 This is ThinkTech. ThinkTech Hawaii at the four o'clock block on a given Wednesday. And what do we do on Wednesday at four o'clock? We always do energy on Wednesday at four o'clock. Hawaii, the state of clean energy, our flagship energy program, supported by the Hawaii Energy Policy Forum, and Sharon Moriwaki. And today we have a special—a beginning of a special month where I'm co-hosting with Mark Glick, and we're talking about HNEI. Welcome to the show, Mark. Thank you. Mark, what is HNEI? HNEI is the Hawaii Natural Energy Institute. And it was created by statute, and it's directed by statute to manage the energy systems development special fund, that's part of the barrel tax money, and also to develop and leverage those funds and other funds into building solutions for our clean energy transformation. And these two young men over here, Matthew Dubary and Mark Matsuura, they have something to do with that. They do. You know, Mark Matsuura was one of the leading engineers and the smart grid specialists at Hawaiian Electric actually was Hawaiian Electric's loss when HNEI was—and Leon Ruse and Rick Roscholo recruited him to help essentially build capacity to design these sort of unbiased solutions within HNEI. And Matthew Dubary—Mathu—Mathu. Matthew. Matthew. Matthew. Matt, am I right? Yeah. I'm good. And you pronounce your last name? Dubary. Dubary. Is one of the top battery specialists in the—or researchers in the world. And this is a form to actually allow us to present some of the most innovative solutions that are being designed, built, and exported from HNEI. Yeah, you know, welcome to the show. Matthew, Mark, great to have you here. We like to be near engineers and scientists as much as we possibly can when erupt shoulders with you. Yeah, we're good guys. Especially in energy. And the thing about it is, you know, people don't realize that HNEI is, it's a scientific organization. It's at the university. I know this because I ran into you at the university, and there you were, and you said your office was nearby, so I know I have confirmation that it's at the university, and it has been there a long time, and it has been doing, you know, science and engineering around clean energy since the beginning of the clean energy initiative in this state. Furthermore, this is a fabulous laboratory, and so this is where you want to be to do science and engineering around energy, right here in Hawaii. So these guys have a great platform to do that. It's a really important organization, and I mean, you can't even think about clean energy in Hawaii without thinking of HNEI. That's true, and you know, I haven't talked much about this yet and won't actually dwell on it today, but one of the things that caused me to want to branch out from being energy administrator for five years, the only thing that could be better than that was to actually be part of the transformation efforts from the inside. The design and also deployment. We're at that stage. We're at 27% renewable energy statewide, 54% on Big Island, and actually to make these changes, we have to integrate large rates of intermittent power, and we have to figure out the transportation solutions and the grid integration, and that has to be done in a clinical, technical, unbiased, scientific way, and I felt like it could be more beneficial to the state in trying to help organize the solutions from policy perspective, but these are the guys that are actually doing the work, and this month is about exposing that really intense and exciting work that's going on. Okay, well, that's just a great segue to actual exposure. Okay, Mark, why don't you expose your work? Try to do this in a nice way. Oh, that's right. You're not appropriate there, right? This is a family show, you know. All right, all right. Yeah, so we're, I'm part of a group within AGI called GridStardis, a grid systems technology advanced research team is kind of an acronym we put together, but we focused on research projects. So when you say laboratory, our laboratory is actually the grid many times where we install batteries on the grid. We install smart inverters on the grid, so we try to look at what are the pressing problems that are potentially impeding the progress or renewable energy in the future, and try to come up with solution sets to kind of address those problems and hopefully clear the way so things can continue to progress. So what kind of experiments, research, development are you doing? For, from a battery standpoint, which is what we're talking today, HNAIs through the funding from Office of Naval Research has installed three batteries in partnership with one of the electric companies, one's on Molokai, one's on Oahu, one's on the Big Island, the Big Island one was first. And each one's kind of looking at different issues. I guess we think that batteries maybe can help address on the Big Island. It was primarily wind energy at the time when it was installed. So we looked at what are the issues with wind power? It's kind of the fluctuations and frequency management. So we looked at wind smoothing algorithms. And then also frequency regulation algorithms and seeing, you know, which ones, what's the pros and cons to those different algorithm approaches. But we also wanted to balance the support of the grid with the help of the battery systems, which is where Matthew comes in, where we want to see if we have an aggressive response and we help the grid very much, does that just beat up the battery or it's going to die in two or three years? And Matthew's kind of helping us with that life. Where do these algorithms live? Within the battery control systems? So there's a battery and then it's a battery control system. Yes. Does that work a lot like a computer? Yeah, the control system's like a computer. I have a graphic that shows the actual battery system. Maybe we can call it up and you can. It can work maybe. OK, watch this. I wave my hands and bingo. There it is. How about that? I'm impressed. So there's the first slide and we can talk about this one if you want later. But these are all the. Second one. The next one is the actual battery projects. There you go. OK, so what are we looking at? Nice pretty pictures. So the one on the top left is on Molokai. So you see the big container is the actual battery cells and modules and things. In that one you can't really see the inverter, but the next one on the right is the big island one at Harvey Windfarm. So you see the battery modules and in front of that is the container for the inverter that converts DC power into AC power. And it also has the control system that controls the power in and out of the battery system. And actually there's control systems throughout this whole thing. So they're all kind of working together, making sure the batteries are not being overheated and it's being operated within its limits and then taking signals from the grid and then providing support when needed. So, you know, we went to think tech went to the Tesla facility in Kauai a few weeks ago. Are they using your technology? They're running separate. And by the way, they're supposed to engage another is still another battery facility over there. I'm not sure what company is for a similar but not related battery, you know, facility and just wondering are they ahead behind? Are they using what you're learning? Could they? Yeah, we've we've been in discussions with them that we haven't been directly involved with those projects. But they're quite ahead and in a sense that they've done a lot of better projects. So if you want to look at the other graphic, if you look at Kauai, they have several large battery systems. And the largest one was the most recent one. I think you're talking about is 15 kilowatts and 52 kilowatt hours. So basically able to take all of that solar energy and shift it into a time when the sun's not shining and provide energy. So kind of a dispatchable, controllable. Yeah, that's what I wanted to get to. I mean, so people understood understand what the the algorithm and the controller do for the battery and the you know, the equipment that's attached to the battery. How does it so it's not only gathering data. It's processing data. And then it's telling some other piece of equipment what to do. What does it tell it? Yeah, so that's there's there's a lot of things batteries can do. So when people say, you know, oh, you're having trouble with integrating renewables, put a battery in. Well, what is that? What do you want that battery to do? And that's where the algorithms come in. So it can it can do energy shifting like you talked about taking energy from a time when it's maybe not so much needed to a time when it's needed more. You can take frequency signals from the grid and say, you know, if there's this frequency slowing down and going to pump some energy in to prop it back up, it can look at voltage. And instead of providing what they call real power, it can put in imaginary power. And don't ask me to explain the difference between real and imaginary. But that imaginary powers helps to support grid voltage on a system. And all of those things when they when you talk about the value of energy storage, you're because it is a little bit expensive and prices are coming down dramatically, but it's still not cheap. You need to try to get as much value propositions out of it. So maybe you do energy shifting with some voltage regulation or with some regulation and try to get as many things out of that battery as can. But you can't do it all at once. So what's the purpose then of this equipment and the and these algorithms you're working out? What what are they intended to achieve, say on Molokai? Yeah, Molokai is an interesting one because over there it's a 5 megawatt grid and a lot of times they only operate with two small generating systems. So say one of those generators stop working all of a sudden or you get a fault or a short circuit on the grid. The frequency deviates very, very quickly and very deeply. And that can cause the PV panels to trip off and things like that. So the purpose of the Molokai one is to very quickly and that jump in once it sees a frequency decay or escalation to quickly jump in and try to bring it back to 60 Hertz. The problem with that is the system reacts so quickly that we had to jump in there within say like 50 milliseconds. I was kind of the target and battery systems, typically with the monitoring and algorithm calculations and then sending the command to the battery and then the battery output. All of that usually is a lot longer than 50 milliseconds. So we have to kind of reengineer that to make it a lot more quicker. Okay, Mark, you want to add something? Yeah, so and these are the critical questions that require answers that allow policymakers to set rules and regulations on how you integrate renewables more reliably. So in these cases, you know, there was a presumption while the battery is the battery fast enough. Well, it turns out the batteries has wonderful responsiveness, but it really was these other systems that have to integrate the battery into the actual grid that was discovered by HNEI's work to be deficient and coming up with solutions to actually make those controllers work better. The implication, though, is from what you said that what's happening, what you're doing, what you're learning, what you're writing in these algorithms in Molokai are different. It's different than the other islands. Different experiment, different result. So you're really using all these three facilities as a three or four, whatever it is. Very different applications. Different applications, yeah. So the Molokai one is a very fast response system stability. Like I talked about, the big island one was wind smoothing and maybe slower time frames of frequency regulation. The one on Oahu at the Camblin Industrial Park, it's on a feeder that has a very high penetration of distributed solar. So that power ramps up and down and voltages ramp up and down. So we want to have that battery kind of maybe smoothing that power so that the voltage control systems that TIKO has at the transformers don't get beat up so much. They kind of kind of take on some of that burden and maybe extend the life of some of the equipment there. When you polish this all off, Mark, what are you going to have? You're going to have lessons that can be used in all of the battery facilities, at least in the Hawaiian electric part of the state, maybe in Kauai, too. And you'll use them all to the lessons. We'll all integrate at that point. You'll be using the benefit of all of these three experimental situations. Right. So one of the missions at the university that I've learned since I moved from kind of industry into academia is you have to write papers. So I'm just joking with Matthew about that. I'm kind of a project guy. Implement the project, get it done, get it working. And then, okay, now you have to write a paper. Oh, okay. Matthew's very good at writing those papers. So we kind of lean on him a bit. So that's kind of the grid scale kind of things. We're also looking at the vehicle to grid kind of project. So where you have an electric vehicle, can those vehicles be used to help support the grid as well? Usually, they're just pulling energy from the grid to charge. You're working on that, too? Right. That's pretty exciting. Yeah. So we have a collaboration project that you talked about with Hitachi. It was funded by the Japan government through the NATO organizations, kind of like their Department of Energy. I think about $50 million worth. And they're trying to now take that research project and turn it into actual commercialized product. So again, working with Matthew is what he knows about the battery cells and the chemistries and trying to forecast if we use the car battery to support, say, grid frequency or energy shifting, how does that potentially impact the battery and how do we make a business case so that, you know, the vehicle owner's happy, the utilities owner's happy, everybody's kind of made a hole and we can further advance. So these papers that you write and will write, they're intended to inform policymakers, but they're also intended to inform the utilities, too, because somebody at the top of the utility in some engineering capacity has got to make equipment choices and connectivity choices and so forth. So it's really for everybody, you know. And it's the key thing, Jay, that if the information is viewed as credible, if it's presented in a scientific way and it's done in a peer-reviewed journal, if it's done in a major conference, if it's done on an appropriate forum, a meeting of scientists within a company or before the Public Utilities Commission, they can integrate those recommendations and make changes. Yeah, so we're all on the same page and we're all working at the best level we can work at. This may result in standards. It may result in new technologies. It may result in more efficient technologies. All of those things improve essentially the way this system works and it's being tested and proven here in Hawaii, but because we have such high rates of renewable penetration, people are paying attention to it and because HNA is an incredible player, they're also paying attention to what's being said. Okay, with that note, we're going to take a short break and when we come back, guess what? We're going to talk to Matthew Jubberry and we're going to find out about some pretty sexy stuff around batteries. We'll be right back. You're watching Think Tech Hawaii, exploring the world we live in, recognizing the changes around us and looking into the future of our lives together in these islands. Great content for Hawaii from Think Tech. Aloha, my name is Steven Phillip Katz. I'm a licensed marriage and family therapist and I'm the host of Shrink Wrap Hawaii where I talk to other shrinks. Did you ever want to get your head shrunk? Well, this is the best place to come to pick one. I've been doing this. We must have 60 shows with a whole bunch of shrinks that you can look at. I'm here on Tuesdays at three o'clock every other Tuesday. I hope you are too. Aloha. We're here on Think Tech in Hawaii, the state of clean energy today. We're talking about market gaps in battery technology as part of an HNEI series this month. The first show today is about batteries. Next time we'll do fuel cells. Time after that we'll do PV. Time after that we'll do fabulous innovation in other areas too. Mark Lick is helping me do these shows. He is the organizer of this month's study of HNEI. Our guest today, we've heard from Mark Matsuura who works with experiments involving integrating the batteries in the grid. But now we're going to talk to Matthew Duberi and forgive his French accent, sorry. I love it myself. We're going to talk about some experiments that he's doing in order to try to understand batteries, the care and feeding of batteries. We need to know a lot about that. Tell us about your work. Yeah, so my main goal is to help Mark and to help our project to understand batteries better and to be able to diagnose them when they're on the field. So it's really easy to diagnose a battery if you can open it and do a lot of stuff to it. But when they're on the field, you cannot do that. So you need to develop methodology so you can understand the battery better, control it better and make sure that everything is safe and everything's fine. And when Mark's talking about a different application, understand which one is more beneficial for the battery or not. And when Mark says a battery, for me, it's not really the case. When you talk about those BSS systems, in reality, you have thousands of batteries in there. The battery is just four volts for each of my own batteries. So if you want a megawatt-sized system, you're going to have several thousand batteries in there. Yeah, I told you, we went over to Kauai. We saw the Tesla facility and then these steel boxes. There's millions of batteries, little ones. Yeah, especially to use really, really small batteries, the same as in laptop computers. So basically, we're trying to understand how they degrade depending on the application. And we started on vehicle to grid two or three years ago, starting from the idea that the utilities say, oh, it's going to be great for the grid. But I wanted to look at, from the EV owner, is it worth it in terms of battery degradation? Meaning that you're going to get some money to feeding the grid, but is it worth in terms of additional degradation of the battery? So we did a study on that that we just published recently. And for the battery we tested, and again, it's every battery is different. So we just tested one for now. You see that the incentive will need to be pretty high for the consumer to be willing to do vehicle to grid. But we need to do that again on different chemistries, different applications. We only did one case figure. We need to do much more to really understand and help making policies on whether or not vehicle to grid can be a thing for the Clean Energy Initiative or if we need to think about something else, add more storage to the grid because people just won't do it. But part of your analysis is economics. The cost, huh? Well, not me directly. I'm trying to help them have a better understanding of batteries. Most of the models right now use black box battery models that are not realistic. And especially not for the Hawaii application where we're not using the battery the same way they might in some other area. So trying to help them understand batteries better, give them a more accurate idea of how the battery is going to degrade or fast. And so they can adjust the cost model. That's not my job. That's their job. But I really like that working. And that's why we're in China where we have experts in all those fields to be able to collaborate to have a real deep understanding on the entire aspect of the thing. On your writing papers too? Yes, I do, yeah. So you're a long term as opposed to others at this table who came from the utility. You're a long term academic researcher. Yeah, I was an academic for a long time. And I intend to try to make it stick for the next 25 or 30 years until I can tell you it's getting a bit out these days. So tell us what you've learned. I mean, because what it sounds like is that we need to know the best batteries. I mean, the best batteries and how they're made in material science and how they're connected and how they're configured and all that so we can tell policymakers what to focus on and utility companies, what to focus on. You know, one of the main problems right now is how to control those several thousands of batteries and the way they control right now is really not that good. You're going to hear about state of charge. It's the way to know how much... Is this the same control as Mark was talking about? And I'm trying to improve them because right now the algorithm in them, they're working okay, but they're not working great. And so what's happening is you keep a lot of capacity buffered on your battery so you don't go into the danger zone. But that also means that you're not using your battery at its fullest. So we actually developed some new algorithm to control battery better and we patented it a few months ago and now we're trying to validate it and it's great to have Mark and all those... So you guys work together on this. Your algorithms and your material science. Yeah, I'm actually testing a lot of algorithms on the B.S.A. system which is great to be able, for me to... When I'm testing small sayers to be able to apply the same idea on one negative battery is it's... It's a great way and that's what... The world of Chennai is great is we can take an idea from really the smallest scale. We make batteries that are that big and we can test it on a megawatt battery just a few weeks later. So it's... When you learn the truth of it... Yeah, you need to. So on one side you want to be efficient. Yeah. You want the battery to deliver. On the other side, you don't want it to deliver so much so that it blows up. Yes, and you want to also to diagnose it but you don't want to consume too much power making the diagnosis. If you think of an electric vehicle if you spend half of your capacity trying to understand the battery then you're losing half of your driving range. So it's important to have metal that are accurate but also really mad in terms of resources so we can diagnose battery without using the capacity of the battery itself. And so big part... As we discussed earlier, a big part of this is testing and measuring and having sensors and reading every little detail about that battery and how it's delivering or not. What do you use? I try to go away from sensors just because when you have 2,000 batteries in the battery pack if I call those guys and say... Too many sensors. Yeah, too expensive. So I developed method just to look at the voltage of the battery and if you're a matter of scientist if you understand the electrochemical reaction in the battery and you track them you can have a really accurate diagnosis without the need for any additional sensors, any... So you're just looking at the current you're looking at the voltage. The voltage. And then when you have the numbers from the voltage that's data and then you have a computer somewhere that is analyzing that data and giving you some conclusions. Yeah, we're trying to have that... Whether it's mostly by hand or we're going towards automation right now and the patent is one way and we have some other methods to automate that but yes, mostly that is from the matter of science and from the voltage response of the cell you can learn all you need to know about the cell degradation without having the need to destroy the cell or add any sensors. So what have you learned so far? Well, a lot of the problem of batteries come from the lithium inside. You have a certain amount of lithium in the battery and that lithium gets consumed by some parasitic reaction through the age of the battery and that's mostly why you have some capacity loss. If you have only that, it's fine. If you start losing negative electrode that's when you get at risk of lithium plating and all those kind of reaction that can be really dangerous for the battery later on. But then you can find a way to prevent that. Not really. We can track them and we can flag ahead of time. I can call Mark and say, stop the BSS right now. Change that module because we are getting into the dangerous territory where we think plating is going to start and so that module might be at risk of bursting into flames soon. But at the end of the day, that'll be automatic. Hopefully, yes. So you don't have to call Mark. You just send another device's signal and it actually works faster than Mark could do it. Yeah, I mean... One of the other things that Matthew is doing is trying to find greater value in that battery pack. So, like in vehicle batteries, they get discontinued at 80% of its degradation. So there's a lot of good use. What are some of the things you're doing there? Yeah, absolutely. And actually, I saw last week a press release from Tesla. One of the main problems is we talked about the state of charge meter and that thing gets really inaccurate with time. And so at some point, you can have the case where your battery car tell you you have 30 miles left where in rating, your car's going to stop. And I saw press release of Tesla. That happened to them. I think somebody doing LA Las Vegas and did a lot of miles and at some point, the car says, I still have miles, but it stopped. And we tried to find method to make that more accurate and the patent again is going to do that. It'd be able to have a better state of charge measurement. And by doing that, we can use less buffer of capacity on each side and we can use the cell longer because we understand it and we know exactly what happened in there. And so we can flag whether or not we can still use it or if it needs to be replaced. And that's going to be also really interesting for battery second use. And we haven't mentioned it yet, but the idea is if they could agree don't catch, maybe we can take the EV batteries and put them on the grid. But there's a lot of problems there in terms of safety that we need to know if the battery is safe or not. When it costs 20%, can you use the 80 remaining percent or is it going to blow up after a few months? So it's going to be a lot of work to develop certification process standards, as Mark mentioned. Are you going to share your results with Tesla? Are they going to share their research with you? Are you sharing your research with collaborators on the mainland? How open is this area now that you're at the front end of energy? I published almost everything and even the patent, we will publish the inner mechanics behind it at some point. Just to make clear what that means. When he says it's published, that means it becomes open record. So that's opposed to patenting or licensing a specific thing and keeping your proprietary right. So the tendency has been to actually allow anybody, including Tesla, to look at this information and benefit from it. That's great. That's great. That's what we need to do in a way. We have to be open like that and we have to be a leader. So I have one more question that I'd like to ask Mark to summarize what we've learned here today in a technical and non-technical way, however you want. So where are we in terms of finishing? I mean, this is like when, you know, when you reach the end of the internet, you know, when are you guys done? When are you going to be in a position where you can make dispositive scientific conclusions that will actually let this all unfold, let it unfold in the utility, let it unfold in the regulators, let it unfold for the public, or we can actually move ahead, do the grid that we have been dreaming about for some years. When are you going to be done? Hopefully the day before I retire. For me, we're going to get new battery chemistries almost every year. In the next decade, we're going to get probably some completely different type of batteries and in a way we're going to have to redo most of our work again, because most batteries are not going to react necessarily. Hopefully not. Okay, well good. Push off your retirement. You're a young man anyway. What about you, Martin? Yeah, I think, like I said, things are going to continue to evolve and I think the research that we do, we try to look a little bit ahead and like I said, try to clear the way so that things keep progressing, but it's not something where the, you know, things need to stop until we have some end date of our, okay, our research is done, we're closing the books here, and I'll build it. So it's kind of, as the grid evolves, as technology evolves, you know, our work is going to continue and kind of inform the process as things progress. So even after we get out there and do this grid building, we've been thinking about there's still a lot of work to be done to refine it going forward to wrap around new technologies and new demand. Okay, Mark, can you summarize all of that in one minute? It's going to be kind of hard, but I'll try my best. I mean, the truth of the matter is that the School of Ocean and Earth Science and Technology and HNEI as a key part of that research institution and part of the UH is to take some of these leading edge solutions and try to build real understanding that will help move us forward on energy transformation. Battery technologies are being currently used, as you could tell, from small individual batteries, battery packs, understanding, monitoring and coming up with a better understanding of where it sits at any given time and then also looking at grid solutions. How do you integrate it into the grid seamlessly to help deal with intermittency? So, you know, all of these things are now being practically put into place to help us make continued progress to get beyond 27% renewable to be able to get to 50% and 60% renewable. If we don't get these solutions, we're never going to be able to tamp down intermittency, wind and solar, and be able to make that reliable, safe operational power. But it is happening and it's happening rapidly and these guys are on the ground floor of that. It's very important. Batteries are so important to make this work. Well, thank you, Mark. Thank you, Mark. And thank you, Matthew. It's been great to talk to you guys. Thanks for having me. We'll be back next week with more on HNEI. Thanks so much.