is stored in New Mexico. When we were in the Biden administration, there was a process called consent-based siting, which meant we asked people whether they actually wanted the economic benefits of storing nuclear waste at their site, and that process has led to a lot of interest from the folks in New Mexico, but other places around the country to store nuclear waste. It is a big deal. It's a great question and one that really should be talked about more. But I think we want to put this in perspective. The total amount of nuclear waste that we have in our entire country is so small that we actually can fit most of it at existing nuclear sites, which is what we do now, and it probably, I think, fills up an entire football field of space. You know, something on the order of around 15 feet high. And so it feels like a lot, but that's the amount of waste that's produced by a coal power plant every hour, from coal fly ash. And so it really is night and day in terms of how much waste is there, and the value comes from the fact that 20% of our entire grid is produced by nuclear power. And so I think there are really good solutions. I think we're on track to implementing those solutions, but agree that it is something that should be talked about more. One of the other options that people have been talking about for years, the French, of course, do it now, is recycling nuclear waste. We have been, avoiding that for a long time because President Jimmy Carter was really against recycling nuclear waste. Today, I'd say that many people are thinking about it again. There's a lot of innovative approaches to how you take the useful part of a nuclear fuel rod and put it back to work. Just to put it in perspective for you, when you have a used nuclear fuel rod come out of a nuclear power plant, less than 10% of the potential energy in that nuclear fuel rod has been used. And so it really is mostly unused. And so it is a travesty that we don't recycle nuclear waste, and I think you're seeing a tremendous amount of movement now out of the current administration, to revisit recycling nuclear waste. Next question is from William Bonagura. What's the best way for retail investors to actually support the development and deployment of clean tech? There are all sorts of green and ESG funds out there, but their impact is murky at best, and they mostly seem to invest in large cap companies that say the right things. I think there's a significant number of high income earners who'd accept slightly lower returns on a portion of their investments if that money were actually leading to projects being built. It's a great question, and one that I've tackled and wrestled with myself. So I've never invested in a green or ESG fund, and mostly it's because I agree with you. It mostly sounds like you invest in large cap companies that, talk about their green investing, and so I don't find that to be that interesting. I mostly just put money into the S&P 500 and go with low fees from Vanguard. But there've been some great opportunities recently. There was a great effort done by Greenbacker several years ago to let retail investors invest directly into funds. There have been other, areas where accredited investors could invest in angel investing or things like that. The most recent one that I've been investing in, that's interesting is a group called Energia, and Energia actually owns solar projects from other countries, so think Colombia or Brazil or parts of Africa or India, and they're actually giving people eight, nine, 10% interest on, investing in those things, and you can pull your money out at any time. And so there's a lot of people that have been putting money in there. I think they have over $450 million now of money that's been invested by mostly high net worth individuals into those types of things. But what I- One of the things I've done is I've owned solar projects on churches and schools and things like that in my community, and those have been fantastic investments, and they seem very similar to people who buy real estate and rent it out to people. And so, obviously do your own due diligence. I'm not here to give you investment advice. But I agree with you that just investing in a broad ESG fund is not the way that I prefer to invest. I like to invest in physical assets, and there's a lot of platforms now that are out there that let you invest in physical assets. Next question is from Nicholas Sweeney: Is the US equipment supply chain, transformers, switchgear, busbar, gensets, about to become the binding constraint on the energy transition, or is manufacturing capacity catching up in time? A trillion in hyperscaler CapEx doesn't matter if the equipment can't be built and delivered. This is such a good question and such a timely, question. I- look, I think that in general, many of us lived through COVID, and during the COVID period, there was a run on toilet paper. People were asking whether we were short toilet paper, and the answer is people were not using more toilet paper as much as they were hoarding toilet paper to be able to try to get through, what they thought was gonna be a period where they couldn't get any. And I think you see a similar issue happening in the supply chain. Yes, there's an increased demand for transformers and switchgear and busbar and gensets, but in general, what you're finding is a lot of people have bought these pieces of equipment before their hyperscaler data center has actually been installed, right? So Tesla famously has many transformers sitting on a lot in Austin waiting to be used in some of their future projects. Gensets are the same, right? A lot of people have bought gensets believing that their data center is going to go into service in the next 18 months, only to find out that they're delayed by three years. And so you see a lot of hoarding. So there is an increase in demand, and there is some new manufacturing coming in place to catch up. But I would suggest to you that we actually have plenty of manufacturing capacity, and we also have a lot of excess manufacturing capacity in places like India and other places you're starting to see, companies bring that capacity over to the United States to try to meet the backlog. You know, people are four years behind right now on transformers. But I would suggest to you that we're not structurally short transformers and switchgear, busbar, and gensets as much as there's a lot of hoarding going on, and I think if we really took a longer view, you're gonna see that there's probably overcapacity in these areas probably six or seven years from now. The next question is from a user on Reddit The username is spelled D-J-I-I-V-U. Thanks for your question. You're asking, "What do you think is the gateway drug to renewables, and why is it balcony solar?" I think when you say the word gateway drug, I think you're talking more about residential installations as opposed to, commercial or utility scale. Balcony solar is a great choice, and what that is there's a lot of folks who, are buying two panels with batteries and just plugging it in to an outdoor wall socket. This really originated out in Germany, has taken a life of its own, and now you've got, I think, over 15 states in the United States that are passing laws to make balcony solar legal. Utah started the craze. Virginia has passed a law recently. I think other states are starting to pass laws as well, so very exciting times. In general, I would say that what's more exciting is that there are companies who are now making manufactured solar arrays that can be placed on the ground and be drop shipped to your home with a battery installed within it, and those systems are going in at roughly $2 a watt, so 14, 15 kilowatt systems going in for 30, $32,000 all in installed. And at that price, you're at below 10 cents a kilowatt hour delivered to the home. And so I think for a long time, the United States really has been a high-cost place to do distributed renewables, right? There are just a lot of margins and aggressive sales tactics and other things that have been happening, but I think what I'm really excited about is that solar-plus-storage at least, you're seeing specific companies going to market later this year with a packaged solar system that is much cheaper than what's been offered before, really breaking that $2 a watt, mythical, area. And not only do you save a lot of money on your electric bill, but if you have a battery, then you can participate in virtual power plant programs, which many utilities are starting to offer around the country, and that means that you can get even more revenue. So some markets, you can get $10,000 over a five-year period just on, virtual power plant revenue, so super exciting times.