The U.S. rooftop solar market has tanked. Residential applications in California, the largest market in the country, plunged 82% from May through November 2023 compared to the same period in 2022. Contractors are going bankrupt. The big culprits are high interest rates and California’s subsidy cuts. But there are some bright spots. Battery attachment rates in California have surged.
So what will it take to revive the U.S. rooftop solar market?
In this episode, Shayle talks to Jigar Shah, director of the Loans Programs Office at the U.S. Department of Energy. Jigar argues that the rooftop solar industry should reinvent itself, relying on batteries and virtual power plants. He also argues that regulations should focus on system-level dispatchability.
Shayle and Jigar cover topics like:
- The pros and cons of California’s latest regulations, new energy metering or NEM 3.0
- Learning from the mistakes of California’s Self-Generation Incentive Program (S-GIP)
- The role of VPPs and rooftop solar in meeting accelerating load growth
- Incentivizing system-level dispatchability
- How VPPs complicate the sales pitch for rooftop solar
- How VPPs could help utilities increase the utilization of infrastructure
- How to make VPPs more reliable
Recommended resources
- U.S. Department of Energy: Virtual Power Plants Commercial Liftoff
- Latitude Media: Defining the rules of DER aggregation
- Latitude Media: Unpacking the software layer of VPP deployment
- CalMatters: What’s happened since California cut home solar payments? Demand has plunged 80%
- The Wall Street Journal: The Home-Solar Boom Gets a ‘Gut Punch’
Catalyst is supported by Origami Solar. Join Latitude Media’s Stephen Lacey and Origami’s CEO Gregg Patterson for a live Frontier Forum on May 30th at 1 pm Eastern to discuss Origami’s new research on how recycled steel can help reinvigorate the U.S. solar industry. Register for free on Latitude’s events page.
Transcript
Announcer: Latitude Media, a podcast at the frontier of climate technology.
Shayle Kann: I’m Shayle Kann, and this is Catalyst,
Jigar Shah: The game of just putting dumb solar in people’s houses, providing them a loan, that game has a short shelf life. And you’ve heard both Sunrun and Sunnova, but also others announce that they are now VPP companies.
Shayle Kann: This week, the one and only Dr. Jigar Shah, exclaimer, not actually a doctor, diagnoses the maladies of the rooftop solar industry and prescribes VPPs as a solution. I’m Shayle Kann, I invest in revolutionary climate technologies at Energy Impact Partners. Welcome. Well, I think Jigar Shah probably needs no introduction, but I’ll give him one anyway. Jigar and I have been friends and collaborators for a very long time. You know him today as the Director of the Loan Programs Office at the Department of Energy in the United States. I know him as one of the founders of SunEdison way back in the day. He worked at a place called Carbon War Room. Very few of you are going to remember that one, but it was pretty interesting founded by Richard Branson. Jigar was leading that for a while. He wrote a book called Creating Climate Wealth before the term climate tech was a thing.
I think he was recently named as one of the Time 100 most Influential people or something like that, though I would certainly never talk about that to his face. Anyway, Jigar and I have been talking about climate tech stuff for a very long time. And it’s always an interesting conversation as anybody who talks to Jigar knows. This one in particular was really interesting. We started out talking about rooftop solar and what’s going on with rooftop solar in the United States. We have not spent a lot of time on this podcast recently, but it’s not a great story at the moment. It’s been a tough year or two for the rooftop solar market, particularly the residential solar market. And Jigar has long been into distributed generation back to his SunEdison days and even before. And he’s a proponent of that market, but he also has some I think, sort of wake-up calls to offer to that industry and what it’s going to need to transform into in order to resume growth.
And so it was an interesting conversation that started with talking about rooftop solar and then morphed into the world of load growth, how we’re going to serve it, and what role distributed energy resources of various kinds, not just solar, can provide. So with no further ado, Jigar Shah. Jigar, welcome back.
Jigar Shah: Thank you for having me.
Shayle Kann: All right, time to put on Dr. Jigar’s stethoscope and diagnose what’s going on in the rooftop solar market.
Jigar Shah: My goodness, there’s a lot of negativity out there right now, isn’t there? As you know, we predicted this back in 2012, 2013 during the first net metering wars. And back then we did the value of solar work in Minnesota and some of these other places, I think California and New York has done that too. I think we’re just at a point now where people recognize that what’s really of value to the electric utility company is dispatchability. It’s not the generation of power on rooftops, it’s the battery that’s actually in the home. But even in 2017, 2018, the battery seemed not that important. Attachment rates are really low, et cetera. But now you’re starting to see attachment rates that are much higher and you’re starting to see an interest from utilities to use those batteries for other purposes that they have to help manage their load growth.
Shayle Kann: Yeah, my sense was always, look, net metering in its full form where every kilowatt-hour you feed back into the grid is compensated at the full retail rate. That was never going to last to infinity. There was always going to be some point where that was going to change, and so that was inevitable. But what wasn’t necessarily inevitable was particularly in California, which is by far the largest market for rooftop solar. This is the epicenter of where this changed. It changed pretty dramatically, pretty quickly, and that changed coincided with high interest rates. And I think the combination of those two things was what led to just if you look at purely the deployment numbers of rooftop solar in California, like a cratering. But to your point, it cratered and the remaining market has a very high battery attach rate. So you could make an argument that’s exactly what the point of the policy was supposed to be. If you believe the value is in the battery attached to the solar, then you need to force a high attach rate and this is what did that, right?
Jigar Shah: Yeah, there’s a couple pieces to this. One is that I think we’ve always known that net metering has declining value beyond 5% penetration. But when you get to 5% penetration, the solar industry has so much political power that they’re able to keep things going for a little longer. And so that’s sort of what happened in California and you see that same thing playing out in Puerto Rico right now where they passed another law. And so then when it really gets to the point where you have this cost shift, then it feels abrupt even though they probably should have been eased in three years earlier. On the battery attachment rate, I totally agree with you. But the thing that frustrates me about California is that they did the NEM 3.0 piece, but there’s no transparent way of getting paid for your battery piece. So they did the peanut butter but they didn’t do the chocolate.
Shayle Kann: So talk more about that, I like the peanut butter chocolate analogy. The basic way you get compensated in a NEM 3.0 type regime, which to be clear for anybody who’s not deep in the weeds of net energy metering is basically you get paid much less for exporting to the grid than the full retail rate. You get paid closer to the wholesale rate. So the argument for a battery then, the economics for a battery, the simple version of it is basically just charge up when you’re going to have excess generation discharge that battery into the grid or into the home when prices are high and you’re doing an arbitrage. Are you saying that’s not ascribing sufficient value to the battery just doing the time arbitrage?
Jigar Shah: Yeah, I think maybe stepping back for a second, the point is what are we solving for? I’m not solving for the solar industry to continue to be able to find a value proposition for customers. What I’m solving for is that we are sitting in a moment right now where we are being told, and the data is suggesting that it’s true that we’re going from 0.4% load growth a year to two, 2.5% load growth per year. And when that happens the question becomes, what tools do we have to be able to meet this load growth? And so, one of the tools that people talk about is the fact that we have 40% rooftop solar in Australia and that we should have something more similar to that here in the United States. And so that could be one source of generation. And if we can sculpt it in a way that makes sense to the grid, then it offsets all sorts of distributions, system upgrade costs and all sorts of other things.
But we’re in this moment where we’re having a conversation about how we compensate homeowners. So NEM 3.0 basically says that if you self-consume the power, then you get paid all this money. And if you export it to the grid, you get paid very little money. So you need to add a battery to maximize your self-usage. But it doesn’t actually force the electric utilities in California or the CCAs for that matter to use the batteries to actually solve these problems. So when you think about the Sunrun 30 megawatt pilot with PG&E, I’m as excited about the pilot as the next guy. But the question becomes we don’t have just 30 megawatts of batteries in California. We have probably something on the order of four or 500 megawatts of residential batteries in California, not counting commercial batteries that are behind the meter as well. So now when does that 30 megawatts become 500 megawatts.
And in what structure do they actually get systemic compensation so that I can actually add that feature set into IQ8 with Enphase, or with the Lunar Energy System or whatever it is that the hardware and the software is that you’re funding. The question becomes, are you confident that you’re going to take NEM 3.0 and make sure that those batteries are getting paid these other value streams within the framework of California?
Shayle Kann: It’s interesting you’re saying this, and I think it’s a good point. It reminds me, I think this is a repetition of a problem we face when we craft policy and regulation for batteries behind the meter over and over again. And in fact, even in California you remember the S-GIP program, which was the program designed to incentivize commercial behind the meter batteries. I don’t know what that was, seven or eight years ago, maybe more. And the problem with that program ultimately is all the reports after the program was operational for a few years came out, were basically that the program incentivized you the commercial customer to reduce your peak demand. But your peak demand may or may not be coincident with system peak demand. And system peak demand is actually the thing that you really care about reducing from a system perspective, obviously. And this is the same thing what you’re saying here, that basically the way NEM 3.0 works is it incentivizes you to self-consume. So it says use all of your solar within your house, don’t export it to the grid.
But if you’re optimizing for what the grid actually needs, and to your point, what the grid is going to need a lot more of in the coming years because of load growth, that might be a different thing. There might be some overlap, sure, but it’s actually not pointing the gun at the right target necessarily.
Jigar Shah: No, that’s exactly right. And this is what Lon Huber and I talked about on stage at plus last year, which was the deal that he negotiated with both Solar and SEIA and Sunrun in North Carolina was exactly that. And what they want to do is systematically discharge your battery from six to eight in the morning because that’s their peak. And so they’re saying, “We’ll give you net metering and we’ll do all this stuff, but only if we have the right to control when we dispatch your battery.” But it only really makes sense if Duke gets to discharge at six to eight in the morning. If they can’t get you to do that, well then it’s not as valuable. And even when I talked to him on stage there, he said, “Well, we’re going to do this pilot of 60 megawatts or something.” And I was like, “Well pilots, the solar industry eats that for lunch. We can get that done in a month. And so how far does this go?”
And he’s like, “Look, if we truly have dispatch ability, we probably could be fine with 20, 30% of our rooftops having solar on it because now we’re using that battery capacity to actually take excess solar power that gets generated during the middle of the day because of all the solar farms that we have in North Carolina. We’re dispatching at the right times and all these other things.” And I do feel like that level of systemic conversation is something that is lacking because folks are trying to sell 45,000 systems a month and keep people employed and keep the lights on.
Shayle Kann: And I think offer a simple customer proposition. That’s the other tricky thing about these VPP type programs in general, I think, is it just makes it inherently a little bit more difficult to go to a homeowner and sit at the kitchen table and say, “I’m going to put solar on your roof and a battery in your garage and it’s going to save you X dollars because of Y and Z reasons. And if it’s because you’re going to participate in some utility program and we’re going to control your battery some of the time, but you’re still going to have sufficient reserve for backup.” This I think inevitably is something that has to get solved. It’s the same thing with smart charging EVs, they’ve got a little bit more customer friction but necessary. But as you said, the companies that are selling residential solar, they’re trying to hit volume targets predominantly, so it’s harder for them to have these projects be parts of VPPs. And I think that’s part of why it’s all these specific pilot programs today.
Jigar Shah: Yeah, look, I think that the game of just putting dumb solar in people’s houses, providing them a loan, making money on that loan, and it’s a securitization story so you get debt from Wall Street. And then you own the equity slice and then you try really hard to make sure that folks continue to find this valuable, that game has a short shelf life. And you’ve heard both Sunrun and Sunnova, but also others announced that they are now VPP companies. So it’s not like I’m the one putting words in their mouth. They’ve already suggested that publicly, even starting in probably in 2021 or 2022. And they’ve got partnerships as a result. And so the question really as you move forward though, is to what end? Because you watch on Twitter and there’s people in Texas who are on Tesla electric and they have negative $600 electricity bills.
And the question is should they be putting four batteries in their garage and overcharging the ERCOT market. My sense is that that’s not probably the most scalable solution, although it is interesting. But I think to me, the what end piece is that as our country is experiencing extraordinary amounts of growth. And for many people, if you have an existing commercial facility, let’s say, and you want to add EV charging to it because you’ve got heavy trucks and medium-duty trucks that you want to charge there, the utility is saying it might be three years before we interconnect you. Now, if you think about the technology that we have today, you can put solar on the roof of that facility and put batteries in their parking lot and you can use the grid connection you already have more efficiently. So you use it at a higher capacity factor and not have to have an upgraded service from the utility.
And so that’s where we’re headed, is that you have this ability to increase loads on severely underutilized distribution circuits. And use the battery, the solar, all these things to actually increase the utilization of those distribution circuits substantially while allowing for all those loads to get added. And not having to spend the 50 to $100 million to upgrade that distribution circuit. But the question really is whether the solar industry is in charge of those solutions, are they actually selling those solutions? They clearly are saying that our hardware has the ability to do these things, certainly Lunar Energy or Enphase IQ8 or others are saying that we have this software by which we can integrate your backup, natural gas generator and your thermostat, and your water heater and all this other stuff. And we can do that but the question really becomes like, is that their business?
Is that what the people who work in this solar industry think that their business actually is? And then as you’re suggesting, we need a simpler sales profile and there are solutions to that too. So like Rocky Mountain Power has made it very clear that they view these behind the meter batteries as a certain value, and they’re willing to pay you upfront for it. And so they’ll say, “Here’s a big rebate and now we get to do whatever we want within this service level agreement for the next 10 years. So we can dispatch your battery four times a day if that’s what it means to be able to keep our natural gas generator peaker plant offline.” Which is what they’re doing.
Shayle Kann: I want to draw out one point I think for a minute, and then I want to talk about what it takes to go from these discrete pilots to this being the state of distributed energy. The first point is all the things that you’re describing for the most part can be done with batteries that don’t have solar attached to this. And the premise of this conversation is the distributed solar industry, the rooftop solar industry is in a tough spot right now. It’s not a growing market at the moment. It’s painful. And I think you’re saying in some ways there’s potentially a bright future somewhere out there for this industry if it can just figure out how to be more than solar, and instead it becomes dispatchable, distributed energy that can help solve the problems of load growth and capacity constraints and so on. But do you think it is inevitably true that you need the generation component of that? You definitely need the flexibility. It’s clear that solar without batteries doesn’t really solve that problem. Batteries without solar might, right?
Jigar Shah: Yeah, so it is certainly the case that many utility companies would love batteries without the solar. And be willing to have that conversation with folks and say, “Here’s what your battery is worth to us and here’s how we would compensate you for it.” And certainly that’s really valuable. But I think when you think about the way in which our grid operates and the load growth that we’re facing, there’s a couple of pieces to this. So the first is to say that for many of us, because our distribution circuits have been upgraded so substantially, you’re down to 20% asset utilization. So 20% of the total capacity of that circuit is being utilized, and that’s because it’s made for everyone plugging in at whatever, 5:00 to 7:00 PM and the rest of the time it’s not being used. A little bit better post-COVID maybe, but still sort of that way.
So now you add a bunch of batteries to it and you can actually get that asset utilization up on that circuit. But at some point you might still need electrons to be added to the circuit beyond what the transformer allows for. And so in that case, you might need to add generation. And when you look at, particularly for commercial distribution circuits, you’re talking about something on the order of 200 terawatt hours of cost-effective solar that can be put on warehouses that are four acres or larger in size. And so then when you look at interconnection cues from utility scale solar and that kind of stuff, it may actually end up being net cheaper to put solar on these four acre warehouse rooftops than to put it in the middle of a field.
Shayle Kann: Or at least faster, even if it’s not cheaper. To your point on interconnection cues, there’s a decent case that you can build the aggregate same amount of capacity behind the meter faster just because of how long it takes to get interconnected at utility scale now, at least in some places.
Jigar Shah: Yeah, I think that’s exactly right. So then the question becomes when you have a system where you have a Tesla Powerwall or Sonnen system or whatever it is, and there’s an inverter that’s a part of that. And that system can be installed without solar, but I may sense is it could become solar ready, then that system has its own payback rubric, whatever that is. And now the question is how much is the solar? So now that I’ve got a battery with an inverter and it’s solar ready, how much is the solar? Today that number is whatever it is, 475 a watt or whatever it is with the inverter. But if the inverter is already paid for over here and now you’re doing solar, you could imagine that maybe you could get a lot more competition with that solar upgrade. And maybe you can get it down to 250 a watt.
Shayle Kann: Okay, so now to this question of how does this type of thing where solar plus battery maybe plus other stuff, backup generator, whatever it might be, serves as a quote unquote, “virtual power plant”, which as always to me it’s like a term that means everything and nothing at the same time. But the key point being,
Jigar Shah: You’re welcome.
Shayle Kann: Thank you for that, really appreciate it. Where it starts acting as a true grid asset that could be valued as such. And we go from pilot to this is how this stuff gets deployed at scale. What does that have to look like to you? You said earlier that this conversation is not happening, what would it look like for it to happen?
Jigar Shah: Well, I think the first thing you have to look at is what are the alternatives, because clearly this is a culture change, norm change issue. And so it is not anyone’s first choice to figure out how to flex virtual power plants. Everyone’s first choice is to do things exactly the way we do them now, which is to have utility scale generation assets, transmission assets, distribution assets, load. That is what everybody wants. It makes everyone’s life easier.
Shayle Kann: It’s interesting, that’s what your traditional centralized utility plan or whoever wants. And as you’ve said, as what your rooftop solar company wants is just to keep throwing as many systems out there as they can basically. And keep the customer value proposition as simple as possible, keep the economics as simple as possible because again, they’re going to go get a bunch of loans based on this and securitize the loans. The whole thing is predicated on a stamp and repeat model.
Jigar Shah: Yeah, for sure. But one of the things that we often acknowledge is that when you look at what has happened to electricity rates from 2019 to today, this has been the fastest the electricity rates have gone up in at least my career. And so when you look at that, the old way of doing things, the business as usual way of doing things is not working. And so we can say, well, look, on the margins this solar power is $25 a megawatt hour or $35 a megawatt hour or $45 a megawatt hour. And now we’re adding one hour, two hours of battery storage, whatever it is that we’re doing. And then on this side we’re saying we’re going to use grid modernization tools and upgrade the grid and get at least 30% more capacity out of the existing grid that we’ve already paid for. Although that even is not guaranteed, because a lot of folks would rather replace old expiring conductors with the same old technology they were putting in before. And then you upgrade the distribution circuits and you now have a transformer shortage because everyone is prophylactically increasing transformer capacity, et cetera.
And so at some point you have to suggest that we need a new approach. And that is only now I think being considered. And you saw that with New York State sort of mandating an entire docket on this recently. You saw the state of California and their desire to do dynamic pricing, which I can’t imagine is a great idea but I’ll let them do what they want to do. But then you see what LUMA is doing on their virtual power plant down in Puerto Rico. And you see a lot of other great things that are being rolled out. But the challenge I think, is that there is a lack of understanding by everybody around what it will take for this particular solution set to be dependable. The biggest challenge with the electric utilities and their regulators is they’re saying, “Well, when it’s super hot out, people don’t want their thermostats to be messed with and therefore they’re not as reliable as people suggest it might be.” And so obviously batteries are more reliable than a thermostat-based thing, electric water heaters, et cetera.
And so part of this is figuring out, okay, fine, let’s just give you that point that today they haven’t been as reliable as one would want them to be. But what would that framework be for reliability, because presumably you’re saying that it might cost a little bit more to make it more reliable to do that. And so when you think about that, there’s software layers, there’s a level of certainty that you need on the software side to make sure you know exactly what’s happening on the distribution circuit, to be able to dispatch when you need to dispatch. There’s a level of integration within the grid operations software, et cetera. And our estimates is that this would cost roughly $35 million per utility to put all the software layers in place from the venture-backed companies that have created them, to be able to actually fully integrate this whole suite.
And that has been done in pieces and parts, and in some cases in full suites for certain co-ops and folks like that. And so to me, I think that part of this is trying to unpack for people what this future might look like, how much it might cost, so that when people say, “Well, we’re ready.” We’ve got all these solutions, we’re happy to get paid to dispatch our battery. That there isn’t this false conversation where the regulator is saying, “Wait a second, I know that you’re ready but the utility actually needs to upgrade their software.” These other things need to occur for this to be a viable option. And all of those costs have to be put into the calculations before we agree that this is going to be a solution set that we’re going to roll out and abandon what we’ve been doing. Even though what we’ve been doing isn’t really working well, but we’re not going to abandon it until we trust this new solution set.
Shayle Kann: I think one of the open questions in this general space is where the fundamental control layer needs to sit and where the interface with the utility needs to sit. So as this stuff gets more complex, so one building be it a home, be it a commercial property might have generation in the form of rooftop solar, it might have more generation in the form of backup power, it’s going to have a battery, it might have smart load control thermostats, et cetera, HVAC, it might have EV chargers that are going to be controllable and controlled. And there’s this complicated ecosystem emerging of all these different players who are doing different components of that and then offering controllable versions of it. And so each of them want to go to whoever it is, the aggregator, the utility, somebody and say, “Okay, I am installing generation and a battery, and I can control it and you should make me part of this VPP.”
And then somebody else comes along and says, “Okay, well, I did the EV chargers and actually that’s far more load than anything else in the building put together, and I can control that so I should also be part of this VPP.” So do you have a view on just to make this successful large scale, does there need to be a single aggregation point at the building level or the distribution level? Or can we just figure out a way for all of these disparate loads and sources of generation and storage and so on to automatically play with each other?
Jigar Shah:
Yeah, it’s a good question and one that I think depends on where you end up with your sort of tribal proclivities. I think it is most certainly the case that if you want to backfeed power into the grid, then you will need a much more robust and expensive structure. So if you were to say… For instance, there’s a lot of people who are obsessed with vehicle to grid. If you want vehicle to grid, then you need a very expensive set of permissions from the utility and others before they’ll allow you to backfeed onto the grid. And so if you make a simplifying assumption, which is that no one is backfeeding anything into the grid, that the batteries basically use case here is to make your home go off grid, sort of an energy efficiency play. So your home now has zero load because you’re getting it from the battery.
You’re not actually backfeeding that battery into the grid. Then that simplifying assumption solves so much of the answer to your question, because once this is basically like, I promise you, I’m not backfeeding into the grid, well now you don’t need to ask the utility for permission because the utility doesn’t care. If what you’re doing is varying your load from zero to something else, they don’t really care all that much. And so that becomes a much easier framework by which to run a VPP and to do VPP whatever it is, 1.0 or 2.0, whatever we’re at now. And for me, that’s important because I don’t have all the answers to the 6.0 questions. I don’t exactly know, because you could imagine us going to a distributed sort of system operator, sort of a DSO. And then now you’re basically saying, “Okay, south of the transformer we can do whatever it is, 50 megawatts, 150 megawatts out of the transformer. And as long as we stay within the 150 megawatts we’re good.”
But that requires a level of regulatory approval, just sensors and software implementation, et cetera, which even the CCAs as far as I can understand, it have no real appetite to do. But they could do it if they wanted to but I don’t think that they’re signing up to that. So I think that the answer to your question is made a lot easier if we’re not talking about backfeeding into the grid.
Shayle Kann: Yeah, that’s an interesting way to cut it. All right, last question for you. It’s a bit of a tangent, but you mentioned it and it’s on my mind, which is you talked about how electricity prices have been going up. I don’t think this has been talked about enough actually, and this is true in lots of places but California is where it’s most dramatic. Retail prices for electricity are way up over the past few years. And I think the general expectation is with all this load growth which is not just in California, in fact most of it is outside California, the rate of electricity price increase. It’s hard to imagine electricity prices declining on a real basis over the next few years and more likely it’s going to go in the other direction. How much do you worry about that in the context of electrification, whether it be of heating or vehicles or industry or whatever? Are we risking slowing down electrification because electricity prices go too high too fast? Or do you think the train has left the station, it doesn’t really matter?
Jigar Shah: Well, I think certainly on an individual decision-making basis, I do think the train is left the station and people prefer the solutions that they’re buying today which is why they’re buying them, whether it’s heat pumps or electric vehicles or whatever it is. On an aggregated basis though, I do think that we need to be very thoughtful about who’s causing the rate increases to go up. And I think that it’s important for us to start being very pointed in the way that we talk about this stuff, because it is most certainly the case that all of the distribution grids that we’ve already paid for can handle all of the load growth that’s occurring on residential distribution lines. That is crazily simply solved.
The amount of energy consumption coming from these homes just needs to be scheduled in the right way to be able to utilize the capacity that’s already there. And if that occurs, then the capacity utilization of those circuits goes up. And by definition, the cost of what we’ve already paid for gets amortized over more kilowatt-hours sold. So on a net kilowatt-hour cost basis, the cost per kilowatt-hour should come down because we’re utilizing the assets that we’ve already paid for more efficiently. So that is clearly the case. Now, not everyone forces them to do that. And so this is negligence, in my opinion, on the part of the regulator and the utility for not forcing the features of those loads to be used. Every single car company has an ability to program when your car charges. If the utility is not integrating that feature, it’s really on the utility.
It’s not like the car companies want to wreak havoc on the utility system or to raise rates. They’re providing all these features in their car and sharing APIs with all these companies that are creating these things. There is separately an entire conversation around peak load growth as we talked about at the very beginning of this call. And it is most certainly the case that the 25,000 megawatts of data center load that’s getting attached will increase peak because why not? And so they have some solutions they could do on site, natural gas generators that they’re putting in, they’re putting in battery storage. They’re doing other things on site where they can reduce their impact on the grid. They’ve got some flexibility that they’re trying to build in. But either way, what would be a shame is if those loads were added to the grid and everyone’s bill went up as a result, as opposed to charging those people for the full impact that they’re having on the grid.
And I think if you talk to those data center companies, they want to pay for the full impact because they’re two trillion dollar companies and they don’t want to have that impact. But a lot of the electric utility companies are saying, we don’t know how to charge you a different price than we charge other loads. And so we’re having challenges with figuring out how not to tax everybody on the system because that’s what we know how to do. And so there is that particular piece there. And then for the other loads in the middle, whether you think about manufacturing loads and things like that that are occurring, most of those folks were interruptible. So most of those folks if you said to them, “I’m happy to give you a 20% discount, but what I need you to do is to actually shut down production for two weeks in a year when it’s 100 degrees outside and have everybody stay home. And if you’re okay with that, then we’ll give you a cheaper price.”
Shayle Kann: Yeah, I think we should do another conversation just on this topic. I think what we’re learning is there’s a difference between an interruptible load, which is better than nothing clearly. You could participate in a couple of demand response events a year type of thing, versus you’re contributing to peak every day. So the system’s built for system peak. And so the first order problem is solve system peak which is only a couple of times a year. But then the second order problem emerges pretty quickly, which is a peak that occurs a lot more often and you still have to solve that problem. So having a bunch of interruptible load is helpful. It’s better than nothing, but it doesn’t solve your problem if you’re adding gigawatts of new load to a grid that hasn’t seen gigawatts of new load in a year since the ’90s.
Jigar Shah: Well, and you’re right, this is probably a different conversation. But the reason why it matters for the virtual power plant piece to me, is that these are not loads that are going to get added to the grid in the way in which people think they’re going to get added. So when you think about how our electricity system has morphed since the 1970s, everyone decided to get air conditioning in the 1980s. That’s why we have natural gas peaker plants and all this other stuff. So as a result, the system capacity utilization of our system has been going down every year since the 1970s. And so when you think about all these loads that you’re saying is getting added to the grid, that does not mean that we have to add to system peak as you’re, I think, insinuating and suggesting. So I think you’re right. And so the question then becomes how best do we deal with that?
And my point to you is we have 400,000 megawatts of batteries and interconnection cues. The reason they’re stuck in interconnection cues is because the ISOs think that they have to have dedicated transmission such that they can actually charge in the middle of a peak. That’s crazy talk. And so they should all be approved next week with certain assumptions around when they can operate and when they can’t operate. And if you had 400,000 megawatts of batteries on the grid with let’s call it one hour, two hour, three hours worth of storage, that’s the entire peak of our system in the spring and the fall. And so I just think that the amount of load growth that we can handle within the actual system that we’ve already paid for is gargantuan. It can easily take us through probably 2034, 2035, by which time we think nuclear will be more cost-effective, enhanced geothermal will be more cost-effective. Some of these other 24/7 clean firm power sources will be more cost-effective. And so I think we have a really definitive choice here, and we’re not describing the situation and the problem statement correctly.
And so this really is, if you care about rates, you will figure out how to use what we already paid for more efficiently. And that is strictly virtual power plants all day every day.
Shayle Kann: All right, Dr. Jigar, nice to hang out with you as always. I think we opened the door to like four other conversations, so we’re just going to have to do this again soon. But in the meantime, thank you as always for your time.
Jigar Shah: Oh, my pleasure. It’s always good to be back.
Shayle Kann: Jigar Shah is the director of the Loan Programs Office at the Department of Energy. This show is a production of Latitude Media. You can head over to latitudemedia.com for links to today’s topics. Latitude Media is supported by Prelude Ventures, Prelude backs visionaries accelerating climate innovation that will reshape the global economy for the betterment of people and planet. Learn more about their portfolio and investment strategy at preludeventures.com. This episode was produced by Daniel Woldorff, mixing by Roy Campanella and Sean Marquand, theme song by Sean Marquand. I’m Shayle Kann, and this is the Catalyst.
