 I'm really happy to be here this afternoon to welcome you to this afternoon's lecture. As you said, it's part of the Powering the Future series, which we're very happy to sponsor and very happy to be associated with, and I suppose the high-level purpose of those lectures is to inform and to generate debate about the critical kind of energy issues of our time. And I suppose against that background, today's talk and today's eminent kind of speaker really fit the bill in terms of the series. Today's speaker is Professor Willis Kempton. He's a research director at the University of Delaware Center for Carbon-Free Power Integration. He's a professor at the School of Marine and Science Policy, and he's a renowned expert in the field of offshore wind generation and electric vehicle integration. The title of today's presentation, Introducing Grid Integrated Vehicles to Europe, in the presentation, Professor Kempton will talk to us about his insights gained from research and kind of practical work in the United States, in the Netherlands, in Denmark, and in the United Kingdom. The whole kind of concept of vehicle-to-grid integration is not new. It's been around for probably a quarter of a century and I suppose very simply it kind of looks at electric vehicles and it looks at their potential beyond simply decarbonizing transport, it looks at their potential to offer support to grid and grid operators to make grids more efficient, to offer ancillary services to grid operators and in particular to upper I suppose a channel to deal with the increasingly kind of significant issue of managing large amounts of variable renewable power, wind or solar on the system. So it's looking at electric vehicles in a very, very broad way, looking at all of the kind of potential and the possibilities that they offer. And I suppose it's not new, it's been around as I say for a quarter of a century, it's kind of been more on the theoretical, the fanciful side, until recent times and it's on a journey now from the fanciful or theoretical and I'm not sure those two words go particularly well together, but it is on that journey through into the practical and into the commercial realm where it's becoming offering real possibilities to kind of grid operators and the fact that it is down to the work of Professor Kempton and his associates in no small degree, particularly through the use of their offshoot venture, Nuvi, which has in Delaware, it now has the first commercially operating bank of electric vehicles offering commercial services to the local grid. ESP has benefited from Dr. Kempton's work through our involvement and his in the IEA's working group on vehicle to grid integration and we like to think of ourselves among European utilities as being proactive in this space. Going back to 2009 when we started installing the basic charging infrastructure across the country right down to our leadership among European grids in the whole area of electromobility. This is something that we see as being very important and as a utility we look forward to a time when electric vehicles are much more widespread out there, when they're making a very significant contribution to the decarbonisation of our society and in that work the work of Professor Kempton and his associates is absolutely fundamental. So without further ado I'd like to welcome Professor Kempton and I look forward as I'm sure all of you do to his presentation. So thank you. Now thank you for that very kind introduction, I appreciate it. So I appreciate being invited and I appreciate your attendance, thank you very much for coming. The title as in the announcement in advance is Implementing Grid Integrated Vehicles in Europe. So you'll probably recognise this island which seems to be moving into the future quite rapidly with very large amounts of wind already generating and an unbelievable wind resource, Scotland and Ireland I think have the best wind resources in Europe. And also a pretty well developed electric vehicle charging infrastructure for in-route charging. So these two, I see a connection between these two, right now they're doing very well each on their own, maybe a little bit faster purchasing of vehicles would be nice but the infrastructure is out there facilitating trips by those who do purchase. So I'll talk about wind and solar as variable generation which I think is more relevant from a grid operator perspective than talking about renewables, hydro is renewable, hydro is dispatchable. We're really concerned here about variable generation which means that it's not variable by when you pull the lever back and forth but it's variable by when the good Lord wants it to vary. So that doesn't always match load of course. So storage can be part of that picture but lest I oversell storage I want to be clear this is not necessarily a near-term problem. We have lots of facilities on the grid for adjusting to generator outages, unexpected outages, unexpected loss of load and those facilities also work for variable generation up to some level and it's not a, it's bright line when you go over something but Ireland is doing quite well now with 24% of electric generation coming from wind. I learned at lunch that there's something like 2% or 3% curtailment there which to me seems not so bad. If you've invested in a wind farm and you're getting curtailed it might seem bad. So transmission use of existing generators will deal with variable generation at some level. I would say up to 50% that's a very soft number. But if storage is built for other purposes it can be very low cost to use storage to help balance the grid. So the idea as alluded to in the introduction is that we use the equipment in electric vehicles that is not just the batteries but the batteries and the chargers which would like to be bidirectional. And we are taking advantage of the fact that this is an expensive resource that's purchased for private transportation in its idle 23 hours a day, typical OECD country average. It's actually more than that in Japan for example. So here's an expensive resource, it's close to the grid, it's got storage in it, it's purchased for the purpose of travel and can we do something with it the rest of the day other than that one hour when it's typically in use. If we pay for that, which we should for using something we should pay for it, then that also brings the payment to the electric vehicle owner which further lowers the total cost of ownership. So the idea here is that it's low cost storage for the grid, it enables renewable energy to provide a large fraction of electricity coming from variable generation and it also provides a second revenue stream or a revenue stream to the EV owner, that's the concept. So we call it the grid integrated vehicle, we would like that vehicle to be able to not only draw power at a variable controllable rate but also to return power to the grid that increases the value and it's not twice as much a rough, rough rule of thumb as that if you can do both ways it's worth about eight times as much and we can go into why that is and it's a different number in different national situations. So we would like vehicle to grid power as well as grid integrated vehicle, that is some grid operator whether it be the DSO or the DSO or wind farm operator, some grid operator is controlling in part the flow of power in and out constrained by the needs for travel which always get highest priority. So what's the current status of this? As Pat alluded to it's been a theory, it's been publications, it's now starting to be real with companies and actually revenue flowing, revenue is still less than cost, let's be clear but there is revenue flowing. So our EV group at the University of Delaware has experienced from ten years of implementation now with partners in other parts of the world, we created technologies to use this, there's an ecosystem of three technologies, developed, tested, patented, supported by technical staff at the university and partners who are licensing these technologies. Those partners include original equipment manufacturers that is of automobiles, auto manufacturers, tier one suppliers, the people who make the parts for automobiles, charging station manufacturers, DSOs, DSOs and so forth and these are partners some of whom are licensees, some of whom are using the technology and just selling it as a product. In several locations we are registered as a power provider that is from the TSO perspective we're a generator, you go in it's a parking lot with cars, it doesn't look like a generator but from a regulatory standpoint and from what we are doing for the electric system we're a generator and we're earning revenue in some of these markets. So the three components that I talked about that make up this ecosystem, the aggregation server, it's an essential location, it does real time operation of a set of vehicles and it provides services to an electric grid entity of some kind. So it talks to the vehicles in one direction, talks to the grid operator in the other direction and provides translations in those cases. Second is the, so the vehicle smart link goes in the car, it links the car to communications to the aggregator, it also controls charging and reports what it's doing to the server and then there's electric vehicle supply equipment, that's a standard term, that map that I showed you on the right shows electric vehicle supply equipment in Ireland. But we add some capabilities there, there's an internet portal to the car, there's of course a power connection which you have to have, there's a TSO certified meter which is maybe a little bit stepped up in certifications from what would be in a charging station where you're just charging the customer for electricity, we have a permit for interconnect. So from the TSO perspective it may be the charging station that's actually the generator rather than the car because they're used to generators staying in one place. So when you file a form for a generator you have to give an address. So that's the charging station, it's not the car. It's a grid location, that is the charging station knows what distribution feeder it's on, what substation it's behind, what the breakers are, both the distribution system and in the building, so it's a real grid location the way a DSO would think of it rather than a GPS location which doesn't tell you what you need to know. Okay, so the aggregator, not much glory here, it's this rack mount server and then the 64-bit encrypted box that we get, this one is in Delaware from our local grid operator PJM Interconnection. So PJM sends signals, they go into the server, they filter through 70,000 lines of code and then they go out to the cars coming the other direction, they get metering reports from the cars, they add those together, it's a simple addition going the other way and then we report the total power. So to PJM, to our grid operator, it looks like one big resource. Of course, if they want to audit us, they can get down to each individual car, they can put meters on each one, they want to audit and so forth, but normal operation, they're giving us a single dispatch signal, we're giving back a report of what we did. Inside the car, the vehicle smart link, found in this case, this was in the mini-E, BMW's mini-E. We found some empty space inside the weatherproof box, which is always great and then we can plug into where we want to and that small bit of electronic communication and processing can also be integrated with an off-board charger if it's a vehicle with DC charging. So obviously you have to work with OEM engineers to do this, to put anything into the car, so this tends to be a fairly lengthy process with telecoms and then, you know, they, in this case, airlift a car out to our lab and then we actually put everything together and it doesn't actually work the way we thought it would, so then we fix it. But ultimately, it does work, so there's actually a tier one supplier and OEM and then my team on the right there. So, you know, this is a bunch of cars in one location. We also have cars that are driving around and parking in different locations. So we have also developed EVSEs, as they say, part of this ecosystem. We really didn't want to do that as quite a bit of effort, but we're not able to get a manufacturer to take this seriously. So in this EVSE, we have communications that we want and we're pushing communication signals through the pilot signal. So we're not adding pins, we're not adding wireless. We always know exactly what car we're talking to. Each charging station is serialized and that serial number goes out digitally. The car has a VIN number, which it also passes back digitally. We know exactly where the resource is. There's never any question. By the way, this is a 19 kilowatt single phase charger, around 1200 euro cost. We've also got a three phase, sorry, this is a three phase one. That's a single phase one there. We also have a three phase one, which can certainly do 42 kilowatts here at 400 volt distribution, 63 amp, but also can go up to 83 kilowatts. So this is more than many DC chargers. It's under 2000 euro cost to the end user. So it just illustrates the cost effectiveness, I think, of AC charging. So types of markets, what are the markets for storage of electricity? It's not really a well-developed market yet. The markets weren't designed for storage, but you can fit storage into existing markets often. So what is the purpose of storage? Well, you can balance second by second fluctuations in the system that can't be predicted. Failures, as I mentioned, loss of load goes the other direction. You can also move electricity from hours of excess electricity to hours of shortage, so-called arbitrage or hourly market, time of day shifting. You can also hold extra electricity and readiness, much like the car is doing for the next trip, and release it only when there's an unexpected failure. Those markets are often called spinning reserves or contingency reserves. So these are markets that exist. They're not quite right, so we kind of struggle to fit into them. We usually have multiple meetings with the TSO before something gets started, but there are existing markets. There are other markets that can't be so easily fit into yet. Let's categorize this whole wider set of markets. So for the TSOs, I mentioned those. They already have typically bidding. We can bid against other providers. The DSOs do not have organized markets. They're regulated, but you could today have a bilateral agreement. You could say, well, you're going to spend all this money to upgrade your system, or how are you going to deal with high voltage on that distribution feeder that's got lots of solar panels on it? Oh, you're going to spend all that money? Well, we can do the same thing for half that much. So you could have those kind of bilateral agreements, but it's complicated because it is a regulated industry. So you would have to get regulators willing to think a little bit differently about the way those things are done, because obviously the DSO has to recover their costs. Third, possible value would be so-called vehicle to building. That is, you could maybe lower peak demand, or you could lower a capacity charge. See, demand charge, it's called in the US, it's called here. This is a charge for the highest kilowatts of the month, their highest kilowatts of the year, what's that charge called? Sorry, maximum capacity. Power capacity, yeah. Right, so if that's a short enough window, that could be lowered. That takes some tricks technology-wise, but it is possible to do. And then, of course, emergency power is what everybody talks about. You can do PV balancing if you're not getting much credit for backfading, like in Germany now. You could just absorb that power into a storage system in the car. What we have found so far is that the TSO markets are the easiest to enter and provide the greatest value several times more than these others, although some of the DSO markets may be equally lucrative. So we have, as a business, stayed mostly with the TSO markets. We're very eager to talk to DSOs who think that we can do something with them and that we can have a three-way conversation with the regulator or whatever. But you'll hear more about TSO markets, just because after looking at all these, it seems to be the most lucrative. So we think you should start with high-value markets. They're actually more challenging, technically. In some ways, regulatorally, they're more challenging. In other ways, they're not. But what we see, most other groups are kind of starting with an easy market. Well, we're going to do emergency power or we're going to do peak shifting. Okay, and then that's easier technically. You don't have to spend as much money on engineering effort and you don't make as many mistakes. Okay, that's all great. But sometimes the hardware specifications for those easy markets won't work for the harder markets. At lunch, we were talking about these fast response markets that prevent the grid from going down, valuable, may require a 10-second response time, one-second response time. We're looking at national grid may be going to one new service that would be 100 millisecond response time. So if you're going to do 100 milliseconds, you've got to design that into everything from the battery, not the batteries, but the bidirectional charger, communications to the car, the charging station, that all has to be pushing down the time. And if we didn't start with the hard markets, we wouldn't be able to go into them without making a bunch of technology changes. So that's why we are working on these more difficult, fast response markets. Okay, so we are in commercial operation today, as noted. Just quickly, their main licensees for the aggregator, sort of turns out as sort of the central piece, are Nuvi, operating in Europe, and a bit in Asia, and EVGo, operating in only the United States. So EVGo and the University of Delaware have a project in the US with the TSO, PGM Interconnection, that is operating commercially. No experimental, no trial, no test. It's just, you know, we have to follow all the rules and we get paid at the end of the month. The Nuvi and the Nu Motion have a project in the Netherlands with Tenet, and that is a commercial trial. So Tenet is enforcing the rules, but there's some rules that we're not able to follow yet. We're in discussions, it makes sense to change those rules, or are we going to have to do something different technology-wise. Third project, Nuvi, DTU, Nissan, and NL. I have a project in Denmark. There the TSO, and again at DK, is evaluating our tests. So we're demonstrating this new technology that doesn't look like anything else on their grid. Works. So we're doing extensive testing at the Danish Technical University, and they will evaluate those tests, and then they'll decide, is this a reasonable power resource or not? And then with similar partners, Nuvi, Nissan, and NL, in England and Wales with the TSO National Grid, we will be announcing a trial project May 10th, which is coming up right soon now, in London. So I'll be in London tomorrow for that announcement. Nissan has a factory in Newcastle, and right now all the cars they're producing are capable of two-way power flow back and forth to the grid if you have the right charging station, and NL is making the charging station. So that's the partnership. Nuvi knows how to create value out of this, and NL makes the charging station, and Nissan is making the cars. So let's take a look at each of these briefly. We're providing Ensler Services Regulation in PJM in the US. We have been validated for 300 kilowatts. We passed their validation tests, so we're approved to operate commercially, and then we bid in 100 kilowatt units, so we can bid 100, 200, or 300. If we add more cars and more charging stations, then we'll have to get retested for four or five or 600, whatever. So they're not used to power plants that grow over time, from 300 kilowatts to 10 megawatts, so we actually have to go back and retest every time we change our capacity. We have stationary cars as shown in the photo. We also have driving customer cars, total of about 30 EVs. All of these are bi-directional power flow. They're between 10 and 19 kilowatts bi-directional, and the revenue there is about $150 per car per month, as long as the customer keeps it plugged in most of the time when they're not driving it. If you only plug in once every three days, an hour before you need to go somewhere, then there's not gonna be any revenue. So that's what PJM looks like, the blue section there, part of the Eastern Interconnect, which is quite large. Although PJM itself is fairly large, 470 companies, 51 million people serve, peak load of 144 gigawatts. So we're a very small part of the area size. So when it first came online and was actually validated as a legitimate power generator, we got press in the New York Times and so forth, so it seems to be a new thing in the world that gets news coverage. So Denmark, we're operating here pre-commercially, as I mentioned in the testing phase now. The purpose is to establish a new resource for Denmark, which is approximately, let me forget my figures, I think it's 40, I guess it's 50% wind powered now. Don't quote that one, call Chatham House Rules on that number, which may be wrong, sorry. 44, okay, thank you. So they're very interested in storage, and the initial service that we'll provide is what's called Frequency Control Disturbance Reserves in DK2, that's Eastern Denmark. The rules are different in the two sides of Denmark. This is part of the Nordic system. They part we're in. Right now we have three vehicles being tested. There's a customer already lined up, which will have 12 EVs, and then they expand to 40, and then presumably, hopefully, we just go on and then that becomes part of the Nissan offering. You buy a Nissan, you can take a little package and sign up for these services. We'll pay, I believe the business model that Nuvi is using is that the customer gets about half of the revenue from the grid services, which should be motivating. Or they may get a free charging station, depends on the customer. Here we have to bid, in Denmark we have to bid, sorry, via a BRP, Balance Responsible Party, which so far seems to add an extra layer of bureaucracy that is not helpful at all, but we'll see, maybe it actually is helpful at some point. But anyway, those are the rules in Denmark. We're using a BRP to go to and again at DK. That's what the system looks like. This is a, some of you probably looked at this snapshot of the Danish grid. You can actually see how much wind is operating and how much is exported and so forth. And the total grams per kilowatt hour of the Danish electrical system at the time this was snapped was 230 grams per kilowatt hour, which is pretty good to run a country on. So, that's the team at, this is actually at Danish Technical University, where we're doing the testing right now. This is a Nuvi, University of Delaware, NL, and a couple of ex-InergyNet DK people. No active InergyNet person there in the photo. I'm prejudiced towards one better over another. Okay, so Netherlands. They're the grid operators' tenant. We're operating on the primary reserves market there. We've got 15,000 EVs providing service, so much bigger than the other locations. What we did there was adapt to an existing EVSC provider. Somebody's, I don't have them on this slide. The new motion is the provider there. So they already had a network and we just connected our aggregator to their network so we could talk to those 15,000 charging stations. When a car plugs in, we do some charging in some level on the top of grid services. So we have a shrunk total capacity and the customer really doesn't ever see the difference. They got an email or something saying this is gonna happen, but they don't even notice it. It's unidirectional, it's not using the full capacity, so we're not, of course, not getting this total amount of revenue we could under a different system, but still it is earning 20,000 euro per year just as sort of an add-on to their existing charging system. So it is considered a trial by tenant, so a tenant is dispatching us, or actually we're running off the grid frequency and they're getting a daily report on what the cars do, they're evaluating that and then at the end we'll talk about are we gonna continue going forward? Do we need to change something? Do they need to change the rules? So that'll be an evaluation that happens in a couple of months. So tenant has the Netherlands branch and then also it's connected. It also runs a lot of another big country next to it, so we're thinking about that, of course, at the same time. So I promised in this talk to explore policy barriers and I'm not sure how fast or slow to go through that or whether that's a discussion point. How am I doing on time? You start 10 minutes to go? 10 minutes, okay, well maybe I can go through the policy barriers a bit. Okay, so I'll just read these out. I mean, if somebody wants to break in or ask a question, please feel free. We'll sort of transition into discussion here if I'm unclear or you think I'm really missing something. So the TSO markets were really designed mostly for thermal generation, maybe hydro which is a bit faster, but it's doing really well if it responds in 15 seconds to 10 minutes, depending on what kind of iron you've got. It might be six hours or 24 hours. So they really don't kind of have a special treatment for things that can respond in one second or 100 milliseconds. So that means there's not a, oh, you guys are really saving us a lot of effort, so of course we should be paying you more. No, we don't hear that when we got walking in the door. So we think ultimately through time as markets get restructured, there will be, as has happened in PJM, partly because there was a federal FERC order, a separation of markets into faster markets and slower markets. And the faster market people bid against each other and the slower market people bid against each other. So far, the faster markets are more valuable. At some point they may not be, they may be a lower value, but we can also run on the slower markets too. So the markets were not really designed for that. Other types of ways that these markets are not designed for it is the capacity in, for example, in some, in so-e areas, including a tenant, the capacity has to be constant for the entire week, I mentioned that before. That doesn't really make sense for cars. In some cases in the Nordic system in particular, the response has to be held for eight plus hours. If we can respond in one second, we'd really like to not have to hold the same thing for hours and hours because it's a limited storage resource. We don't have a great big pile of coal. We just keep shoveling into the hopper. We've got a battery and it either empties or fills, depending on the frequencies, too low or too high. So we would like to have a certain amount of time that we'll do what we can and then some other resource comes in. And of course, if we're not holding for eight hours, we shouldn't get paid as much as if we would. We'd like to hold for 15 minutes, 30, maybe 45 minutes. But then we're just, I mean, we can hold for eight hours, okay, but our capacity is gonna be really low. We're gonna bid three kilowatts or one kilowatt instead of 10 kilowatts, which the resource is capable of. So again, we come in the market's like, how long do we have to hold? Well, as long as the frequency stays low. Oh, great. The Nordic system does not stay on 50 hertz very well. Okay, so lack of markets for DSO services, the second policy barrier actually should be indented a bit. So these should be indented under that. No, they should be indented under that. But the, no, sorry, this is the separate bullets here. So there's not markets for DSO services I mentioned. We'd like to see up and down requests balancing over time, maybe within a half hour. Not be, we have discharge, discharge, discharge, don't do anything for half an hour. But now discharge, discharge, discharge. We'd rather be balancing. We'd rather be cleaning up the last big generator's errors. So sometimes they're high, sometimes we're low. That's the perfect thing for storage resource. And whether it's a building with a bunch of batteries in it or a car, it doesn't matter. Or I pumped hydro for that matter. They all do better if it's sometimes high, sometimes low rather than you're always doing extra or you're always absorbing. So we've got regulations for electrical devices which don't yet match vehicles very well. For example, UL standard that we would put the car, subject the car to, requires that the inverter be firmly bolted to the building. So that doesn't work for a car. So just the electrical standards are not always well matched. We're working on that and that's taken time and money. Those new standards are getting done in some jurisdictions at least. So second set of policy barriers. I mentioned already, ECHEVSC is a power plant. RET plums need to register that way. For a power plant, you build for three years and you do one registration process and if it takes six months, that's okay. We'd like to do 10 units a day and register all of them. So we don't want to wait around six months or three years or whatever. Right now, there's not really storage behind the meter regulations. So we would like to be able to do net meter and you can say, well, why should you get free use of the distribution system? Okay, there's an argument there. But we're trying to serve the distribution system. So we shouldn't get charged retail when we pull the electricity in and then get only paid back wholesale when we're pushing it out. We're adjusting the frequency either way. So if there's that kind of a toll gate that only charges more one way than the other, it makes it much harder to make this a cost-effective resource and you can say, well, okay, go home. You don't have a cost-effective resource. Okay, fine. But it's gonna cost a lot more for big buildings with batteries, building hydro, building compressed air storage. So I think it's in society's interest in the electrical system's interest to not excessively toll the back and forth flow here. That's a debate we can have. That's that last bullet as well. Yep. Okay, the last set of policy barriers. At first one, we can debate that at the time when I can try. Yeah, BRP, hourly balancing, is that really relevant if you're doing fast frequency adjustment? Do you really want to be in a BRP regime where you've shifted a little bit of power from one hour to another and now you're trying to balance that? We split the meter here. The TSO cares about power response in response to a grid event or a frequency deviation. And so we've got a kilowatt meter in the charging station that's a certified meter. The DNO is still metering kilowatt hours. That works great. Any kilowatt hour, any net kilowatt hours that go through, we or the customer, car owner, pays for it. No short change in the TSO. But then we've got a separate kilowatt meter that's used for the TSO. Other ways to do it, for arbitrage, you wouldn't do that, you'd have two kilowatt hour meters maybe. So that's not a typical way to split kilowatt and kilowatt hour metering. But it makes the business case much easier. So these are all solvable problems, but there's a lot of them. We spend a lot of time in meeting rooms and in various cities around the continent and the British Isles here. And they're not gonna get solved right away. So it makes life more interesting, I guess that's the positive. So result, kind of quick summary for bullets. We're in pre-commercial and even commercial operation of grid integrated vehicles already. More locations launching. The technology definitely works. It's not a theory. Well, there's theory behind it, but it also works. And it pays for itself. You don't need subsidies, although somebody's got to invest initially because you don't have positive cash flow initially. And in some markets, it's self-supporting. So we're seeking out those markets. If the rules don't allow it, if there's a much cheaper way to do frequency balancing, for example, then we wouldn't go into that area with this technology now. And there are many policy barriers still remain. So thank you. Thank you very much. Thank you. Thank you.