 Good afternoon and welcome to today's Energy Seminar. We have a real treat with us today, a former student in the Management, Science, and Engineering Department, Ryan Ledek from the Brattle Group. He's a principal there now. I do remember Ryan as being very precocious. I think he had come from Charles River Associates and worked with some people we work with pretty closely and he came back to school. The way I remember it is those on the faculty spent the whole time he was here trying to talk him into going into the PhD program, but he was wise enough to turn us down and very productively went and now at Brattle, which is probably one of the most highly regarded consulting firms in the energy space. Oftentimes, consultants are either thought by the academics is not being sound enough on the theory side and thought by practitioners is not being practical enough for them, but I think Brattle, in my way of thinking and understanding what they do, has been able to strike that balance perfectly and that leads to his topic today, which is one of several methods people have thought of to try to get more renewable energy on modern electric rates as we now seem to want to do very urgently. If I read his bio, he's worked on all the other options and now he's going to talk to us with a clever dual meaning title about increased reliability using virtual power plants. So he'll describe to you what that means and how it fits in with all the things one could do to keep the reliability up and reduce the cost of pushing for a very high penetration of renewables on electric rates. So with that, Ryan, take it away. Great, well, Dr. Wyatt, thank you very much for the introduction. It's great to be back here. Just to give everyone maybe a little more background on the Brattle Group. So we were started in 1990. Our first office was in Harvard Square in Cambridge, Massachusetts. The first office was at the intersection of Church Street and Brattle Street and the founders of the firm wanted to name the firm after the street that our office was located on. So that gave them two options, either the Brattle Group or the Church Group and they opted for the Brattle Group. So that's little history. We have a large energy practice which is the practice that I'm in. And as Dr. Wyatt was just describing, the focus of my consulting practice is on regulatory and planning and strategy matters that have to do with all the really exciting stuff that is happening on the customer side of the electricity meter. So electrification, EV adoption, rate design, energy efficiency, demand response, rooftop solar adoption, distributed generation, all the really exciting stuff that's starting to happen when consumers start to become more active participants in the energy ecosystem is what I'm focused on. So today what I'll be talking about is a study that we released in the spring at some work that we did for Google. And they came to us and we're looking for some help really articulating the value that you can get from being able to manage consumers, utility customers, electricity consumption in a more flexible way to provide services to the power system. And so this is a summary of that study. Maybe just to kind of quickly calibrate here, just in terms of a quick show of hands, how many of you have already heard the term virtual power plant and feel generally familiar with the concept? Okay, so maybe half, maybe more. And really in my view, we're talking about virtual power plants today. It's really an evolution. This is to an extent something that's been happening for decades. Years ago it was referred to as demand response. Then more recently we started referring to it as demand flexibility. And now that we've introduced the idea that customers could push electrons onto the grid as well as consume them more flexibly, the term has now become virtual power plants or VPPs. So just to make sure we're all talking about or thinking about the same thing, just to start I'll give you a quick overview of how we've defined virtual power plants in our study. So the way we defined it is a portfolio of distributed energy resources that are actively controlled to provide benefits to the power system, to consumers, and to the environment. So what you can see in the green box here are just a few different examples of potential elements of a virtual power plant. Could be smart thermostats, electric vehicles, smart grid interactive water heaters, rooftop solar PV, distributed batteries, or on the commercial side, automated commercial building energy management systems. So the idea is to take some portfolio of those resources and then either the utility or a third party aggregator will bundle those together and control them in order to provide benefits to the power system. Could be the bulk power system, could be the more local distribution system. And by providing those benefits to the grid, ultimately the goal here is either to reduce the cost of providing power to consumers or to reduce emissions. And if those are achieved then ultimately what we should be able to do is take all of those benefits and share them between the utility, the program administrator, the participants in the program, and hopefully there's enough left over to benefit all consumers within the greater power system as well. And there are a variety of business models that exist for basically carving up that pie and determining who gets what. But that's the idea at a high level. When I presented this slide recently, some people have asked me, why isn't energy efficiency part of your definition? The definition explicitly says actively controlled resources. And if your goal with a virtual power plant say is to reduce demand during peak hours of the day, if there are energy efficiency measures they can do that very cost effectively and you know will provide you with energy savings during those peak hours, why wouldn't you include those in your definition of virtual power plants as well? And in my view that's a really good point. I think taking a broad view of virtual power plants and trying to make all of this happen more effectively is an important thing to think about. So just wanna give you a couple examples of virtual power plants that I think are kind of interesting innovative ideas to be aware of. The first is a program that's been run for a while now by a utility in Vermont called Green Mountain Power. And this is a program that's focused on deploying distributed batteries to individual households. Basically the way it works is the consumer leases a battery at Tesla Powerwall from the utility for about $55 per month, which is very discounted price relative to what they would pay if they were buying the battery themselves. And in return for receiving that discount under very specific predefined conditions, Green Mountain Power, the utility gets to operate that customer's battery in a way that will benefit the power system by relieving grid constraints or producing energy from that battery at times when it otherwise could be really expensive to generate the electricity themselves or buy it from the market. So the really interesting aspect of this program is the customer who's participating gets the resilience benefit of having that battery basically as a form of backup generation. They get to use that battery however they want during most days of the year. And then when it really makes sense from the utilities perspective, the utility gets to use the battery to reduce costs or improve reliability for all their customers. So it's a really interesting way to sort of stack all of those different benefits that you can get from one single distributed energy resource. And then on top of that, what's a really interesting feature of the Green Mountain Power program is they're specifically targeting the deployment of these batteries in this program to areas of their grid that tend to have customers with lower than average incomes and parts of their grid that have less lower than average reliability. So that's just further compounding the benefits that you can get from this really unique type of program. The other example is a program that's offered by Excel Energy in Minnesota. This is a topic I've become particularly interested in. And it's the idea of bundling this virtual power plant model with something called subscription pricing. And what Excel Energy is doing is they're going out to customers that have electric vehicles and they're saying you can basically pay a fixed monthly subscription fee for unlimited EV charging. So basically you're paying for electric vehicle charging at your house the same way that you're paying for Netflix or Spotify or any other streaming services, for example. So from the customer standpoint, that's really interesting because it's very simple, very straightforward, you can plan around it exactly what you're gonna pay. The catch is the charging of the electric vehicle has to happen during off-peak overnight hours. So in return for giving customers this simple, very transparent pricing structure, the utilities benefit is that they're ensuring that that electric vehicle charging is happening during those off-peak hours when it's less expensive to serve energy and when those EVs are less likely to be contributing to congestion or capacity constraints. So pretty interesting model and part of that subscription fee from the customer's perspective also includes a utility-owned electric vehicle charger. So that's kind of the other commonality between these two programs is they both involve the utility actually owning a resource that's located on the customer side of the meter and being able to earn a return on that investment, which is, I think, potentially a way to get utilities more engaged in this. Those are a couple examples. There are a lot of other case studies that we could talk about later on. So I think it's a really important time for us to be talking about virtual power plants because it really seems like we're at an inflection point when it comes to consumer adoption of all these different distributed energy technologies that are going to be the backbone for making all of this happen. And there are a lot of reasons that we're seeing, we're expecting adoption of these technologies to pick up rapidly. The cost of distributed energy resources is coming down, particularly as you've seen with batteries and EVs. Technology is advancing quickly, particularly when it comes to the algorithms that are needed to manage all of these different technologies in a coordinated way. The Inflation Reduction Act is reducing the price of a lot of these technologies to consumers. Something called FERC Order 2222 is designed to open up wholesale electricity markets to make sure that distributed energy resources can participate and basically compete against larger scale supply side resources in the same market. There's been an explosion of model availability of a lot of these different technologies. We've seen it with electric vehicles, but the same applies to other technologies as well. And then lastly, behind all of this is the decarbonization imperative. Consumers, policymakers, regulators, utilities, so much of what's driving their activity right now is decarbonization, that that's basically the foundation for making a lot of this happen. So the statistics that you're seeing on the right hand side of this slide, these are just various analysts' projections of the likely adoption of these consumer technologies by 2030, relative to where we are today. And what you can see is, depending on the technology, in most cases, relative to current adoption levels, the forecast is in the next few years, we could see an increase in adoption of anywhere between three times and 10 times the level at which these technologies are adopted today. So very timely to be thinking about these issues. So now thinking a little bit more specifically about our study and what we set out to do. Really what we're looking at with this study is to try and understand the extent to which virtual power plants can provide resource adequacy or capacity to the power system. 120 billion dollars has been invested in capacity over the past decade. And our view is that's likely to continue. So the chart that you're seeing on the right hand side of this slide is just showing the capacity that's been, the generation capacity that's been built in the US for purposes of providing resource adequacy to the system. It's been about 110 gigawatts over the last decade, mostly gas, peak or power plants. But recently you can see that utilities and power developers are starting to build large scale transmission connected batteries to provide this service as well. We think this need for capacity is gonna persist into the future. It's gonna be driven by a combination of electrification driven load growth, retirements of coal and gas units, and then our growing dependence on renewables, which are a great source of decarbonized energy but don't always provide a lot of capacity at the times when it's needed. So with our study we really set out to answer two questions. The first one was can VPPs provide the same capacity, the same resource adequacy that we've been getting from these traditional options? And then if they can, what do the economics look like? Are VPPs economically competitive with gas peakers and with batteries? So those are really the two questions that define the study. I'll spend just a couple minutes talking a little bit about some of the modeling that we did and then we'll jump straight into the results. So with this analysis we conducted the analysis at the utility level and it was done on an hourly level and we compared the ability of the VPP, a gas peaker and utility scale battery to contribute to reliability. So we defined our utilities having about 1.7 million customers as kind of a mid-sized utility in the US. This is a forward-looking study and we assumed that by 2030 half of this utility's load would be served from renewables. So this is a decarbonizing power system. And partly because of that because of the load characteristics of this utility, ultimately what that meant was that this utility would have a need for capacity or for resource adequacy in both the summer and the winter. So this utility now has a year-round need for resources like VPPs. And then to define the need for resource adequacy we just sort of looked at the load data of the utility and said, all right, let's assume that by 2030 this utility is gonna need 400 megawatts of new resource adequacy on its system and that could be due to replacing retirements or load growth or other things. And in order to provide that 400 megawatts of resource adequacy, the resource, the VPP, the peak or the battery needs to be able to serve in any hour all load that is contributing to those top 400 megawatts of peak demand over the entire year. And so what you can see on the chart is this is just an hourly chronological plot of this utility's net load. It's net demand for electricity. So that's the utility's electricity demand minus generation that it's already going to be getting from wind and solar resources. And you can see that in order to serve those top 400 hours of load, the resources need to be able to operate in 63 different hours of the year in order to provide this benefit. So now we'll jump into the results that we found in terms of whether or not VPPs can actually provide resource adequacy. What you're seeing on this chart is the utility's net load profile on a peak net load day in the summer. So the navy blue line is showing you what that utility's net load looked like before the VPP was dispatched. And then the dashed orange line is showing you what that utility's load looked like after the VPP was dispatched. The different colored segments are the four different components of the VPP that we modeled. As we were talking about a minute ago, there are a lot of different ways to define or create a VPP. We focused on four commercially available residential technologies. So behind the meter batteries, like a Tesla power wall, for example, managing the EV charging of the customer's electric vehicle at home, kind of like the Excel case that I was describing. Smart or great interactive water heaters, so controlling the heating element in a water heater, or smart controlling the customer's cooling and heating through a smart thermostat. So there are a couple of things that I think are important to note here. The first is if you look at the evening hours when the utility's peaking, when we simulated the dispatch of those technologies, we were able to reduce the load by 400 megawatts in any hour when that load reduction was needed. And then what you can also see is some of that load is being shifted into the middle of the day when on a utility system like this, you might have excess solar generation and actually a need to curtail solar output. So building load in those hours can help as well. This is just the peak day. What we actually found with the analysis was in order to provide this benefit over the entire year, the VPP would need to be dispatched in both the summer and the winter in seven different months of the year for 63 hours of the year, and in the case of this day in particular for seven consecutive hours. And so one question that comes to mind is, that's very different than the way utility demand response programs have been operated historically. And so one question is, can you actually get that level of performance given various constraints around, from the customer's perspective, their willingness to actually let you tinker with their thermostat or tinker with their EV charging? And the answer is yes. All of our modeling took into account what we've seen as being real world constraints around customer willingness to have the load of various end uses controlled. So just to kind of give you a sense of how we modeled this, what I'm showing here is the average charging profile of a fleet of home electric vehicles before a virtual power plant event is called and then after the virtual power plant event is called. So the dashed line is showing you what the average charging profile across a bunch of EVs would look like in the absence of the VPP. And then the solid teal line is showing you how that changes when a virtual power plant event is called. So the first thing that I'll note is, we account for limits on customer tolerance for the number of interruptions. In the case of EV managed charging because there's a lot of flexibility there, we assume that you could shift the EVs charging load on a daily basis. For something like a smart thermostat program where you're managing the customer's cooling demand, there we impose limits and I think assume that you could only do that 15 times per year. So that limit on the number of events did vary depending on the program. You also can see that when EV load was charging, it didn't drop to zero when a VPP event was called. It dropped to something less than above that. And the reason for that is typically the way these programs are designed, customers will be given the option to opt out of an event. If this is a smart thermostat program and you wanted to manage the customer's air conditioning, if they were having a party that day, they probably don't want you turning up the temperature in their home, even if it's just by a couple degrees. So customers are given the ability to opt out of these events. And we take that into account in terms of the load reductions that we model. So that's what you see there. We are also accounting for the fact that we're only looking at the actual available load. So if some EVs aren't plugged in and charging during those peak hours, we aren't assuming that every EV is plugged in and charging at the same time. You know, we're also accounting for the fact that if you're gonna curtail a customer's charging load during evening hours, you're gonna need to charge the EV after that or before it, depending on the customer's needs. So there is load building increasing in demand, increases in demand that's happening around these events to make sure that you're maintaining that level of service to the customer. And then lastly, well, the primary goal here is to demonstrate the reliability benefits of VPPs. We also, as you'll see in a second, are accounting for other benefits that VPPs can provide at the same time. So when we're modeling how you would curtail a customer's air conditioning load or how you would shift their EV load, we're taking into account not just the need for reliability, but also the benefits that that can provide to the system in terms of avoiding energy costs or reducing emissions. So all that gets factored into how we've modeled the dispatch of each one of these technologies and we have a specific set of parameters or assumptions like that for each of the technologies in the VPP. So that was kind of the story around how we modeled reliability. And now the next question is economics. What do the economics of the VPP look like compared to these other resources? So first we'll just start by showing you what the economics look like for the gas peaker. This dark blue bar that you're seeing on the left, this is the total annualized cost of building a gas peaker and running it to provide 400 megawatts of resource adequacy. So that's gonna cost $70 million per year over the life of the peaker. But when you build a new efficient gas peaker, you're gonna get some benefits as well because that new gas peaker, when you build it and you run it, it's gonna displace older, more expensive, less efficient units that otherwise would be running. So that gas peaker is going to provide you with an overall reduction in energy costs on the system. And that's what's represented by the green and teal bars. Those are benefits that you get from building this new gas peaker beyond the reliability benefit. And when you subtract those energy benefits from the cost of building and running the peaker, you're left with a net cost of $43 million per year. So that's the net cost of getting 400 megawatts of reliability from gas peakers. We did the same thing for utility-scale batteries. Here you can see that to get 400 megawatts of reliability from utility-scale batteries, costs on the cost side, things look pretty similar to the gas peaker. But because of the economics of the way batteries work and the fact that they're gonna be running more frequently than you would see with a gas peaker, in our specific case, the utility-scale batteries provided greater energy and ancillary services benefits to the system than the gas peaker did. So in the case of the utility-scale battery, we're left with $29 million of net cost to get the same 400 megawatts of reliability. So those pictures look similar with the economics looking a little bit more attractive for the batteries. Then we get to the virtual power plant. And here, the picture looks pretty different. So first, on the cost side of things, you can see that just that annual cost of building the peaker, basically getting customers enrolled in this program and paying them to be able to control their various technologies, that cost is lower than the cost of building and running the battery or the gas peaker. It's coming in at, I guess, let's call it 40 or 50 million instead of 70. Then in addition to those costs being lower, we're also seeing that you can get a broader range of benefits from a virtual power plant than you can get from the transmission-connected utility-scale battery or the gas peaker. And the reason for that is virtual power plants are located basically at the edge of the grid. And when you're controlling load, that's all the way at the edge of the grid, what that means is you have the potential to be able to avoid expensive upgrades to the distribution system and capacity investments that would need to be made all the way down to the customer's meter. So that's a big chunk of the additional benefit that you get from VPPs is that potential to defer or avoid investments in both the transmission and the distribution system. And then on top of that, if you look at kind of the purple colored bar at the top, our analysis also did take into account the social benefits of reducing carbon emissions. So in this case, we just went with a round number of $100 per ton as the social cost of carbon and it factor that into the analysis as well. And in particular, what we see driving that is the efficiency benefit of smart thermostats, the fact that they can be more easily programmed, some of them have learning capabilities that program, so the thermostats actually program themselves. We saw that those energy savings actually result in a pretty significant environmental benefit. When you stack all of those up and subtract them from the cost, including the societal benefits, you're left with a net cost of only $2 million per year associated with getting 400 megawatts of reliability from the VPP. So that's pretty remarkable. I think when we went into this analysis, I was expecting to see the economics working out for VPPs. I didn't think that we would actually see the net cost dropping all the way down to zero and essentially getting free capacity from the VPP. So just to put these numbers into context, as I mentioned earlier, we did this analysis at the scale of an individual utility. We also just wanted to think, what could this mean at the national level? So RMI, Rocky Mountain Institute, in a paper that they put out recently, estimated that there's the potential to deploy 60 gigawatts of VPPs in the US by 2030. And if we just extrapolate from these results out to 60 gigawatts of national buildout of VPPs, what we're finding is that VPPs would save between $15 and $35 billion in resource costs relative to the alternatives over 10 years. And so resource costs is just referring to those things that are actually recovered through utilities rate, building out the transmission and distribution system and buying capacity. If we also factor in those societal benefits of avoided greenhouse gas emissions, that adds in an additional $20 billion of benefits associated with VPPs. So the story here is pretty remarkable, right? I mean, that's a pretty striking opportunity when it comes to doing this. And as I mentioned earlier, everything, all of the technologies that we modeled here are commercially available. You can get smart thermostats today. You can get electric vehicles today. So the big question is, why isn't this happening? I guess before I get to that, one other quick point I'll make is we also did a lot of sensitivity analysis. That was just kind of our base case result. We did want to understand how robust those findings were. So we ran, I think, a dozen different sensitivity cases, testing higher and lower avoided costs, different assumed costs of carbon emissions and things like that, and found that these results held pretty consistently across all of those cases. There are a couple of things that I want to just highlight. So these blue bars are the net costs that you were seeing on the prior chart. The teal bars are just showing you the spread of the outcomes of those different sensitivity cases. So the first thing to highlight here is there is a point where the net cost of the VPP is higher than the net cost of getting capacity from batteries. And so the economic competitiveness of those two resources in particular, virtual power plants and batteries, is something we'll want to keep an eye on going forward because that's going to depend a lot on the specific characteristics of a given market and also the pace at which the cost of these technologies declines as we move forward in time. The other interesting thing that we saw in the sensitivity analysis was the low end of the VPP bar actually drops into negative territory. So you could actually get resource adequacy from a VPP, not just for free, but with all of these other benefits actually outweighing the cost of the VPP. Okay, so that's the positive story. And then the question is, if it's so good, then why isn't it happening? And we do need to spend a little time talking about that. I have one more slide in here. Okay, I'm gonna keep talking about the benefits for two more minutes, then we'll get into why this isn't happening. The benefits just keep coming. So there are a few other benefits that we didn't quantify in the study, but I do think are important to highlight. And just in the interest of time, I won't talk through each of these individually, but there are a couple that I just want to point to because I think they're really important considerations today. And the first is this benefit that we refer to as faster grid connection. For large-scale utility resources to come online, the resource has to get developed, you need to get financing in place, you have to actually put steel on the ground. But what has become a major barrier to bringing capacity online and meeting our urgent need for resource adequacy has been something that's referred to as the interconnection queue. And basically, utility-scale resources need to get approval to connect to the transmission system. That hasn't always been an easy process. I think some statistics that I was seeing were suggesting that that could take a year historically, maybe a little bit longer than that. We're at a point now where there's so much capacity that's trying to come online, it's gotten so complex for the transmission system operators to manage that it now on average can take about four years for new resources to make their way through that interconnection process. That is a huge lag when you have locations of the grid that have a pretty immediate need for new capacity. Virtual power plants don't face that same constraint. Basically, you can build a virtual power plant as quickly as you can sign customers up to the program. So in locations where you do have an urgent need to get capacity quickly, I see that being a huge underappreciated benefit of virtual power plants and demand-side distributed resources in general. The other benefit that I wanna describe is what we refer to as flexible scaling. Right now when we think about forecasts of the rate of growth in electricity demand, looking out five, 10, 15, 20 years, I think this is, we're probably at a point of pretty unprecedented uncertainty in our ability to predict what the demand for electricity will be over that timeframe. There are so many assumptions that are going into the rate at which customers are gonna be adopting electric vehicles, customers are gonna be adopting electric heating through heat pumps, customers are gonna be adopting rooftop solar, that I don't think we've ever seen a broader spread of the uncertainty in these load forecasts. So what that means is we need to mitigate the risk associated with that uncertainty. And when you're going out and making a decision to invest in a large scale, traditional grid capacity investment, that's a 20, 30, 40 year decision that you're making once the steel's in the ground, right? Once the resource is built, you're committed to it and now you just need to sort of live with that decision. If it turns out that your forecast was based on an optimistic expectation around the extent to which electrification would grow load and that doesn't materialize at the rate that you expected, you're stuck with having a lot more capacity than you potentially needed. And VPPs give you more flexibility because basically you can scale the growth of a VPP with growth in load. And then on top of that, to the extent that you've overbuilt with something like a VPP, like demand response, you actually can scale it back. And if you discover that you're overpaying for that capacity, you can reduce the incentives in that program. You can't do that on an annual basis because customers aren't gonna really wanna live with the uncertainty of those incentives necessarily changing every year, but you do have that ability to right size the VPP in a way that you don't with other resource types. So those are the benefits. And now we will talk about some of the challenges in making this all happen. So we, as we kind of conclude this study, we focused on I think four areas in which we wanted to define conditions under which we really would see the value of virtual power plants being maximized. And in a lot of cases, these are conditions that don't currently exist today. So the first area is market design. When it comes to market design at the wholesale level, we need wholesale markets that provide a level playing field for demand side resources, for VPPs, for energy efficiency to come in and compete with large scale supply side resources. There's a lot of work that's been done in this area already. I mentioned FERC Order 2222 earlier. That's also designed to make sure that there is, you know, apples to apples treatment of both resource types in wholesale markets. At the retail level, we need programs and rate designs that will incentivize participation in these programs in innovative customer centric ways. So the Green Mountain Power program, the Excel energy program that I mentioned earlier, those are two really interesting, innovative examples. And I think if there's anywhere that we need to see more innovation in this space, how we design the programs in a way that counts for a lot of the interesting sort of behavioral constraints and nuances of engaging customers is where we need to see some development. Then there's technology innovation. So DERs are increasingly available. I mentioned costs and prices of these technologies are coming down. We also need these technologies to be able to communicate with each other and also with the grid. And there are improvements that are being made in that area but increased standardization around communication is gonna be important as well. And then the algorithms that optimize the use of these technologies while maintaining customer convenience and comfort, that's critical as well. There are a lot of companies that have entered this space and are developing software that can manage and integrate all of these technologies. And this is an area that's changing very quickly. But those algorithms, particularly as they relate to making sure that this is a positive experience for customers are gonna be key. Policy support, codes and standards that can promote the deployment of these end uses will be really helped to sort of get them deployed. And then R&D funding that supports the removal of some of these technical barriers will be key as well. And we've seen recently the US Department of Energy being very active in this area. Some of you probably know the name Jigar Shah. He's at the DOE loan programs office and has become a real advocate for making sure that this happens and also making sure that loans and other forms of funding are available for this market segment to grow. And then lastly, the regulatory framework for myself as someone who does a lot of regulatory work with utilities at an economic consulting firm. To me, this is the key and probably in my mind the biggest barrier that has kept a lot of this from happening today is that utilities just don't have the same incentive to go out and build the VPP that they do to make conventional investments in grid infrastructure. They're in a return on capital investment to build out the grid. They don't have, typically have that same profit opportunity when it comes to paying their customers to use less. So the other key area in my mind where a lot of innovation is needed is around those regulatory business models. And I think if we can get that right and I don't think we've gotten it right yet but if we can sort of crack that code then I think we will start to see utility resource planning initiatives sort of more fully accounting for the potential benefits of VPPs. So the study that we did, there's a link to it at the bottom of this slide. It concludes with, that was a lot on the last slide so it concludes with just kind of three low risk actions that we think utilities and regulators can take today to start moving forward with VPPs. The first is to conduct a jurisdiction specific market potential study to really understand the opportunity. Those results that I showed previously as I mentioned, we ran those sensitivity cases and those findings were pretty robust but the benefits of a VPP in a market where your need for capacity is being driven by retiring coal plants, that story can look different than a market where you need the VPP because your goal is to really integrate a bunch of intermittent wind and solar generation. So getting a utility level or a state level view of the benefits is key here and then establishing some procurement targets around the results of those studies would be important. Also establishing a VPP pilot and I think the key here is a lot of times a pilot is kind of seen as a way to check the box, prove that you've done something and then kind of go back to building out the power grid. So I think two things, one is when we see utilities or other companies piloting VPPs, doing that with an assumption that it will succeed and a plan to scale after the pilot has completed is really important to making sure this happens. And then additionally, these pilots are typically used to demonstrate that a technology will work and make sure that everyone's comfortable with the technology. I think also using the pilot as an opportunity to test out some of these new business models and regulatory models that I was describing. That's kind of an opportunity that we've missed so far and is something I'd like to see as these pilots move forward. And then lastly just reviewing and updating existing policies to make sure we're comprehensively accounting for the value of VPPs are important. Sometimes we'll see a smart thermostat program being treated as an energy efficiency program in energy efficiency analysis, a demand response program in demand response analysis. And the cost of the thermostat is being counted twice in both cases rather than a single resource that's providing both of those benefits at the same time. So ironing out some of those wrinkles in the way we think about the value of these programs is key as well. So that's it. I think I'll stop there. Definitely would encourage all of you to take a look at the report. It's got a 30 page technical appendix for any of the modeling nerds like me that really wanna see the modeling detail that's behind all of this. And I guess to conclude just the last thing that I'll do is make a very shameless recruiting plug for Brattle. We're hiring, definitely would love to bring in anyone here who has an interest in energy economics, regulation, the types of issues that I've been discussing in this presentation. Find me on LinkedIn, come see me afterwards. I have been to this cards, but would love to talk to all of you about that. So I guess I'll stop there and we have a few minutes for a Q and A. Thank you. Thank you. Thank you so much. I can't believe that you were saying such an easy to understand way. I personally really appreciate that. And I was gonna ask you where the students could apply for a job, but you beat me to the punch here. So now we have about 10 minutes for questions. Do you have any student questions? Here we have a student entrepreneur right in the third row. Imagine that who works in part of this business that you talked about, Fulham. Hi, thanks for such an incredible presentation, Ryan. I have like four pages of questions. I'm like trying to choose which one to ask first. I'm glad you asked. But I'm curious to hear more about the customer profile composition of some of those results that you showed. And you talked about a couple different pricing models. Subscription and then sort of like the fixed monthly fee or something. I'm also curious to hear how customer preference for different pricing models was kind of viewed. That would be interesting. And I'm curious to hear what your thoughts are on what would make a good VPP platform? Cause when it comes to cost, a lot of the issues are, there's different sort of resources, different players who have different price and capacity beating behavior. So it makes sense to think about aggregation, but what would make for a good platform for an aggregator? Yeah, those are great questions. So I'll maybe start with the pricing piece of this. There are kind of two camps when it comes to thinking about how we financially incentivize customers to participate in these types of programs. Kind of the more kind of the purist economists will say, well just get the retail prices right, get the rates right that customers pay to consume electricity and all of this stuff will happen naturally, right? It'll happen in the most economically efficient way because customers will just naturally respond to those price signals. The other camp says customers are very, that's like that would be the economically rational approach, customers are very limited in the amount of time that they have to spend thinking about their electricity rates and their electricity bills because they have a million other things happening in their lives. So just give them the simplest possible value proposition, I'll pay you $50 per month to manage your air conditioner and then sort of take it out of their hands and don't force them to think anymore about it. People tend to be somewhat polarizing, people tend to take one view or the other. I personally, my view on this has always been that those two things can coexist. Something like a pretty simple time of use rate for residential customers where the price is, say higher during the evening hours and lower during all of the other hours and maybe very low in the middle of the night. That's something that most customers can wrap their heads around and the benefit of exposing all customers to that type of price signal is that that gives all customers an opportunity to respond and save money, right? You aren't just focusing on customers who can afford an EV. You aren't just focusing on customers who can pay a premium to buy a smart thermostat. You're setting those prices in a way that any type of behavioral response that works for the consumer will allow them to save money. So to me, that's the baseline, but then if we really want to get sort of the advanced level of control and management of these technologies that we really need to provide all of the services that I described in this study, then we do need to have more active management of some of these really flexible loads like EV charging or batteries. So that's where the second piece comes in and by giving customers that simple payment option, and then on the back end, doing the really complex work of managing the load in a way that benefits the system. That's where you can get more, additional incremental bang for your buck beyond that baseline pricing signal. Great. Other student questions? Yep, yeah, yep. You here? Thank you for speaking today. You mentioned Jigarshaw and the IRA. You mentioned Jigarshaw and the IRA. In your experience, do VPP programs and specifically VPP pilot projects qualify for EIR or Energy and Infrastructure Reinvestment loans and have they been successful in allocating money because sometimes they don't have $100 million projects? But yeah, so you're referring specifically to whether VPPs could qualify for the types of loans that like the DOE loan program's office is lending money for. So I think the short answer is yes. And one of the reasons, that you're hearing DOE and Jigarshaw talk a lot about VPPs is because they do wanna drum up interest in that part of the energy ecosystem in coming out, going out to the DOE and applying for those loans. So the answer is yes. I think you hit on kind of the key point is for those loans to make sense, particularly for utility, it needs to be a big loan because the utility does have access to other forms of capital. But I know for a fact that DOE is out there pushing VPPs in part because they do view that as being an important ingredient in the decarbonization solution. Great, we have two questions up here. We're gonna, it's all the uptime for it. Great in the middle. Hey, it was a really nice talk, thank you for it. And my question is sort of like a technical one and it's about the 400 megawatt power of the VPP and how is it calculated? Is it like the effective load carrying capacity of the VPP or is it like, so based on like what sort of parameters it's like the 400 calculated, like I'm sure the portfolio would have higher capacity but I just wanna know like what's the parameters for the calculation. Yeah, so the way it was calculated as we said in any hour that contributes to the top 400 megawatts of peak demand for this utility, the resource had to be able to serve all of that load in each of those hours. So we essentially required, with that definition, essentially required the VPP to be fully available to provide capacity and essentially give it the equivalent of a 100% ELCC across all three resource types. One last question in the middle here. I wanna say thank you so much. This was a fantastic talk. Truly really interesting stuff and especially hearing about the implementation and different proposed ideas for how that's gonna look like in the real world. My question was kind of piggybacking off the first question that was asked which was how do you see VPP programs developing in parallel with utility scale battery storage and then in the eyes of utilities are these two investment targets and priorities different or mutually exclusive for utility scale battery storage and then VPP programs? Yeah, that's a great question. I think the short answer for me is not either or. I think both resource types can provide strong benefits to the system and will be needed. I think in particular, we will eventually reach a point where we've maxed out on how much response we can get from VPPs. There are only so many customers who have smart thermostats and EVs and we typically work with utility to develop a planning study around VPPs or demand flexibility. One of the metrics that we'll report back to them is achievable potential. So what can you realistically expect to get subject to all the constraints around customer interest and participating in these types of programs? So there will be a limit on that that definitely does, even if VPPs are the most cost effective option, does open up broader opportunities for other resource types as well. I think the one thing I'll say is we haven't found a utility yet that has maxed out that achievable potential. Typically there's still a lot of low hanging fruit that exists. Great, well Ryan, thank you very much. By the way, we did pick up on your Jigar Shah reference and Chaitanya actually suggested we get him to speak next quarter. We did have somebody from that part of DOE last year which was also very popular. This is kind of boots on the ground, steel on the frame action rated things. So as predicted you did actually bridge the gap between the academics actually right at the end in response to follow this question between the academics and the practitioners and I think we need a lot more of that. So with that, thanks for a great talk and thanks to the audience for great questions. Thank you.