 This is Wednesday. We know what day that is. It's Hawaii, the state of clean energy day. I'm your host, Mitch Yuan. And this show is sponsored by the Hawaii Energy Policy Forum with funding from the Hawaii Natural Energy Institute. So today I've got some new technology made in Hawaii. In fact, made at HNEI, the Hawaii Natural Energy Institute. And I'm very happy to have Dr. Kevin Davies as my guest today. And Kevin is going to talk about his new invention, what problem it's trying to solve, and where he is in the technology development phase. So welcome to the show, Kevin. It's nice to have you. Yeah, thank you, Mitch. Good to be back. So, Kevin responded in a very short notice. Thank you so much for pulling all this together in about 10 hours. I just saved my life. That was really great. So Kevin, let's talk about what's the overall problem that we're trying to solve here and that you've been working on and HNEI has been working on for many years. Yeah. So as we know, Hawaii is trying to progress to 100% renewable energy. And more specifically, we're seeing lots of distributed resources, particularly rooftop PV. There's a whole lot of people that want to have rooftop PV on their houses. The utility is struggling to accept all that. And that's the general framework that this fits under. More specifically, and this is my first slide of introduction here, is the challenges of grid integration of PV. At the distribution level, which is kind of the local neighborhood level, there's really two major issues with integrating distributed PV, voltage variation due to the bidirectional power flow. The grid was designed for power to flow one direction, because that's traditionally what has happened. The utility produces energy, the consumers consume electricity, it always flows, power flows one way. But now with PV, power flow is going both directions, and that creates voltage variations more than we're originally designed for. The thermal limits of the wires and transformers, the infrastructure, the actual equipment that's in the neighborhoods is only designed for certain power levels. And now that we have large amount of PV, in some cases, PV, power flowing backwards at a larger magnitude than it was even flowing forward in the original intent. And so that can stress certain infrastructure and sets limits on how much PV we can integrate. And those are both at the distribution level, like the smaller neighborhood level, at the system level. Let me ask you a quick question. I want to interrupt, I'll interrupt you every now and then. So like how close are we to kind of maxing out the grid right now? I mean, you know, when people first started like eight or nine years ago and when the big rush for PV came on, there was a quote, lots of room on the grid. So how are we doing in that department? Well, there are quite a few feeders. I don't have the statistic in front of me, but there are quite a few feeders, let's say on Oahu, that are already, you know, at 250% of their daytime minimal load. That means that the daytime minimal load, literally like the amount of minimal amount of electricity that customers are going to use on that feeder during the daytime when they're generally away from home COVID, that amount, 250% of that amount is what the PV is rated for. Like if you, the nameplate PV that people have on the houses in some cases is up to 250%. That's been a limit that's been set, it's sort of a threshold, but there are feeders in Hawaii that already reached that limit. And then there are issues where feeders are voltage limited, where the utility has to, by regulation, meet the standard voltage at your outlet has to be 120 volt, plus or minus 5%. If it's outside of that range at the point of connection to the grid, they're not serving their needs. And we already are seeing that as a limitation to PV hosting capacity. So you also see progression in Hawaii away from obviously net metering as long in the past, but, you know, even towards, you know, no export like customer self-supply where you aren't exporting electricity to the grid, you're only serving your needs. So that is to protect against some of that say, you know, it also is at the system level. So I can't say all that's a distribution level, but that is the overarching concern is that if, you know, obviously if it were to continue with net metering forever, you know, the grid wouldn't have been able to handle it as designed, right? So I don't have the statistic in front of me of how many feeders are voltage limited, thermal limited, but it's significant. It's like hundreds in Oahu. Okay. So it's a real problem. And the fix, unless you come up with some innovative stuff, like H&I yourself have come up with, would be quite expensive, like in order, if we wanted to have more PV or otherwise, hey guys, sorry, we can't put PV on your roof, which was something we don't want to have happen. So I'll stop interrupting now. Yeah. Well, yeah, that's a good setting. We'll get obviously to the solutions, but the potential solution, but I wanted to first cover at the system level, right? That's at the bulk level, the whole island in our case. You know, three major things that happen with large amounts of PV and really, in general, any renewable energy wind included is energy. There's a concern with energy shifting to meet demand. The utility has been built up around dispatchable power, like power that can be increase or decrease on demand as needed to meet supply and demand to meet, right? Because if supply and demand don't match, then good frequency goes out of spec and we have big issues, right? So, you know, you need flexible resources that you can change on demand. Well, PV and wind aren't like that for the most part. Like they, you know, PV and wind go up and down depending on weather. You can, in general, you can curtail PV. You couldn't tell a PV inverter to not export all the power that it's capable of at the moment, even though the sun is shining doesn't mean you have to use all that potential power and exports the grid. There's concern there where, first of all, that's only on the downward motion and downward direction. You can't, you typically are going to be by default exporting all the power you can. So there's no headroom to go off, typically, right? You know, the other aspect is distributed resources like rooftop PV. Utility doesn't have any direct control of that to, to, to limit the power export. It is what it is, according to the tariff that you're connected to, right? But they're not like they can send a request immediately at, you know, the disconnect or anything. There's, I think, a trend towards more, more intervention, but it's a long way off to have some nuance control that the utility is used to for, you know, fault generation. So there's that concern of energy meeting demand as well, just, just at a bulk level, you may have heard of the duck curve or the Loch Ness monster curve or whatever, right? So that generally summed up is that, you know, demand on, well, first of all, PV output peaks at midday, right? Load, load, load peaks in early evening when people come home. So that used to be, you know, utilities had to deal with the peak in the evening. There was already infrastructure and plans to do that, that was handled, right? But now it's sort of a double-edged thing where not only is a peak increasing because load increasing, but the sun's going down and your production's decreasing. So it creates quite a ramp rate that the utility has to have flexible generation to manage, right? You know, so that's a concern with more PV. Increased variability and thus greater need for operating reserve. So utility is used to managing unexpected things, right? Like the grid has to have a high-level reliability. And one of the ways that that's accomplished is maintaining reserves. So for instance, utility keeps one reserve the capability to cover for any large generation plant to go offline. And they do that by keeping the other plants operating less than what their maximum capability is, right? And at any moment they could be one short order, they could be ramped up and cover and the grid would hiccup a little bit, not go out, right? And so that's now on top of the variability of, let's say, a contingency of a generating plant going offline and also variability of load of people changing their activities and load changing. Now you also have variability of wind and solar, right? So additional variability and therefore likely greater need for operating reserves that has a cost associated with it. Because these generation plants are designed, they're actually typically the most optimal, the most efficient when they're running full bore, like all the power. But now you said, oh, we have to back them off a little to have this extra headroom. Well, there's a cost to that, right? And it's big dollars when you consider the big dollars that Eco's paying for fuel, right? So that's a concern. The other concern, and I'm sure there are more, but I wanted to highlight the major ones, is frequency regulation with smaller system inertia. So traditional generation is synchronous rotating machinery, right? So big industrial heavy equipment that has a lot of rotating mass, right? You're not going to stop that reading in mass immediately. It's like a low commotive engine, right? So whenever a load, additional load hits the grid, you know, increases due to whatever happened, some disturbance, that inertia is immediately just physics based, no controls. It just is, it hits the grid and that extra energy slows it down a bit, but there's so much bulk behind it that it can ride through a lot of things, right? Before controls have to then, controls are a bit slower and then they can adjust, right? But the thing with PV and wind is it and batteries is they are inverter based, right? At least modern turbines are. And so there's not that physical inertia. There's no direct sort of rotating mass, right? So that's an issue and somewhat an opportunity, but an issue that needs to be addressed, right? Well, let's have the next slide. That was a great explanation, Kevin, because that's something I didn't actually understand. And the way you described that made it quite understandable, particularly the thing about inertia and the difference between a rotating machine and sort of the electronics inverters are basically electronics. They don't spin, but they have no inertia. So it has to be all done through, you know, through algorithms and control systems. So let's talk about DER, which is distributed energy resources and cell resources. Right. So just, you know, context distributed energy resources includes PV, rooftop PV, things that are resources, things that can affect energy, either import or export energy that are at distributed, not central like traditional generation. So PV grid batteries as well. In addition, demand response devices like maybe a water heater that can be turned off if needed. Those are all resources that could be called on potentially to help with the grid, right? You just defined the word, I mean, ancillary. The word ancillary might be foreign to most people that are listening. Right, right. So ancillary services are also just grid services is all the things that utility has to do to operate the grid reliably besides the bulk power and energy. Right. Like we're used to our utility bills as a residential customer says this may kilowatt hours or charges or this and a large amount of it is that kilowatt hour energy rating. And we think that all that's what we're providing or paying for. But we're also paying for all the extra stuff the utility has to do to maintain reliability. Right. In under disruption under things that happen. Right. And delivering that that power to us at the proper voltage and frequency and all these things. Right. So, you know, as we put more renewable energy PV on the grid, these PV does not provide grid services in the same manner that a traditional generation was. And I think I pointed out inertia as a thing. Right. Like freak inertia is associated with frequency support. Right. That inertia helps support and maintain frequency. But, you know, inherently the electronics don't do that. So something has to help with that or else we have a much different behaving system. Right. Let me. Sorry, let me find in just what why is frequency control so important? Well, we maintain a very precise frequency around 60 hertz. I mean, the deviation is very small, typically around 60 hertz. Now, if you, if that varies, you know, far enough, equipment will be damaged ultimately. Like, I mean, just to put it in terms of a household in the extreme, if frequency was really, really very alive, you could wear out your motors and things like your air conditioning or your refrigerator more because every time frequency is changing, it's creating a bit of a stress to those components and voltage as well. But there's strict, you know, guidelines on what the utility needs to provide in terms of frequency and voltage. And so they need to do that or else, you know, you're going to see issues, voltage is also, you know, you can literally see flicker in your house when, you know, voltage increases or decreases, right? Do they have real world consequences? And then at the extreme, you know, things do, I mean, eventually the grid will fail if frequency decreases or not because things start tripping offline. There actually are built in protection mechanisms called load shedding that the grid does help with that. Like if things are going down in the grid, like frequency is decreasing, either take the whole ship down or you kind of throw a couple of things off board. And that's what, you know, load shedding is. Like if you're the unlucky one that gets shot off and thrown off board, it's a rotating thing. So it's not only the same neighborhood or whatever, but the grid does have the facilities and capabilities and they do have to shed load or rolling down ground out essentially in order to protect the integrity of the whole grid and not bring us all down, right? That's why some people get shut off during Super Bowl. Right, right. Yeah, yeah. And just another real example is the inverters in Hawaii for PV used to have a tighter band for frequency. They would shut off if the frequency was not within that certain range. That range has now been widened for, it's called frequency over frequency, over and under frequency ride through. So that allowed more inverters to stay online because you can imagine if the grid frequency is going down because there's not enough generation and now you hit some level and now more inverters start going offline because you have gone down, well, you've just increased your problem, right? Okay. So yeah, so these grid services since PV doesn't inherently provide these grid services in at least in the same manner as traditional generation, we need to do something about that because as we all know, these generation, new generation sources are becoming a larger portion of the grid and we want them to. Now that's not to say they can't support these ancillary services. There are capabilities to do that, but they require more careful design and control and coordination and that's a newer thing that is still honestly being worked out of how to properly utilize these devices that now are increasing numbers of consumers actually have PV and possibly batteries and eventually EVs more and more. Those things can provide services for the grid and how do we do that in both the technical manner as well as the economic and the social aspects too. But just for instance, let's say you had a new EV charger at home and you're charging your electric vehicle. Now as long as you have your electricity, your vehicle charged up in the morning when you need to go somewhere, you may not care whether it's charged at 9 p.m. or 1 a.m. As long as it gets there, then you're probably okay with it. So that's a flexibility that the utility could call upon to help meet its need and to have another lever or knob to fix things if things were wrong. So to expand that into other services, other devices, you can see where there is capability there. And this isn't an entirely new thing. I mean, water heaters in Hawaii have a lot of people for a long time have had an energy device that will cut off your water heater if things get really bad for a short period of time and you get a small credit on your bill for purchasing that service. It's the same sort of concept. If you had devices that can provide services, then you could be compensated for it. But you're happy as well as the grid is happier. That brings us to the subject of how do we monitor this? So let's go to the next slide and talk about how it's done today. Right, right. So we have, you know, at H&I, we definitely have an interest in monitoring the grid for research purposes, right. And, you know, going back four or five years ago, we had a project on Maui where we deployed meters in the neighborhood in Maui, 25 to 50 units that it could monitor both at a household level with volunteer households, as well as at the transformer and see the very parameters that were talking about frequency and voltage and current and report that back for research purposes. That gave us a lot of experience, not just the data itself, but a lot of experience on failures of devices, communication issues, like weatherization, like all these things that are like potentially issues, right, that you don't need to immediately think about. We also saw that these devices are quite pricey by the time you buy a device that probably isn't weatherized. You now need to put in another box, you need to get more sensors, you got to get a cell phone modem, you got to get something to read from it to send the data over the modem. Like, it's a, you know, by the time we're done, it's $3,000 per unit. And even then, it didn't really provide the capability that we would have ideally like, right, so we saw a need there just for our own purposes of research that we needed to do something different or better for our case. So here I just, I'm just showing a sort of lineup of some commercial devices, you know, utility, utilities here in Hawaii rely on in particular grid 2020 and the PQ, those are used by the utility. So, but again, they're kind of a one off thing for the most part. The number of transformers that actually have these is still a pretty small portion, right. And again, these, most of these do require extra sort of boxes and infrastructure. And in addition, and maybe going forward, even very importantly, they're not, they're entirely passive with respect to the grid. They just read what they read, they turn it, send it back, right. Well, if you think about that location, we call that at the edge of the grid, like near the customers, right. The transformer, you know, that might be on the utility pole is about as close of an equipment location that you get to the customer beside the meter itself. And it's a very strategic location where you're serving, you know, five to 10 customers. So you're fairly close to the, to, to the edge of the grid to actually, you know, communicate and have that nuanced, you know, control possibly in negotiation with, with resources and how they might be deployed. And so we believe in others do, too, that this is a strategic location to put more intelligence and controls at the grid. So you not only tried to address, you know, just monitoring voltage and current, but also putting an infrastructure and computing capabilities to leverage that for analytics and controls, right. Let's, let's have the next slide and tell us the solution. Yeah. So we've been working on a, on a device and system for at least three, three and a half years now. In designing up a power monitor and control system from the circuit board level up. So literally selecting, you know, measurement chips and, you know, microcontrollers and, and the comms and, and everything that goes along with it, laying out the circuit boards, initially water, sorting things by hand, like working the whole way up into software, into web-based interfaces, the whole deal, and then eventually laying, layering in control system from top of that. So this whole thing has kind of been built up over some years and a lot of students involved with University of Hawaii. So it's been a good educational experience. It also has helped bring in some research funding. It's, it's helped, helped us be recognized enough to participate in, on new federal funding. And it's, it's proved to be kind of a foundation for our future research in, in one of our groups here, like we're building new software and applications on top of this platform now. So, yeah. So in, in, in summary, it's a box that says, you know, several circuit boards and a bunch of components that can be built for, you know, off the shelf, you know, the component cost is around $400, like that, that's sort of a one per unit. And it does AC voltage and current. It has a GPS for very precise time, stamping and coordination. If we want to do that for standard research, it back up power to measure things as the grid actually goes down. If that happens, it has built-in communication and data storage and power supply. So it's a very integrated device and allows us to do our research. Okay. Now, below there, we see some of the graphical interfaces that are web-based that we can do. All data is reported to a database here in Hawaii, so in, at University of Hawaii. So we have both historical data, you know, archived, as well as we have real-time data that is delivered at, you know, very quick time, time periods, you know, basically only the delay of the internet itself to get measurements back to us, right? It was current status. Yeah. So the current status, summarizing that in the next slide, we, in our latest, you know, we've had four, at least four renditions of the hardware and now we won't want it. It's very stable and has all the features we want. We've built 25 of them, 11 of them have been non-deployed, one in Thailand, the remainder at this point in Arizona. We're getting, we have a patent pending waiting for review back from the patent office. We're pursuing the commercialization option. And again, we're using, using this now as a foundation for our current, you know, research. And so some of these sort of thrust of research, just for example, include online capacity, PV hosting capacity analysis. So what that means is when you, when you apply to have PV put on your house, the utility has to go through review process. They have to make sure that that PV isn't gonna cause issues to that particular area of the grid. That's done in a, you know, in a, I guess sort of a one-off, sort of manual process where an analysis has to be done on that circuit and it's, it's, you know, it's not refreshed. It's only refreshed when needed. What we could do with this device is, is try to do that in a more nuanced and regular interval using data from the device itself and knowledge of the architecture of the topology of the grid and model that online and give us results about how is, how stressed is the grid at that point and, and, and therefore where are we at with our PV hosting capacity, right? Maybe find out that there's extra margin here or there and maybe that could lead to extra PV hosting. We're running out of time, Kevin, so I have like room for one more slide and then I have to sign off. Right, right. So the next slide I have is, is just sort of a conceptualization we've included in some of our proposals of, of sort of the integration of this device and system coordinating again to the edge of the grid as well as upstream to the, to the rest of the grid operation and we're working on again pieces of this we, in discussion with utility and trying to do our part to help bring this, you know, concept into question and it is this concept in general is, is very much aligned with Eco's grid monetization strategy. So it's not a, it is research and we're taking it in advanced sense but it's not, it's, it's not out there. This will be happening in, in the near term. Okay. So it's leading edge technology, you're solving a problem here in Hawaii and I'm going to get my little pitch in for HNEI and the University of Hawaii like solving today's problems with very practical applied research. You've applied for a patent which conceivably, if this takes off, could bring income to the University and to the Institute. So this is, hits all the, the hot buttons, helps the electric utility, helps Eco. So this is a really good, really good work, Kevin. So how do we, I think we have our, throw up our, throw up our last slide which, which tells us how to contact you. Sure. Those out there that are interested in this technology and would like to be involved either commercially or applying it. Let's, there you go, there's your contact numbers. So unfortunately, we're out of time. I told you it would go fast. Yeah. And so I've been talking to Dr. Kevin Davies from Hawaii Natural Energy Institute, solving today's problems at the University of Hawaii. So Kevin, thank you so much. Thank you. Appreciate it. Okay. Aloha everyone. I'll be back next week with a, with a person from the Hawaii State Energy Office. So until then, Aloha.