The Concrete Benefits of Virtual Power Plants — Episode 203 of Local Energy Rules

So I really started focusing on that as a staffer and created sort of leadership role for him then and helped him to write a bill maybe 12 years ago or so, 14 years ago called the Free Market Energy Act, which was a federal bill that would have put parameters around state level net metering fights and went from there to the industry. And then my tactic in this industry has always been to stay six months ahead of everyone else and then you’re kind of an expert because everything changes so quickly. And so that’s how I find myself running VPPs at Sunrun.

So just this past summer from August through the end of October, we ran a roughly 30 megawatt virtual power plant in contract with PG&E, the largest investor owned utility in the country in California, and that was about 8,500 customers with Sunrun solar and batteries. And every day for 90 days straight, we provided a two hour block of power to the utility at the end of the time of use peak period and the hour following it. This allowed the utility to help knock back their new net peak as it’s called, which is the new surge in demand on the electricity grid that happens when the time of use window peak period window ends. And really exciting about that, we went from contract to operation in about six months. You can’t build any other resource at that scale in California in six months. I don’t even know if you could build a new fossil peaker plant in California at all, even if you had a hundred years, given all the permitting and all of the siting issues and all of that.

Then we’re also running a VPP in Puerto Rico, a partnership with the utility operator there, Luma. Totally different use case. That VPP is intended to literally keep everyone’s lights on the island. The utility is very good at predicting demand, so usage on the grid but has a hard time predicting supply because the grid is pretty fragile there due to the hurricanes and then under investment over many decades. So oftentimes the utility will see supply starting to fall off. Maybe a generator is faltering and is going to go offline, and historically that meant that the utility would have to force rolling blackouts around the island to reduce demand and keep the grid energized and operational. This VPP instead calls on us to immediately push max power onto the grid as soon as possible to help the utility create a buffer and not force rolling blackouts. So you see perhaps the most rudimentary grid and operating room in our country, maybe even in our hemisphere, in Puerto Rico, plus the largest IOU, maybe the most sophisticated utility in our hemisphere, at least in our country, and we’re running VPPs in both, today. We’re not relying on utility software, we’re not relying on fancy utility meters or anything like that. This is all in-house with our own software and our own metering. Two ends of the spectrum. So that tells me we can run virtual power plants at scale anywhere in the country right now.

So in terms of the services that we can provide just on the hard utility services, there’s always some level of what’s called line loss. So when you generate power at great distances at a centralized fossil plant and then you transmit that power on the transmission system and then you step it down and voltage and then you distribute it on the distribution system, there’s line losses let’s say on the order of 8%. So you’re losing 8% of power just getting it to homes. That’s number one.

Then number two, centralized plants of any kind cannot perform local distribution functions. So that big power plant that you see that you might drive by, that can’t help reduce the need to build a new transformer or poles and wires in your neighborhood, but residential solar and batteries can. In addition to that, there’s all sorts of customer level benefits. So our customers are still receiving backup power from the battery site in their homes. We never deplete the batteries down a certain percent so that customers always have backup power even if we’ve used the energy for a grid service event. Number two, we pay our customers for participating in virtual power plants. I never receive a check from my utility because they dispatch the fossil fuel power plant. Number three, virtual power plants tend to be cheaper, provide the service at a cheaper level, which means that we reduce costs for all of the rate payers on the utility system. So not just the person that has the solar batteries and is participating in the program, but also all other customers on the grid.

I’ll say number three and then I’ll stop there because I know this is a long answer, but number three, we’re selling a service, not an asset, which means the utility and or the wholesale market operator can sort of price this service. That’s very different from if a utility goes to the utility commission and says, Hey, let us spend, I don’t know, 10 billion on a nuclear power plant to build that asset and own it in perpetuity and then hopefully we’ll be able to have electricity generated from it. So very different when a utility procures a service and not an asset.

That’s fascinating, right, and the reason it’s only a doubling is because electric motors are so much more efficient than their internal combustion counterparts. So we’re going to double the kilowatt hour usage on the grid as we electrify. Saul Griffith has also modeled that if we put solar on the built environment in all available spaces, so every rooftop that’s a good candidate, carports, et cetera, we can meet about half of that double demand that we’re going to see in the future. So for me, this means a couple of things. One is that due to the lack of barriers that we see for residential solar and batteries that I mentioned earlier, we need to start there because obviously no regrets and we’re going to build out very quickly much of what we need to meet the demand on the grid, but then there’s also clearly an opportunity for utility scale resources and other clean energy resources.

So I don’t think we should be having a binary fight about this. However, if we double the kilowatt hour usage over the year on the grid and we also double the kilowatt demand on the grid in any hour, day, week, month, or year, if we double that demand, we’re going to have to build out the poles and wires 2x. We can’t do that. That is insanely expensive, and what will happen is it will actually increase electricity costs dramatically, thereby slowing down or stopping electrification. So how do we solve for that? Well, virtual power plants are actually the only way to solve for that. Flexible demand side solutions, smart charging of your car using your stationary battery to sort of modulate your demand. All of that together in concert is the only way that we can keep costs down as we increase the electricity demand on the grid.

It feels very much like we’re, as you said, kind of like in the 10 years ago conversation about whether or not rooftop solar or utility scale solar is cheaper, and I’ve done some analysis that suggests they’re about the same, depends a little bit on where you are, but very little of the analysis seems to be talking about this kind of future about managing the grid demand. Let me pivot a little bit though and just ask you about how do virtual power plants, obviously they can address issues like affordability because you just addressed this issue of grid infrastructure needs and being able to handle that. You talked about people who participate in a virtual power plant getting paid by the utility for the service that they’re providing, which is pretty awesome. How does it help with climate change, which is obviously one of the driving factors behind electrification and behind folks’ interest in clean energy?

That means that three different homes in that neighborhood are using stored solar power after dark just because one of those homes is participating in the virtual power plant. So the other side of that coin is those three homes are not creating demand for fossil fuel power plants at that time because their demand is being met by local resilient solar and batteries.

So the question is almost moot because it’s like what is being built today and what is working today? Let’s double down on that first. Separately, we should have the conversation about transmission obviously, but again, not a binary thing. And let’s take the no regret step now that we have that we know it works. The other way to look at this is any demand that you reduce on the distribution system or on the transmission system means that that capacity is then freed up for some other service. So if your goal is to bring large scale renewable hydropower, well, I won’t say renewable because there’s debate about that. If your goal is to bring large scale hydropower from Hydro Quebec through the transmission system to the Boston Load Center, if more and more demand is being met locally, then that means there’s more available capacity on the existing lines or the new lines don’t need to be built out quite as big.

So what that means is the rules just simply do not exist and the markets don’t exist for virtual power plants to participate unless we go out there and create them. So you have very forward-looking progressive utilities like Green Mountain Power in Vermont that has been running virtual power plant programs, both utility owned as well as competitive ones, for nearly a decade now. And then we have partnerships with PG&E and others where we’ve gotten something across the finish line, but we are still looking towards the future of what does that partnership look like? And then you have places like Puerto Rico where they have no other option and they’re seeing that they need to rely on virtual power plant services today. But going to be honest with you, I’m based just outside of Portland, Maine, so we have a saying it’s still tough sledding and we have to fight these battles in the policy and regulatory spheres for some number of years to come, I think, in order to create these opportunities.

Hey, thanks for listening to Local Energy Rules. If you’ve made it this far, you’re obviously a fan and we could use your help for just two minutes. As you’ve probably noticed, we don’t have any corporate sponsors or ads for any of our podcasts. The reason is that our mission at ILSR is to reinvigorate democracy by decentralizing economic power. Instead, we rely on you, our listeners. Your donations not only underwrite this podcast, but also help us produce all of the research and resources that we make available on our website and all of the technical assistance we provide to grassroots organizations. Every year ILSR’s small staff helps hundreds of communities challenge monopoly power directly and rebuild their local economies. So please take a minute and go to ilsr.org and click on the donate button. And if making a donation isn’t something you can do, please consider helping us in other ways. You can help other folks find this podcast by telling them about it, or by giving it a review on iTunes, Stitcher, or wherever you get your podcasts. The more ratings from listeners like you, the more folks can find this podcast and ILSR’s other podcasts, Community Broadband Bits and Building Local Power. Thanks again for listening. Now, back to the program.

Distributed energy resources are very good for the grid and for people. They reduce costs, they address climate change, but utility business models make them very resistant to it. And as such, I don’t know if you want to talk a little bit about this, but there’s not a lot of places where you can do this, unfortunately, despite how valuable it could be for everybody who’s on the grid,

So every state in New England has some version of a virtual power plant program generally called Connected Solutions. There are a lot of nuances here, but I’m just going to speak in broad strokes. Generally, Connected Solutions is run out of the energy efficiency budgets that the utilities have, the utilities get paid to administer those energy efficiency budgets. Then for the virtual power plant programs that are run through those energy efficiency budgets, we use those VPP programs in partnership with the utility to knock down a cost driver on the system that the utility is apathetic towards. It’s just a pure pass through from the wholesale market down to their customers so they don’t lose money when that cost driver goes down and they get paid money through the efficiency budgets to administer the programs. Obviously New England is part of a unified wholesale market, ISO New England, the cost driver that we’re knocking down is generally the cost of the capacity market, which is a pass through from the market through the utility to the customers.

But this can also work in vertically integrated places, even in Minnesota. Just figure out what the budget is there that the utilities are already recovering on to operate, likely the energy efficiency budget, and slot the virtual power plant programs in there and then just create a shared savings mechanism, right? So for every a hundred dollars that are created in savings, utility is able to pocket, I don’t know, let’s say eight to 10%, just like their guaranteed rate of return. The rest is then split between participating rate payers of virtual power plant participants and then the leftover savings are spread across the rate base, so everyone’s bills go down. It’s really not rocket science and those programs in New England, we’ve been running those in partnership with utilities for six or seven years now, year in, year out, tried and true. It’s proven. So I said earlier, there are no technological barriers to this, no technical barriers. There are also no business or program model barriers. The only barrier is putting those programs in place.

The other piece here is talked a lot about stationary batteries, Tesla Powerwalls, LG Chems, et cetera, being the backbone of virtual power plants. And that’s true today and I think that will be true out into the future. But increasingly we’re also seeing exciting stuff happening with electric vehicles both on the managed charging side, which I wouldn’t necessarily describe as a pure virtual, that’s really just demand response, but also increasingly on the bidirectional side, so electric vehicles that can push power from the vehicle to the home and to the grid. And of course Sunrun, we have a partnership with Ford on the F-150 Lightning and through that partnership we were the first to market with a true bi-directional electric vehicle offering in the U.S. which we unveiled I guess maybe two years ago now. And so I would say keep a very close eye on that space as well because I think you’re going to see an explosion of both bi-directional vehicle offerings as well as programs for bi-directional vehicles to form virtual power plants with those.

Let me ask you a serious question actually though. You mentioned Hawaii and then you said Hawaii had gone off the rails, and I was particularly interested because earlier in the podcast you had talked about Hawaii when we were talking about business models. Hawaii is like this one state that has made that full transition to performance incentives for the utility. The order first just came out a couple of years ago, so I know that it’s early yet. I was just curious if you could talk a little bit about what Sunrun’s experience has been in Hawaii and if that business model change is actually helping, or I guess what I’ve read, what I’ve read is that there was a particular program that was working really well and they decided to completely get rid of it around storage. So maybe it’s not working as well as we’d hoped.

However, I think that a lot of the virtual power plant action in Hawaii has gotten overly complicated. You have all these competing dynamics there now between what value you get for electricity that’s exported to the grid outside of a virtual power plant program. Then what value you get if you are exporting during a virtual power plant program, then how is that value captured? Is that bill savings? Is that paid out to an aggregator or directly to a customer? Then there are heavy, heavy complexities around the software connections, right? What software, what type of protocol do you need to connect to a utility derms? Should you even need to connect to a utility software platform or can you do that in-house with an aggregator? What level of dispatch is required, how many calls? And then just at the end of the day taking out your calculator and looking at all of that and saying, is the customer better off if they participate in this virtual power plant program or are they not? And a lot of times the answer is they’re not better off, they’re just simply not better off. And so I think it’s just gotten a little overly complex in Hawaii, and I like to say complexity kills and simplicity sells, and that’s true both when it’s an aggregator providing virtual power plant services to a utility or that same aggregator trying to get customers to sign up for solar and storage and be in a virtual power plant program.

So I think it’s a high level answer, but the truth is it’s just gotten a little too complex and they’ve lost sight of where they’re trying to head.

And so our role at Sunrun as the aggregator, as the VPP provider, is to manage all of that complexity on the backend for the customers. They don’t even have to think about it. They just get a check at the end of the year for participating. And then manage all of that complexity for the utility so the utility doesn’t have to interact with 8,500 customers and respond to customer inquiries when their battery isn’t dispatching the way they thought it should or when the check didn’t come through the mail and the way they thought it would or whatever. So that’s our fundamental value proposition is managing that complexity from both sides to deliver that service to the utility and that value to the customer.