 My name is Jeff Byron and I just wanted to take a moment to introduce this panel. Actually, I want to introduce the panel. I'll leave that up to the moderator. But I wanted to introduce our moderator today is Lori Tenhope. I think Lori's been in the research and development area for a long time, maybe upwards of 20 years. She did some pausing in the middle of that stint to help out a few commissioners at the California Energy Commissioner, at the California Energy Commission. And I was fortunate to be one of those. And she was my chief of staff. But about eight years ago, when the Brown administration came into office, she headed over to run the R&D department or division at the California Energy Commission. And she helped transition what used to be the public interest energy research program. It ran out of steam. It had a 12 year life and the legislature was not inclined to renew it. And so under Governor Brown's direction, Lori Tenhope, working with the Energy Commission, the governor's office, the PUC, the investor owned utilities and the legislature, she helped shepherd in this new program called EPIC, which I can never remember what it stands for. And really has returned to California what was an enormously successful program in a different way. I think they spend on the order of $140 million a year of public funds directed towards energy and environmental research. And I just wanted to introduce Lori. I guess there's a certain pride associated with some of her accomplishments because she worked for me and also because I'm older than she is. But Lori, I wanted to thank you very much for being here today. And I think all of us owe you a tremendous debt of thanks because we've benefited from the probably over billion dollars worth of research that you've shepherded through your organization. And I wouldn't even imagine to guess how many people that you've brought on there and trained. But I wanted to thank you very much for being here today and for leaving us on this panel on one aspect of what they do very well at the EPIC program and in her division. This one's on microgrids. So Lori, thank you for being here. Thank you, Jeff. It was a great opportunity to have a chance to work for Jeff. And it's kind of funny to be on the other side because I would prepare the presentations and he would go off and I didn't have anybody to prepare mine. So here we are talking about innovations in microgrids. And I wanted to set the stage just a little bit before I introduced this wonderful panel. As Jeff was saying, we have a pretty large public interest research program in California, about 120 million dedicated to electricity related research and about 20 million related to natural gas research. And it's all really designed to be a catalyst for achieving our policy goals and helping customers test and get familiar with technologies that will really help us get to these super aggressive goals of transforming our electricity system, dramatic reductions in GHG, doubling our energy efficiency, bringing on electric vehicles, ramping up renewables. You're all familiar, I'm sure, with the transition in California. One of the tools that we think is a potential enabler of these policy goals is microgrids. And the Energy Commission's research program has really made a significant investment in microgrids. We funded so far 17 different microgrid projects. They average about 5 million a piece, so that's really a significant chunk of the research investment. And the projects have, you know, they have multiple goals, but at the end of the day, it's how successful are they in helping us get to these goals and provide customers the electricity resilience that they're expecting. So we've kind of done the microgrids in a couple of solicitation rounds. The first one in 2015 was to really focus on microgrids for critical facilities and microgrids that would really maximize the renewables that were available on site. So, I mean, these microgrids are like a microcosm of the macrogrid, but they have a lot of added sophistication of incorporating customer demand, their renewables and storage, and trying to do it in a way that has sophistication in the controls, so that the customer is saving money, has more higher power quality, and is if sustainability is the goal, they're also able to maximize renewables. So we're going to hear about some of those projects today. They range from wastewater treatment plant that you certainly want up and running if you have a grid failure, a Kaiser hospital that's really focusing on using a microgrid for resilience in its medical building. So the hospital would typically have backup power, but your medical facilities doesn't always have backup power. This allows Kaiser to continue medical appointments with a grid outage. We have a couple of microgrids that are really focused on remote areas, like up in the north coast there at the end of a PG&E feeder. Reliability in that area is a real challenge, and we have a microgrid with the Red Cross emergency shelter area to really make sure that there's a resilient service in area of the state that would be particularly vulnerable. We have a couple of other interesting microgrids that are active right now. One is on a DC microgrid. That one is really focusing on improving efficiency by eliminating the conversions from DC to AC and back to DC and proving out the efficiency in an industrial application. In addition to the active portfolio right now on critical facilities and renewables, we just recently awarded 10 new microgrid projects, and now we're looking at what are we learning from this that's going to inform commercialization of microgrids. We're treating this as a portfolio where the principal investigators and the stakeholders from the various microgrids are talking to us and talking to each other. We want to capture what the technology innovations are, what are the challenges in terms of permitting, what do we need to know in terms of cybersecurity, and then capturing and translating this to really provide information on what are the best applications for microgrids and what are the best practices for configurations in microgrids. I just wanted to set the stage for why are we doing this and why are we investing basically your money, ratepayers' money to explore these questions. Before I introduce the panel, I also wanted to just say if you're interested in this topic, I'd welcome the audience members and our panelists. We're going to have a conversation this afternoon at 2.30. We'll continue to talk about this if you have ideas that you want to share or some suggestions on future research. That would be a great opportunity to have that discussion. So now I want to introduce our panel. I thought it was really interesting reading the bios of our panelists because when we think, like I said, microgrids are such a microcosm and really kind of require that you think about the integration of buildings, electric vehicles, renewables, storage, and the bios for each of our individuals are pretty broad and deep across these areas and bring that integrative thinking to this opportunity. Our first panelist is Doug Black. He's a research engineer in the grid integration group at Berkeley. He leads the microgrid and V2G integration demonstration projects and he's been involved in some really interesting V2G demos at military bases and with the city of Alameda. Our second speaker is Vipol Gore. He's the president and chief executive of Gridscape Solutions. They're designing renewable microgrids for critical facilities and really focusing on the integration and control to maximize resilience at those facilities. He also has a prior experience with ChargePoint and deep software expertise. Last is Silla Kilichote and she's a chief scientist and leader at the Gizmo research program. I'm wondering if you came up with the acronym first and the name later and also is a managing director of grid innovations here at Precourt Institute. She's also led demand response at Google and the demand response research center at LBL among other expertise. So our format for this is first each of our panelists are going to just give an overview of their microgrid work then we'll have a moderated discussion where I will ask some questions of our panelists and then we'll open it up for audience participation. So first up Doug Black. Thanks Laurie. Thanks everybody for attending here. I'm going to present on a recent award as Laurie just mentioned. The most recent call from Epic on microgrids. This has been recently awarded not contracted yet but so I'll just be giving a high level overview of what our objectives and our goals here are with this microgrid at the parks reserve forces training area. Next slide. Do we control the slide? So this is a army reserve training area located in Dublin, California. It's about 30 miles east of Oakland. It's a small to medium sized military facility. The microgrid here the military facilities have particular resilience needs have different structures of operation of procurement of installation so they have their own unique characteristics regarding microgrids but they also there are things we can learn here too for other similar size campuses. This is on the small to medium size for a military site but is somewhat unique for military sites in that it's as Laurie mentioned some of the other ones not at the end of a line. This is in a suburban area in Dublin so we're not really resource constrained here but it gives us the opportunity to demonstrate the capabilities of having onsite renewable storage and how that can either by supporting itself lessen the impact on the grid in outages or case of emergency so that PG&E the supplier here can provide other critical needs in the area one particularly being a county jail that is nearby. So the main innovations we are looking at to maximize resiliency commercial viability and uptake by military bases and hopefully other similar type of campus locations are is to island with critical loads with 100% renewable generation and control of a set of a diverse set of distributed energy resources that I'll describe on the next slide of showing the system diagram but using an LB&L developed Durcam based supervisory controller that will integrate control of various stories from different manufacturers so PV battery electric vehicles. One of the innovations as well is to develop a nested modular architecture with what we're calling an integrated resilient node. This is basically a packaged medium to low voltage transformer switch gear communications and control equipment that can be used to cluster critical buildings, cluster different groups of buildings, particularly critical buildings and use local rooftop and locally cited storage for supporting individual or clusters of buildings in the case of even an onsite failure or outage. With the military there are also special procurement issues, special installation things that we're going to be documenting, streamlining, optimizing, working with the military and their procurement people to come up with templates that make this easier to develop and and implement at the other military bases in California and the US and even really beyond. So just to give a quick overview of our system here really I just want to highlight a couple of the main components here are our control system supervisory control layer with Durcam and a market participation layer that will be developed by customized energy solutions in which we will use the storage that we have and for participating in wholesale markets to generate revenue. The controller will also optimize PV storage, building load control, the electric vehicle charging control to minimize electric utility charges, particularly demand charges. We're going to start, the project is going to fund one of the integrated resilient nodes with a cluster of buildings that will also have PV and 100 kilowatt PV and a 400 kilowatt hour battery storage. This is a real partnership though between the military and the state of California in that the military is providing a good deal of infrastructure as part of this project as well in which they are going to provide the two megawatts of PV, a two kilowatt for a, I'm sorry, two megawatt for a megawatt hour battery as well as the integrated resilient nodes for an additional up to 14 clusters of buildings. So this is a great demonstration for developing a really key kind of market here. There's over at least 30 military bases in California alone that can benefit from this and funds more across the country. Good morning everyone. Thank you so much for giving us an opportunity to Stanford University to present our company and our microgrid project. I also really want to thank the California Energy Commission as well as EPRI. They've been our brilliant partners in this projects that we have done. Gridscape is a clean technology company based in Fremont. We were five years old. As Laurie mentioned, my background has been in software systems and IoT and energy for last 10 years and around 2013, I saw that there's a big opportunity to integrate storage, solar, electric vehicle charging into microgrids. And so we embarked upon then and we bid for the project. As Laurie mentioned, there was a project in 2015 to deploy or create microgrids for critical facilities. And we were one of the awardees that time. And then subsequently this year, based on the results of what we have achieved so far, we became an awardee of the second grant to commercialize the whole technology that we have developed so far to spread it out into different critical facilities and different types of emergency operations centers across the state and the nation. So thank you so much for the California Energy Commission for the work that they've been doing and allowing us to basically bloom and foster in this kind of environment. Our mission in life is to lead clean energy transition through building renewable microgrids as well as vehicle to grid PV transportation. This project, the Fremont Fire Station project, was the first funded project through California Energy Commission grant back in 2015. We got about close to $2.4 million from CEC to deploy three microgrids behind the meter microgrids. This is different from the military kind of, you know, a campus kind of microgrid. This is more behind the meter customer centric microgrids. And what we did was we deployed about 50 kilowatt solar canopy systems at the fire stations, about 120 kWh lithium ion battery systems and basically have a microgrid controller that manages that site along with the software in the cloud. It does about six to eight hours of islanding. Again, that's basically a factor or function of the size of the battery. Initial goal was to just do three plus hours of islanding. But recent tests have revealed that we can go beyond eight hours also, especially during summer months, winter months, you know, when you have a little bit cloudy weather is different. But otherwise, you can do that. On an ongoing basis, this project saves as of today about 75 to 80% of energy savings back to the customer. The good thing about this microgrid is, you know, it works in a grid tied mode as well as an off grid mode like any microgrid would do. But you know, when we talk to the city of Fremont and, you know, the fire department, they say that, you know, we have this diesel generator and we have invested into it, but we seldom use it. We don't use it. And so this system actually produces energy savings on an ongoing basis. And then we have got grid resiliency on top of it, which is really a great thing for us. So it's really producing great results. And the first site, which you see the one on this side, I don't know how to use this pointer. Oops, sorry. The one below here, that has been operational since September of last year. And the other two sites are going to go live this summer in about months time. And so I think we'll be finishing our demonstration by end of this year, at least for these sites. What we have done with this technology is, you know, we actually ended up inventing our controller and the distributed energy resource management system. And the reason for that was when we first built this project and I looked out at partners or, you know, vendors to provide such systems, but they were very expensive because they were mainly utility scale, you know, large assets and, you know, they were built to kind of manage large networks. There was nothing in the space of this particular market segment about, you know, 50 to 100, 150 kilowatt solar systems and those type of systems. So we ended up building our own controller and software that is one-tenth of the price of, you know, other available controllers. So we are actually able to achieve now the economic scale that we need to kind of take this forward. Typically, you know, this is how our microgrid works. You know, you have a facility that has a panel in the center here and then the utility connection, some backup generator connected to an automatic transfer switch. And then, you know, there's a critical load and a non-critical load kind of, you know, obviously, you know, when you're going to facility, you have a discussion of what's critical, what's not critical. When I did that discussion with my fire chief, I said, okay, sir, can I sit down with you? Can you tell us what is critical load in your facility and what is not? And he says, everything in my fire station is critical, period. And you can't argue with a fire chief. So, you know, even if, you know, they run, have a TV running, it's still critical, but that's okay. We don't argue with the fire people. They do great job for us. So in the fire station, everything is critical. And then we add solar panels and some smart inverters with a battery storage system, and then connect it through our controller. The controller basically manages each point in the whole microgrid, whether it's an energy meter or the smart inverter or the service panel or with the battery system. And then each site would have one controller and that goes up to the cloud that basically aggregates all the data. In the cloud, we do utility tariff integration, API integration with different types of demand response program. We also do electric vehicle charging infrastructure integration in the cloud. So this piece, we have standardized into a package that now we are actually commercially also selling throughout lots of customers in California and beyond. There you go. So now, you know, basically that this architecture allows us to kind of scale it up to a wide area microgrid network, all the assets, the microgrid actually behind the meter, but then you have a scale. There are 3,000 fire stations just in California. We did three in Fremont and city of Fremont, we have actually agreement with them for a 10 year power purchase agreement. And they told us that, you know, when economics work out, you want to do all 11 of them. In the next project, we are doing three in Stockton and Stockton has about 17 fire stations. So they're interested in doing all of them, not just three that has been funded by the CEC. So we absolutely see there's a scale happening here. The economics are working out and just the energy savings that each of these sites provide just basically allows them to just pay for it over a, you know, over some period of time. So it's been a great project. And then just not just fire station, but you can basically scale it through all the other kind of CNI retail, you know, emergency operations center, school campuses and across, you know, all the different types of customer segments. So thank you. Looking forward to the panel discussion. Okay, so a little bit about Gizmo grid. It stands for grid integration systems and mobility. That's the group that we have at Slack. And before that, I was at gig, which was at Berkeley lab. So we come up with cool acronyms for grid related research. So I don't know how it happened, but I somehow snuck into this panel without having an active micro grid project. But we are doing some really interesting things with distributed energy resources in terms of integration and coordination that can inform a lot of the micro grid work. And I know that many of us think that the distributed energy resources are gateway drug to micro grid. So I'll talk a little bit about what we're doing here at Stanford and Slack with distributed energy resource coordination and integration. So first of all, thank you CC for funding our work. And thank you all for being here. This work is also funded by the ARPA E program under their notes program where they asked many participants to coordinate distributed energy resources to deliver ramping and regulation products from many, many different end uses. So at Stanford with the leadership of Dr. Ram Rajagopal, we are developing three key technologies to integrate distributed energy resources. So let me start with, oh, sorry. Okay. It does this. No. Okay. So let me start with the, okay, I can't do that. So let me start. Okay. I can't do that either. So let me start with the smart dimming fuse. Smart dimming fuse is basically a fuse that sits in a home panel. And at the bottom, you can see that there is the time scales of operations for these technology. So smart dimming fuse is the fastest reacting technology in our technology toolbox. And its job is to do several things. We can remotely control an entire circuit using this fuse, or we can actually change the voltage levels at that circuit and control the power based on that. Now, not every time you change voltage, you can control power. So the idea is that we will figure out using machine learning techniques and other things, build some intelligence into the dimming fuse to figure out if any time that we change voltage, we can change power, and then control only those that we can actually do change the power levels. So that's sort of the smart dimming fuse. Then we have a home hub, which is a typical, in a sense, home gateway device. It maintains the stability of the system with distributed energy resources behind the meter, including a solar PV installation, a storage installation. And it also, its job is to really send data upstream and get commands from the cloud coordinator, which is the third technology. The cloud coordinator actually receives signals from the CalISO and takes those signals and disaggregates them to individual homes as commands. So it has a lot of intelligence, and it does a lot of the coordination and distribution of the commands. And the time scales of operations are varying in different technologies. And we organize the technologies or develop them in such ways that they're a clear open API so that if you want to use one piece of it, you can use that one piece only, and then you can use other products upstream. So I think being able to have open interfaces is a huge need in this space because we want to build microgrids very quickly and cost effectively in having those open interfaces, the key element of the microgrids. What we're doing with this technology is it turns out that we're also working with the military. The idea is once we prove that it works in our labs, there's a bits and watts lab here in Y2E2 in Stanford, and we have a gizmo lab at Slack. So we're building these virtual homes and actual homes in our labs and then looking at how this technology performs in those virtual and actual homes in the lab. But unless you take the technology out to the field and understand how it's installed, what are the costs associated with its deployment, how is it being used, what are the interfaces and how people interface with those technologies, it's very difficult to gauge if that technology is going to be successful. So one element of it is that we're taking it to Navy Housing in Vista Ridge in San Diego and testing them out with 20 homes that already have solar deployment and we're putting these technologies into test their functioning. In some of these homes, we're also putting some storage devices so that we can look at more sophisticated controls and coordination of the system. Last one. So quick things to think about that I wanted to share with you is that we're really serious about lowering the cost of microgrids. We really need to think about the life cycle of microgrids. So the first thing that we know today is the integration and labor associated with designing, building, operating microgrids is expensive. It costs more to do all these things than to actually buy the components and build the system. So how can we actually start thinking about lowering the cost and the design phase? Well, if we could have open libraries for technologies where people can pull information easily and start designing these, that would be a very streamlined approach and could potentially save a lot of money. So we want to see more machine-readable technology descriptions so that tools can actually use these machine-readable technology descriptions. Then we'd like to see some machine-readable tariffs in that in the sense that when we design a microgrid today, we use the components and we design it for the tariffs or the rates that customer is in with the information that we have at that time. But over the life of these technologies, the rates will change, the load will change, the usage of the systems may change. And so while we're designing for a snapshot, we need to think about the continual improvement and optimization of these microgrids and we need more dynamic data from data sources. I think it would need, if we think about community-scale microgrids, we need to develop more resource sharing mechanisms and we have been looking at blockchain applications and resource sharing using community-scale microgrids and, of course, the settlement mechanisms because if you collectively put together resources for a community, then how do you equitably share those resources is still an open issue. And then we think that there is a need for tools, a growing need for tools for stability analysis of these microgrids and distributed energy generation as a whole. So we are big proponents of developing tools for stability analysis as well. So I'll stop there. Thank you. So each of you needed to convince a facility owner that this was something that they should participate in and it's a research project so it comes with some risk. What was the compelling factor for facility owners to participate in these projects? Well, in my case with the military, we actually had a very enthusiastic facility manager who had applied to be one of the net zero energy facilities for the military. So he was already going to have a lot of renewables and storage to meet that. He also has a very stringent resiliency need, 14 days to operate critical loads. So combining those two, it was a pretty easy sell for him. This is microgrids. As far as the microgrid kind of gets used a little loosely for anything with PV and battery storage, but really it has to be able to island and that's the resiliency component. And the military puts a high value on resiliency, but that still is one of the challenges in this area though for I think other customers is them being able to put a dollar amount on that resiliency of saying, okay, this is how much a microgrid system is going to cost you and it gives you this much resiliency and it's just a fairly big number and being able to, well, they have trouble answering themselves and people could probably speak to this more directly, but is that a real value for the level of resiliency protection against outages that will give them? In our case, there are three main factors, forces that basically allowed us to get the sites that we're looking for. In the first project in the city of Fremont, and that was an example because that's how all the other cities are following, have similar programs. The first one was energy savings with putting a system that basically lowered their cost of energy. Second was grid resiliency, obviously for critical facilities that are very important. And then the third was every city has a climate action plan. So they have to basically, they want to basically reduce the GHA emissions by certain percent, by certain date and time in future. So all those three factors allowed us to basically work with the city to basically get, look at the sites that we want to basically deploy this at. And that value proposition is actually also reserving quite well with all other cities that we are working on now. We are probably are talking to about 20, 30 cities in California, and all of them are basically, want to basically go down that path. To Doug's point about grid resiliency, the energy savings has a clear economic value. I mean, you can see, if you are going to pay 19, 20 cents a kilowatt hour or $35 per KW for demand charges and upwards of those kinds of rates in California, and the solar systems today can bring that down to 10, 12, 13 cents kind of PPA style energy cost. It's a no brainer. But then the resiliency is difficult, but here's the perspective. I think this is what resonated very well with the customers that I'm talking to. Monterey Unified School District, they have a food nutrition center in the Monterey Peninsula that sells food for all the students in the community over there. And last year there was an outage. Every day they lost $75,000 worth of food just because not having electricity in their cold storage. The busiest airport in the whole world, Atlanta was down for 11 hours last year during a storm. We all know that it basically has 260,000 passengers basically going through and they were completely stranded. The Northern California fires by December of last year, the state and the total loss created by those ravaging fires was about $9 billion. So I think it's hard to kind of quantify the resiliency in a small project that we deploy, but when you put this picture in front of them, it's so much taxpayers money and the rate payers money going into just after the effect of an issue like this, then I think we can get over the discussion and say, okay, let's go and do this project. That's what we are seeing here. Thanks. Silla, how about the Navy? Yeah, so we always try to identify the early adopters and for the military, it has been a huge goal for them to adopt renewables and resilience has been a huge goal as well. So we have had many discussions with them around resilience and renewable adoption in their facilities, but this Navy housing facilities is fairly unique. One of the reasons I like to work with military is because it's a command and control sort of a situation. You go to the commander and you convince the commander, then you get 20 buildings at all at once. So that's a really nice, easy way to recruit buildings for us. But all joking aside, they're very, very eager and they're very knowledgeable. So the successful projects that I've seen are, if you have a site where the facility manager or the site operator is eager and willing to participate and willing to grow and very knowledgeable, then the projects really seem to go much more smoothly. So that's sort of our selection criteria. How involved are they? How much it is in their mandate to adopt these technologies, how willing to take risks with us? So they're motivated. Are there certain things that it was, you know, certain conditions like I'm really worried about X or make sure you don't mess with Y or, you know, you need to prove this before I'm going to allow this, you know, this technology to be deployed. Do you have those kind of conversations? Yeah, we do have, I don't know about you guys, but we do have those conversations. One of the things that they care about is they don't want their utility bill to go a higher. So if we can give them some sort of an indication, some sort of a modeling effort, some sort of a study that shows that whatever we do is not going to increase, it's not going to cause an increased risk. It's not going to increase their utility bill. It's something that they're willing to do more of. And if there is a way to offset their risk for the duration of the project, they're also willing to participate more and take more risk. Yeah, in our case, with the emergency operation centers and fire stations, you know, the biggest, one of those discussion was that, you know, it's a critical facility. So you really cannot have it down at all. And so, you know, one of the conditions that are put up on us was, you know, get the system deployed. If you don't want anything on the rooftop, it should be separate from the rooftop canopy system. And then, you know, just, you know, don't disrupt our regular operations. So even while we are doing the trenching, we had to really coordinate it so well with them that, you know, we do not basically come in the way of the fire trucks and everything, all of those things. So those are some of the things. But, you know, as you probably already heard from with the fire chief, I think, you know, there are ways that you can basically now get things done and it becomes very easy deployable for them, especially the standardized configurations. The ease of operation is another big one, not requiring a lot of new training for facilities managers. That's a big component. So what surprised you? Nothing goes as planned. So what surprised you in the execution of your project? So for our part, it was really the education, we educated the base commander and then we educated the housing unit, the company that was running the housing units. And then we actually needed to go down there and educate each and every person whose house is going to be participated in the study. There were a lot of questions, very good questions that were asked about the technologies. And one of the biggest questions was the privacy and security. So that was that sort of, I didn't realize that it was going to be such a big concern around privacy and security of the homes because they're already behind some firewalls. You know, we communicated to them that we're not going to collect any personal information. But that was a big concern and that was a big surprise for me. In our case, we actually implemented the microgrid and our original estimation was to just get it done within about 18 months of time because it's a smaller system relatively to the ones that Doug and Sheila has been talking about. But we had a lot of our problems. We had the city itself to get our agreement through the city. It took about seven to eight months through council approval and all of those. So because it's a new project, a lot of education, awareness building and all of that. After that, the PG&E interconnection was a challenge for us. So we wanted a net energy meter tariff agreement. And the technology that we used was actually not fitting the, this is all our end days, new just out of the cuff. And so it did not fit the regular process that CPUC and PG&E actually have prescribed of doing a net energy meter interconnection. And so we had to basically work with PG&E. We had a crew of PG&E engineers pretty much camped at one of our fire stations for two weeks, came with all tools and measured every point of our software to ensure that we are not doing any energy arbitrage. And it's just pure micro grid solar, charging the battery from the sun and all of that. And then they gave us a check park. So that battle was one, but that took about almost five, six months. And now we are actually now working with regulatory process of CPUC that's taking even longer. So that was a surprise for us. And then finally, during the, I mean, I'm probably this, the people who do architecture and they do structural engineering probably for them, this is not an issue, but it's a problem for us because a couple of our site had a liquefaction problem in the Bay Area, you know, a lot of the soil is in either liquefaction area. So you really cannot put up, you know, the canopies that you really want to support these things. So you have to actually design the canopy in a specific manner. So the only, only vendor that we found for that was a German vendor who basically knew how to do this very well. So that took some time as a surprise and I will supply the element for us in the project. Yeah, but the biggest surprise has been that already installed equipment in building management systems and data logging. And this is kind of minor. So it's not really a major surprise, but the biggest I've run into personally working with this was we're looking forward to having all this individual meter data that they said, great, we've got all these data loggers, we'll bring it all in. And it took, we're still going to have half of it. It's been months of just how either the system wasn't set up the way they expected. They didn't have the right person to download and there's just the, and it kind of points to how important it is for technology that's easy to use and reliable. So some of this older things that you think are in place that you have this data that you're going to be able to plug right into your control system, you know, isn't really available. That was one issue. So I have a question of most of these are microgrids that are designed to be able to operate for a few hours. So if we're thinking about some of our climate risks of drought, earthquake, fires, we may be out for much longer than that. Are the technologies that you're working on scalable to a microgrid operating for multiple days? Or should we be thinking about a whole different technological and regulatory model for a longer operating microgrid or emergency operation? Yeah, I can take that. In our case, you know, keeping the cost of the microgrid is very essential because we wanted to keep it to the level where it becomes, you know, efficient and economic for anybody to deploy it, you know, we're taking the kind of the great resiliency uncertainty out of the equation really. So for that reason, we kept it low and we proposed a certain size of the system, the battery size and the solar size. In our ongoing, you know, data collection and analysis, we have found that you actually can do a little bit more optimization and you can extend the time. Initially, we had thought that we could island for three hours or four hours, but in our actual application, we have been able to island up to six to eight hours. And, you know, just by increasing little bit battery size, you can actually pretty much do 24 by seven. There are customers who today are asking us that, can you just run the whole site on a 24 by seven basis? And yes, we can. So, you know, our models show that even with about three hours or four hours of sun or, you know, just, you know, even it's cloudy, but there is, you know, a little bit of, you know, solar, you know, output, then you can basically create a size of the battery in that economic model that can allow you to do 24 by seven. But regardless of that, you know, because it's a microgrid and you don't want to always, you know, kind of rely on your intermittent sources of power, solar, we actually, our controller at the end moment lets the solar is out, the battery is out, it will switch to the diesel generator as well, and it can basically run on it. In our newer project that we are going to do with CEC funding, we actually are also going to experiment with vehicle to grid technologies, as well as mobile battery systems. So, you can actually hold a mobile battery system and put it there. And, you know, I see every, you know, Harish nodding the face. I mean, we are going to work with them on that technology. And then we are also going to bring in, you know, medium-duty trucks, you know, into the microgrid so that it can power the facility with that too. So, I think that's coming up. And probably, you know, in the next three, four years, as cities electrify their fleet, this is going to become a reality as well. We're doing that at another military base. We have vehicle to grid to support a building scale microgrid. So, we can talk about that. As far as at the military base scale, we're going to do 14 days of critical load. So, a critical building. To scale out beyond that takes, I think, first some real energy efficiency efforts. That's the first, you know, the loading order of just get it as efficient as possible. And then, as far as technologies are now, it's really just the, you know, more storage. And right, then that's going to either require, you know, lower cost of storage. But right now, as far as doing a base at a base scale, that's really about the only chance we have is to focus just on the critical loads. Yeah, I think more storage is a good one. Maybe longer storage. I think today's technology is up to four to six hours maximum. We may need more longer duration, not May. I think we will need more longer duration storage to write through some critical events. The other thing that we're doing with another project is really virtual islanding. So, trying to identify nodes where we can do zero P zero Q types of controlling the loads to deliver zero P zero Q at different nodes or very minimal reactive and real power. And then supporting that very little reactive and real power with storage that you can reel into those nodes and the distribution network and plug in easily. That's a different kind of an approach to a typical microgrid, but it could certainly support writing through some great events for a period of time, just using mobile distributed energy resources. So, I have one more question and then I'll open it up for the audience questions. So, at the Energy Commission, we're really serious about bringing clean energy to disadvantaged communities and making this transition and revolution available to everybody. And also, those communities often have issues with resilience and power supply already. What would you offer in terms of strategies that you think would be most impactful in disadvantaged communities? Do you think microgrids are a good solution? Do you think it's maybe not the right solution for that particular challenge? The microgrids, renewal microgrids are a great solution because it's an asset that's going to actually save energy to the disadvantaged community 24 by 7, whether you are in an outage situation or not. Basically, in a grid tide mode, you're going to basically leverage your solar and storage assets to offset your electricity bills. It's a great, absolutely great product solution for that market segment. That's basically the new grant that we're going to do. We are going to deploy close to five microgrids in Stockton, Richmond, Fontana, those types of cities, which are disadvantaged. When we spoke to the stakeholders there, all of them basically just said, yep, we want to do this because it's important. It brings in not just the energy efficiency, it basically improves the overall environment and the climate for the disadvantaged communities, also fosters some local growth, job growth. It basically allows people to get more used to this type of sustainable living. It's actually a really good way to basically invest in those communities too for a sustainable future, for sure. Other folks? Yeah, I have a little bit of a different thinking around this. I'd like to hear from the disadvantaged communities what their needs are in a grid event situation and really figure out what does a critical load mean to them because it's different for a fire station or a school than a community. What are the types of resources they'd like to have for a short period of time versus long period of time and then think about the resources around those needs as opposed to delivering any community something that they will only use a few hours a year. So really understanding the energy efficiency opportunities, demand response opportunities, the needs of that community and then building resources around them to support those needs. I'm not really sure we have done a great job in reaching out to these communities and figuring out what their needs are and we'd love to support that kind of an effort. We'll take you up on it. Open up to the audience. Why don't you come up to the microphone and just introduce yourself and ask your question. Hi, I'm Rick DeGolia. I'm vice chair of Penisal Clean Energy and I'm on the Atherton City Council and I think two things. One is there's a big difference when you're going to put in a microgrid as a retrofit versus an original building where you have to build the entire building including that microgrid and with respect to a project like the big project, I don't think it's the facility owner that's the challenge. I think it's the builder and therefore I think that silly year systems approach is really critical. My experience in Atherton just designed which if it were to be built would be the first zero net energy civic center in California. We got two bids in. One is 140% of the estimate and the other is 170% of the estimate and they're unbuildable and that presents a challenge at the facility level because there's people that want to build them without any of the energy innovations and I think the problem there is the builder. I think they just didn't understand how you build this kind of a microgrid as part of a whole project rather than as a retrofit because in the retrofit you can get somebody specialized. So if you could comment on that I think that'd be helpful. Yeah, Rick, thanks for the question. I'm very familiar with your project. So absolutely right spot on. I mean the builders and the architects of the world, they need to be really educated on energy efficiency, microgrid technology, sustainable technology. In contrast, there's a builder in San Francisco that we basically just been working very actively with. He's built a nice house. It's actually a luxurious apartment but he's made it completely net zero net positive. The house pays for the electricity for the usage. In the house plus it pays for about 15,000 EV miles if the owner has an electric car. So it's a great example. We need to have more and more builders and architects like this to understand this and then kind of bet them into the program and work with the code to basically kind of make that part of the code going forward. So I think that's where CEC and all of them can help us. The initiative that we have now by 2020, all the homes need to have a solar installed or a plug. I think that's great. Those kind of initiatives need to be developed. Jerry Gertler with Oracle, just a couple quick questions on the microgrids. What is natural gas left in place for water heating, heating, cooking, drying clothes, etc. And two, were the emergency generator diesel sets left in place at the fire stations? Yeah, they are. They become part of the microgrid. Obviously in our case, this is a backup of the backup because that's how they kind of view them. But they are left there. I mean, for the generator question that you had, on natural gas, maybe Natural gas is still in place at both our military, all three of our military applications were focused on the net zero electric at this point. And the diesel generators, the same as with people, are still tied in our control objective of those to use them minimally and hopefully not at all. But they do have to be connected. Yeah, in most cases, in the housing units that we have, they are using natural gas for heating as well. But there is an understanding that we need to electrify cooking and heating. But it hasn't really been a priority for a lot of the military basis that I've been working with. Just an FYI, the new title 24 does have an all-electric package to encourage electrification. But not many. Correct. Smudd is doing a big incentive program to convert from natural gas to electricity. As an all-electric utility should, right? They do have their motivation, but it's still good. I'm John Fox from Stanford, and I'm thinking a little bit about the islanding capabilities that all of you have talked about. And you also have controllers in these things. I'm wondering how you handle load scheduling. I'm thinking of cyclic loads like air-condition compressors or refrigerators, or anything that cycles on and off, usually autonomously. Do your controllers have mechanisms to load schedule so you don't try and essentially turn everything on simultaneously and exceed the capability of what you have in the island? Yes, we do. We do it at the building scale, but not at an appliance scale yet. Is that something you have to design? Or is this something like CELO was having under smart views where the system is capable of figuring out which loads do this and then figuring out how to schedule it? Yeah, we write algorithms that optimize based on information that's coming in from each individual building's building management system from the PV system, from the battery storage where we're at, and then optimally selects and on a schedule of what the critical loads would be. Yeah, same. It's a very essential piece of the whole control of software and the logic, and you have cyclic loads. They have a lot of reactive power, active power mixes, and you have to kind of make sure that in-brush currents and all of them are balanced across your microgrid all the time. In our testing, we actually had some issues, and we are working very closely to optimize the controller as well as the smart inverters to basically make that happen. So yeah. And we're actually testing two different approaches. One is a queuing theory-based approach, and then the other one is more of a measurement and evaluation using machine learning. So two different approaches, trying to see, you know, do research around these kind of issues and get a variety of controllers in place. Thank you. Amber Kerr, UC Berkeley. I hope this isn't too basic a question, but I'm a little more familiar with microgrids for developing countries where the microgrid is the only grid. It's not integrated into a larger grid. I'm wondering what are the differences in design principles between a microgrid that is the only source of power 24-7, 365 versus these more backup grids that you're talking about? Cost. Expand on that a little bit. In the developing countries, we are active in India as well quite a bit because, you know, we have a large engineering group in India and we actually are associated with some of the research institutes in India as well, an IIT, Indian Institute of Technology. And, you know, they, the idea is, you know, for them, it's, you know, to really reduce that cost to such a low level that, you know, it basically provides just basic power in a plug for people to charge their cell phones. You know, I mean, they don't have TV sets, you know, radio is okay, you know, sometimes they may have a few other things, but to basically keep that cost really down and lithium ion does not actually, it's an expensive technology for them. So, you know, those are the things that, you know, for development developing countries, that's kind of an important factor, especially there are completely isolated microgrids that basically need to provide 24 by 7 power. One thing that I may add is that if you're building a microgrid here in, you know, Bay Area versus Africa, there are two different things in terms of the existing infrastructure and legacy systems and communication control of those, of those systems as well as the scale of the projects in my mind. You know, if you're thinking about a community, we are used to a level of service that some other countries are not used to it. So, there's that kind of a flexibility and culture that sort of you bring into that deployment as well. So, there's sort of a lot of facets to it, but I say that integration with existing systems, existing controls, that's already in place, plus the reliability and availability requirements here are a little different. Thank you. So, I'm Salie Barros from City of San Leandro. I'm interested in the use of microgrids for critical facilities like water treatment plants that have very high energy use, constantly high through 24-7. We have a one megawatt solar array thanks to the CEC being built now. We'd like to include a microgrid in that, but that one megawatt only covers about half of our water treatment plant electricity use. So, there's just a lot, it's just a much more demanding system. So, I just wanted to see what, I think you mentioned that there were some water treatment plants that have gotten recent grants. We have a couple of projects, so I'd be happy to connect you. Well, I want to, we're just about out of time. I want to thank everyone for coming and especially thank our panel. Thank you very much for your insights.