 Welcome everyone to our panel today on energy storage renewables. I'm Harish K. Muth. I'm from Electric Power Research Institute. I'll be the moderator today. I'd like to thank you all for being here. I know this is going to be an exciting panel. We're competing with some of the other very interesting panels, but you have chosen the right one. This one's going to be fun. So I'm just going to talk a little bit about how we're going to do this from a logistic standpoint and I'm going to let my panelists introduce themselves. Just as an initial statement, as you know, what we're here to talk about is energy storage and renewables. It's a very broad topic with a lot of, there are lots of things you can talk about. It's a continent. So we have to restrict ourselves a little bit in what we want to discuss. And so today we've decided to discuss the state of energy storage and renewables, particularly with a focus here in California. A lot of what we say will have applicability in other parts of the country, other parts of the world. But some of it may be unique to California. One of the things that we have in California that we're blessed with is phenomenal solar resource, very mild climate. Those things influence the way that we are going to be able to use renewables in a way that's different from what it might be, for example, in Minnesota or in Texas or New York. So in many ways, you know, we had a discussion, you know, lately and recently and we had on a different panel and we noted that, you know, California in many ways is more similar to Spain than it is to, say, New York. So although, you know, lessons learned can be drawn from what we're talking about here are applicable to a lot of different places, we were going to focus a little bit on what we have here. So the way we're going to do this is we have three, my three esteemed colleagues here on the panel, are going to each give a very short presentation on where they're from and what they're doing in this area. And then afterwards, we're going to have a number of questions that I'll start off and then at the end, I hope to have a few questions from the audience. So please have your questions ready when we call you up. So my three panelists, as you may be aware, are Magnus Abo from SolarEdge. And he will be talking, he will be kind of our representative talking about residential storage and applications. We have Andres Boris from STEM, who will be talking more about the commercial and industrial aspects of storage and renewables. And then finally, we have Todd Strauss from Pacific Gas and Electric. And Todd will be talking about the effect of storage and renewables and how the approach to storage and renewables from the utility perspective. So I'm going to start with Magnus and his presentation on this. Do you want to come up? Yeah, I just pressed the. Fantastic. OK. Yes, good afternoon, everybody. It's it's an honor to be here for this conversation about storage and the impact on the grid. I know we've been talking about a lot of different things today. A lot of what I've heard has been policy. And I just want to warn you, I'm a very nuts and bolts kind of person. I started off in as an electrical engineer. I've been taking things apart, putting them together for a long time. I personally went and got an MBA at a certain point. I've been working on the management side of making things. Like that. How's that? OK, I don't want to get in your lap here, but OK. So so my my background is I am a head of product marketing for SolarEdge here in North America, where the number one manufacturer of residential solar inverters in North America. We're now the number two manufacturer of commercial and industrial inverters in North America. It's a global company. We've shipped over seven and a half gigawatts of production capacity where in virtually every every country that has a meaningful grid and we're participating in storage in combination with solar on a global basis. So a lot off a lot of exciting things that are happening in North America right now with regard to renewables plus storage. But it fits in as a puzzle piece with with what's happening on a on a global scale. One thing I wanted to talk about and I was thinking about this as a as a parable over the weekend, I suppose, I was building a 12 year old a treehouse had to be strong and reliable and above all safe. So I was there. I had a lot of power tools in order to do that, spent a lot of money getting the right kind of lumber, lumber, you know, read up on codes to make sure that it was close as possible. And then all of a sudden the 12 year old came out and her mom said she wants to help. And here are her friends and they want to help too. And so now all of a sudden my job was not just to create something was strong and reliable, but I had to create rules so that these people wouldn't hurt themselves. And although they were helping, I really appreciated the enthusiasm. They were actually slowing things down a little bit. It feels like this is close to the way the that residential solar had its relationship with the with utilities during its first wave. Really appreciate that the enthusiasm, this is great. But we need to have some rules here because there's a great chance that that by participating, you're going to make the situation worse. So we have put in place rules. And from my point of view, that's that that's nothing but a good thing. At this point, however, renewables have become a large enough portion of what's what's happening out there that we need to move beyond the point of simply benignly allowing them to be part of the grid. They actually have to participate in a way that's going to be very helpful. We are designating that as smart energy, right? So not only is is solar going to be generated when the sun is out, it's going to be done in an intelligent way. The first part of that has been putting together rules about how the solar can connect and disconnect from the grid, making it so that it ramps down slowly so that you don't shock the grid, creating bars when that's helpful to the grid, being able to do things autonomously. From our point of view, though, that's a good start. The next step, however, is that there needs to be greater interaction that's already foreseen in terms of what the CPUC here in California has has required, where we have to communicate with the utility how exactly that's going to work is going to be fleshed out as time goes on. But we do have the capability to communicate with the utility in order to make things better. But it needs to go a step further, even than that. And our view is that the next step is, number one, the integration of storage. As Todd will be talking about, there are tremendous problems with the interaction of solar when the solar generation, when the sun is out versus consumption, which happens when everybody comes home and the sun goes down. So storage is a key part of creating smart energy from a renewable resource. The other part of it is is interacting with consumption. So up until now, solar has been operating and generating energy completely independently of what happens to be going on inside of the household at the moment. And we see that that's going to be changing over time as we start deploying smart energy solutions that include that includes storage at the residential level. It's going to take us from our point of view, three things in order to achieve that out. And none of these three things are probably the most important things, which is policymaking and laws. But from a technical point of view, the inverters, the electronics have to get more intelligent. The good news is that that's already happened. There is an excess of CPU processing capability on these devices. They can do an awful lot of very intelligent things, but they also need to be able to communicate intelligently with the network in a safe and secure fashion. And then in addition, we need to be able to create systems where it's not just the inverter creating energy in absence of any knowledge about what's going on in the household. You need to be able to combine all of those capabilities into one box that they are coordinated. So for instance, if you have an excess of PV generation in the household, you might want to have a command to put it someplace, put it into a battery for use later. Or if you happen to have an electric water heater, take that moment to preheat the water or take that moment to pre-cool the house before everybody gets home. There's an awful lot of things that can be done in order to take near energy and turn it into something intelligent that's helping to stabilize the grid. The third thing that needs to happen is this interaction. And this interaction amongst all of the assets to create what's been popularly called a virtual power plant. This is happening now. We are doing this with several utilities, actively around the globe. We're doing this right now in pilots here in the United States. And I actually got off the phone this morning with a utility that would very much like to do a more extended pilot where they would be co-locating storage with residential solar and then commanding it all at once so that they can offset the impacts of wind. So you would actually be using solar co-located storage as a way of storing excess energy from wind, which means that it needs to be dispatched very fast. So these are all things that technologically there is no issue. We know how to do this. So the next step is primarily, as I alluded to, coming up the policy and working directly with with the utilities in order to make these things happen. But but we're ready. And as I said, I'm mostly a nuts and bolts kind of guy. So keep your questions towards me towards that. And thanks very much. Thanks, Magnus. Next, I will. Next, I will let Andres talk briefly about this. Drive it from here. Stem is about a, let's call it, nine years old company, privately owned. We raised $80 million as around the D in earlier this year, where we are financed. The business is going very well. This slide was created, I think, last week when I sent to you. And that time we had about 416 slides. I said that was last week. This week, we had probably over 400 tonic. The energy capacity in this week, probably what we installed, it's over 50 megawatts and probably around 90 hundred megawatt hours. And most of the system that we are deploying, it's a tour system. And we are in the CNI market behind the meter. Deploying our primary market is California, but we have Hawaii, New York, Texas, Massachusetts, Japan and Canada. Canada is became a very hot market for us. We deliver value, not just for the business and government customers, but utilities and ISOs. And I don't really want to go into details because we will talk about later with the panel, but primarily we are using machine learning. Algorithms, AI bagged in the runs in the cloud with very accurately forecasting load based on weather information and other methodology inputs that we get, historical data, usage data. So we can basically provide savings both for the other customers on site as well as we have helping the utilities. How do we help the utilities? If you take a look at what is the difference between a peaker and the storage, we can, I don't want to say that eliminate the peakers, but we can substitute or we can help the utilities so they don't have to build a new peakers. The storage, little Mayan storage or better storage is clean. It's renewable, has no carbon footprint, so has all the advantages that basically we are talking about all day long here. We participate in the LCR with the Southern California Edison and that's an interactive grid program that basically SC tells us that when and how much energy they need. And we basically discharge that energy. What's helping to the utility that they don't have to build extra infrastructure when they are short of the grid supply. They don't have to bring in the peakers. And basically they save them a lot of money just because a few times in a year or a few times in a month, they need the extra energy. The company has a contract for 85 megawatts with SC and we have similar programs with the other major California utilities also. From the CNI's point of view, man, you can take a look to this chart, turn around. Basically what the company is paying is there is the power, how much power they get and that's a significantly more expensive than the energy. You can see in the chart in the last 10 years, the prices is going up and basically we can provide the big saving that provides the big savings to the customers. This is just an example for one of the school that we installed the 600 kilowatt system and you can see in the chart that we are basically saving on the power side, significant amount, saving their peak and that saves the schools and basically they have a payback in several years. Usually we have a 10 years contract and today we can have an installed site that basically is profitable even without government subsidies or the other incentive programs. And finally, Todd, do you want to speak from there? Sure, just pass the, do you have it? California is doing marvelous things and PG&E is aligned with California's vision in fighting climate change. It's all about providing clean energy affordably while maintaining safety and reliability. So for those of you counting here, that's an eight-dimensional problem. Safety, reliability, affordability, cleanness in time and space, let alone what are the metrics for cleanness and the dimensionality, sub dimensionality they're in. So that's one of the reasons why I and most of you in this room are really interested in these kinds of issues and working in energy. Before I hand it back to Harash and we really go into the details of the panel, I just want to check on three things. Who here is a PG&E customer? If you're a PG&E customer, raise your hand. All right, awesome. I appreciate you. PG&E appreciates you and it's working 24-7, 365, some years, 366 to make sure you get safe, reliable, affordable, clean energy. Second, who here is from the California Public Utilities Commission? Raise your hand. Don't see any hands. No one, OK, it makes it a little easier because actually I have very strict reporting requirements with respect to the California Public Utilities Commission and depending upon who from the commission might be here, but not an issue at this moment. If someone walks in from the commission, it's a different matter. Third, who here has heard of the duck chart? Raise your hand. Oh, many. I'm not sure about most, but many. OK, great. And so I just wanted to talk a little bit about the duck chart. And so it's a forecast. It's a forecast that was originally made in 2012 by the California Independent System Operator. This picture here is from 2014. OK, and it represents a typical day in March and the horizontal axis is our hours from midnight to midnight. And the vertical axis is net load. That's not demand. That's net load. So what's load? Think of it as meter load. And so that's not exactly usage because it's like, well, it's what would be consumed minus what doesn't get consumed because of energy efficiency and what doesn't get metered because it's actually being met by resources that are behind the meter, like solar rooftop. OK, so that's that's kind of, you know, important. What meter load is and then net load is that minus utility scale wind minus utility scale solar. And so the point is, wow, this looks like the duck in particular, we have 2012 actuals. This is from 2014. We have 2012 actuals and then we have 2013 actuals. And each year kind of the belly of the duck gets a little bit lower and lower and lower. And so the forecast is in 2014 was for 2040, 2020, right? There to be in the middle of the day, not that much demand in March and a lot of solar. And so it's pretty low. OK, but yet by the time the sun goes down, you know, demand kind of gets back up. You know, the resources are, you know, going away. And so the net demands higher. So there's this ramp, this long neck of the duck. And so these are some of the challenges. There's a whole menagerie of animals throughout the year. OK, there's just one of the challenges. So here's where we are in California. OK, and this year, the neck of the duck was not the forecast for 2020 of 10,000. The neck of the duck was almost 15,000 megawatts. And what was the belly of the duck? Not 10,000 or so, the forecast for 2020, it was a little bit more than 7,000 megawatts that a net load in the middle of the day. So we're ready, you know, the forecast in 2014 for 2020. We're beyond that now in 2018. So these are some of the challenges now in managing the grid to make sure you, our customers, get safe, reliable, affordable, clean energy in time. In time and space, 365, 24 set. So with that, let me now hand it back to the rest. We can get into some of the details a little bit. Great, thanks, Todd. That was a great explanation. So what we're going to talk about, I think, in the next few and what we're going to do is I'm going to see if I can blank out the screen here so it's not a distraction for us. So what we're going to talk about in the next few minutes is how we address this issue. Because as you can see, I mean, renewables is a thing. It's real. It's really happening. We're really seeing it all across California. It's a real opportunity. It is a real deployment and it is a challenge in some respects. So one of the things is about just the planning process. Historically, the electric power industry has been very heavily planned and they look at these things years in advance to try to understand what the load is going to be and they try to create a arrangement by which we know how to build and where to build assets and who builds those assets. That's become complicated, not just because it's not only the utility building anymore. Now we have a bunch of 12-year-olds on the grid. But now, so how do we do this in the future? How is this going to happen? And what I'd like to ask is, I'd like, Todd, if you could start out from the utility perspective, how does this begin? And then I'd like to add, I'd like to ask, Magnus and Andres, how do the 12-year-olds help? It's going to make me regret that. Just want to point out a couple of things to keep it rolling. And so, no doubt, right? 20th century marvel of engineering the electric grid, right? Yes, it's an engineering marvel. Actually, the 21st or early 21st century version is actually not just an engineering marvel, it's an institutional marvel. It's a sociological marvel. There are many different entities who are engaged. It's not a monolithic, vertically integrated utility in control, certainly not in California and it hasn't been for more than 20 years. And so, we have a variety of different entities. So even when we talk about the grid, there's the role of the California independent system operator focusing on the bulk power grid. And then PG does various things to maintain reliability and operate distribution grids. And a key challenge that we're now focusing on is not just renewables or solar, but it's distributed energy, right? So it's not just large wind farms and large solar power plants out in San Luis Obispo County, right? It's a whole bunch of, it's orders of magnitude more units and resources deployed on our roofs in parking garages, in micro deployments in neighborhoods. And so that just becomes a whole different kind of problem on scale and how to think about this time space issue. And so that's one of the key facets of the problem issue challenge that we have in California today because California has really embraced that. We have 900,000 solar PV units in California for about 6.7 gigawatts, 6,700 megawatts. So that's a little bit, maybe that's about 8% of the total California deployed resources. And so it's beginning to be mind boggling. So PG&E in 2017, 33% energy delivered renewables provided to customers and we deliver even more 80% carbon free, big challenges. So how do we handle that? And so our institutions are beginning to fray because they were designed in the 20th century, not the 21st century, so over to you Max. Yeah, I was gonna say, one of the things that can help out here I think is, as you said, partnerships, effective partnerships and maybe one of them are the inverters that you talked about, Magnus. And especially, is that something that's a partial solution to this approach to the issues that we're facing with massive adoption? To a certain extent, I mean, first of all, just to parse it a little bit. An inverter by its nature is doing a very simple thing. It's just taking DC and turning it into AC but it usually comes with control capability. And it's really the control capability that we're talking about here. And especially the control capability when we combine it with storage, with site deployed storage behind the meter. The reason we think that that's a good idea is, number one, it allows you to take that energy that would be at the belly of the curve. And because you have an overabundance of PV and to use it in the evening, that's number one. But the fact that it's dispatchable once you actually get that communication in place so that it's dispatchable by the utility, all of a sudden you can fix a host of other issues. Here's an example. Right now we're seeing in Hawaii, which has gotten abundance of solar and great big fat puffy white clouds or cloud comes over parts of Hawaii and all of a sudden all the diesel generators are just grunting and grinding, trying to keep up. The fact that you can now dispatch batteries in order to make up for that, solve certain problems locally that are affecting the grid globally. And as I said, the ability to simply move energy back and forth within the distribution area is very useful. There is a limit to it. I mean, the reality is that our local distribution areas are not designed to take energy that's here in Palo Alto and move it to Fremont. It's not set up to do that. And there are certain structural problems so as long as we look at solving the problems within local areas, there's an awful lot of possibility in order to allow the utility to use storage as a way to solve both local problems and global problems. One of the tricks that we've all got to figure out though is how to do that without antagonizing the consumer. So the example I like to take is the Nest thermostat. Here's a device that nobody had done much innovation on for a couple of decades. And all of a sudden, the company comes in, makes it smart, but also makes it very attractive, very easy to use, very unobtrusive. And all of a sudden, not only is it being deployed widely, they're getting higher revenue for it. So I think that people are willing to pay in order to participate in something that is gonna be useful for the grid as long as it doesn't inconvenience them. And the control capabilities that we're trying to build into these devices in order to do that need to focus on that. It needs to be very, very simple and just sit there quietly saving people money and making things better without really impinging on people's lives. That's a, I think that's, I think this is particularly compelling also from a business perspective. Residential consumers will certainly appreciate that there's no impact on their lives. But business, I think business customers will want more of a solid value proposition. So maybe Andres, you can tell us about that and why they would want to put an energy storage system behind the meter that would help the grid in addition to maybe provide some service to them. So let me start with the batteries. Two years ago, the systems, what we installed, it had seven cabinets, like 19 inch cabinets and it had the capacity of about 100 kilowatts. Last year, we installed a system that had about five cabinets and the capacity was about 235 kilowatts. This year, in a few months that we are going to install the next system that has basically two cabinets and the capacity is about half of what was a five cabinet. So my point is that the battery technology is improving. It's the footprint is shrinking and that's a big thing for the commercial because they don't have the place to install it. What happens that for the same real estate we can install higher capacity and with the higher capacity, we can provide more value. And as your name, Magnus indicated that controlling this asset centrally, basically we can help the utilities, we can help to basically put a diet on the dock by discharging the batteries in the middle of the day. Having the very intelligent and what STEM is doing that we have a local control and we have a control in the cloud. The two things works together so we can have very accurate forecasting than what we need to discharge the energy and basically helping both the customer reducing the bill and the same time helping the utilities and the ISO shrinking the belly. Great, so I mean it sounds like, it sounds a lot like from what you're saying that energy storage is integrally linked to the renewable growth in California. And any future renewables growth that we have has to be coupled with energy storage for maybe I'm reading too much into that. Tell me why Todd. Sure, right? And so the challenges we've identified, okay? The point is there are many different kinds of potential solutions, okay? So energy storage is one of them and putting smart inverters on solar, feathering the blades of wind turbines, demand side solutions, pricing solutions for customers, load management, having a wider geographical footprint in which to integrate the grid. So not just California, but let's go to Oregon and the western part of the United States. Possibly even gas-fired peakers that the gas were renewable gas. So there are many different kinds of potential solutions to these issues of dealing with the system with the grid. And so the question becomes, well, what are their characteristics? And at the end of the day, what's it gonna cost? Great, I think that's a good answer. So we all know energy storage is gonna be the best one though. So tell us, tell us, Magnus, so my contention maybe from what I heard so far is that the only question about energy storage is where it goes. Is it gonna go behind the meter or is it gonna be owned by the utility somewhere? So Magnus, tell us why residential storage is the best place to put energy storage. So I think that there are an awful lot of advantages to locating storage residentially. I don't think it's the only place that it needs to be located, but there are a lot of advantages to doing it there. Essentially, if you're consuming energy at a location where you generated that energy, that means that the net impact on the distribution network is zero, right? So if we do that, there's an awful lot of benefit to be gained by simply not having to invest in the distribution area. There's a couple of places where storage is being used now residentially as a requirement. This is in the APS territory where they're using it in order to make it so that they can actually avoid investing in the network overall. Those situations come up fairly frequently. I mean, all of us know that there are more people piling into our most populated areas and investing in the existing infrastructure isn't necessarily attractive. Having a co-located storage with solar actually resolves part of that quite nicely. It also does solve another problem for the consumer, which is that if you have a storage device on site, you can also have backup power. So if there is any grid unreliability, and I lived in PGNEE territory, and it's not that bad, it's pretty good here in California. Out on the East Coast, up in the woods, it's pretty bad. Vermont, New Hampshire, you get a lot of snow, falling trees, so here it's pretty good. But people are still extremely attracted to the idea that they have energy if, for whatever reason, the grid goes away. And again, this is actually a proposition that people have shown time and again that they're willing to pay for. So they are willing to subsidize the cost of the battery in order to have that backup capability. So that's a huge help. So in my opinion, it does matter where you put the storage is going to help. I think there is room for the, next to the solar plants for the utilities. There is definitely room and a good business case to behind the meter for CNI as well as residential. As your storage is bigger, probably it's less expensive per kilowatts. So what I see that the residential business is probably harder to justify today. The CNI, it's probably it's in a break event today, average. And the least expensive is the utility scale storage when you install 20, 40, 100 megawatt hours. Probably you guys read about Tesla installed 100 megawatt hours in Australia, right? In 100 days, that was a marketing ploy, but. So I think there's a future. The price is coming down and it's going to help the utilities and the residential. He taught it eager to say so. I just want to check in, right? So when you're talking about storage, Andrews, you're talking about batteries, right? I'm focusing on batteries, but there are other storages. And you know, Magnus, when you're thinking about storage, you're thinking about batteries? Residentially, yes. Okay. I'm a battery guy. Although you will allow that one other type of storage that we see a lot of help for is thermal storage. So hot water. Right. I think it's important to recognize even energy storage is actually a pretty wide diversified class of resources, right? From small lithium-ion batteries in consumer electronics to deployment behind the meter. We talked about utility scale. It's largely the same technology. Lithium-ion batteries is just kind of aggregated together, but they're lead acid batteries. There are other kinds of batteries. There's compressed air. You can take air and put it into a pit underground and use that in and out to generate electricity. There's some efficiency loss. But again, the reason why we're doing this is that there's some economic benefit that overcomes that physical efficiency loss. And of course there's pump storage, right? Which is 100-year-old technology, right? And PG&E has a 1200-megawatt pump storage facility built in the 1980s up in the Sierras. And so even energy storage is a wide variety of technologies. I think it's important to keep that in mind. Yeah, you have a point. So do you think, Todd, that utilities are interested in owning storage or being a part of it? Are you more interested in coordinating storage that's behind the meter? So most importantly, I wanna deliver to you, our customers, safe, reliable, affordable, clean energy. And so the question is how to do that, right? Where does the energy come from? Because storage in and of itself doesn't produce energy. It's actually net consumer of energy. Sometimes it produces, other times it consumes. And so we need to think about, so where's the energy coming from for that? By and large, again, storage is one choice among many. We tend to be kind of technology agnostic. California policy has a clear preference for certain kinds of energy resources rather than others, right? For energy efficiency and demand response and conservation and renewables and distributed generation rather than gas-fired power plants or coal plants or nuclear plants. And California policy has a particular preference today for certain kinds of energy storage, relatively small-scale distributed storage, whether it's in front of the meter or behind the meter, okay? And the intent isn't because it's so cost-effective today. California's stated intent is to transform the marketplace, okay? So by spending a bit more today, you are customers and everyone else in California, right? We're basically trying to stimulate the market so that costs, we talk about costs, will be much lower tomorrow when we need to deploy it and when the world needs to deploy far greater quantities of it to integrate the massive amounts of renewables into the grid. And so the key question is, what's that cost trajectory? Because when we look at solar PV panels, in 2006 the forecast was $4.50 a watt. And it looks like a lot of other technologies then seem more promising for renewables. And you, our customers have benefited from Chinese subsidies and federal tax credits for solar PV and the California Solar Initiative. And so we've done that in California. We were kind of third. Germany and Spain were one and two going first. On storage in California, we're going first worldwide. And so the question is, this is the $100 billion question or more worldwide, right? Is what will that cost curve look like? If it looks like what happened in solar PV, more storage and there are lots of opportunities for every one of us, all three panelists, all of you to participate in the value chain to own, deploy and so forth. If the cost curve looks like, oh, it doesn't decrease like that. If we don't get Moore's law for energy storage, hey, maybe customer pricing programs and some of the other kinds of technologies may be more promising to do this renewable integration that we've been talking about. So, Magnus, can you talk about the cost reduction and what you're seeing on your end in terms of how costs are falling and whether they're falling at the rate that Todd thinks they should? So certainly we've seen a huge reduction in the cost of PV and primarily that's been because of as I was talking about things that have happened in China and their enormous investment in plants there and then subsequent overproduction and things that we now consider to be dumping and but the end result is that what used to cost several dollars just for the raw solar capacity on the panel now is a fraction of a dollar. I don't personally expect to see that and now this is a personal view and not a solar edge view. I don't personally expect to see that for batteries for a, to the same extent for the simple reason that we're gonna have one positive thing which is that we're riding the coattails of transportation, right? You know, the Tesla battery that you buy for your house is largely the same as the Tesla battery that goes inside of a Model X. So the good news is that all of that investment is going into pushing down that cost curve. However, we won't see the same thing that we saw in PV which is that massive overproduction because there is strong demand for lithium ion batteries. So that's gonna continue on. We are seeing a continuous downward movement in the cost of lithium ion in the same way that you're seeing it in the transportation sector. But the reality is that within certainly for residentially located storage, a lot of the costs are not the pure chemistry. It's the other components. And to the tune of three, four, sometimes even five times the cost of the chemistry is actually represented by the other parts. That's where we're working on it. Okay, so from our point of view, we're pushing down the cost of those components. Now all of a sudden we get to get involved with things like A6 and following actually components that do subscribe to Moore's Law. So the cost of the ancillary components to deliver that chemistry on site is coming down and we do expect that to come down over time. Anders, do you agree with that assessment? Partially. Most of the industry expert and they predict a continuous price reduction. Maybe, I mean, we're not talking on Moore's Law but the 10, 20% per year. The issue is, and when we're talking about storage, I'm talking about Magnus talking about lithium ion. We're not talking about molten salt or compressed air. The issue is with the lithium ion batteries that the supply chain. Cobalt is manufactured, it's mined in Congo. Congo, it's not a stable country and the processing of the Cobalt, 90% happens in China. So it's a limited supply and the EV industry has a much higher demand and clot. So when there is a short supply, the EV industry is gonna get the, basically the batteries, not the energy storage. And we have seen in the past that actually a certain manufacturer raised the price versus dropping the price. It's not the general tendency. And the benefit in the CNI that are installed, the installation, the permitting is not five or four X, it's probably, I would say, between 30 to 40% of the total cost. So there is room for improvement on that one if the utilities can help us. And the local authorities, so it's definitely is a policy issue to make the permitting easier and faster and less expensive. And the same time, I think the hardware costs, chemical chemistry is probably coming down in 10 to 20% per year. I just wanna follow up on that because I think that's exactly right, Andres, is that in fact a good portion of the costs as Magnus Anderson described, it's not so much the technology at this point, right? It goes back to their policy-driven costs, right? What's the cost to permit and install and interconnect and the specs, and there are a variety of things that are not just the components to get to full deployment. And so what does that cost? And what's the benefit? And we've talked about benefits to residential customers, commercial industrial customers. And so there's also, there's the financing benefits to those customers who install, right? And then there's the economic benefits or costs to the overall system because we can actually have deployment by a customer and they can basically save off their particular pricing structure, the tariff, which has a demand charge. But the overall economic deployment and the cost for that storage may actually be inefficient relative to something else. And so economic efficiency is one of the things the utility and our regulators are really keen about. And so it's more than just, okay, here's the price and what's it gonna cost for a particular customer? What does it look like for overall customers and particular customer segments and the fairness and allocation between who pays and who doesn't? Because, you know, at the end of the day, who pays and who doesn't is at the heart of a lot of these cost issues right now. So if I may, there is another issues that helping this industries. It is not just the cost curve is coming down, but actually the better is getting better. What does it mean that we have more cycles? A few years ago, you had 1000 cycles. Today we can get better is two or 3000 cycles. That means that the lifespan of the better is extended, which means that you have a longer lifespan. It's not five years, but 10 years or maybe 15 years. We are not braving up at this point that to sell 15 years storage. But I think in a few years, we will see that the storage is going up to 15 years. Same way as the solar start is probably 15 years of their went up to 20 and 25 years. Or even today may have a 30 years warranty on solar panels. That sounds really promise. So it sounds like, despite my best efforts, that my panelists are all agreeing with each other, which is... Sounds, I said it sounds promise. Yeah. And to some extent, you know, I mean, one of the things with PG, we're trying to figure out like how does this actually work, right? So we actually have a few deployments and pilots. We own some batteries. We try to put them in the grid. The first to actually connect to the ISO's real time market and try to bid it into that wholesale market. I mean, there are just so many of these, again, this is complicated technological institutional interface that really is gonna lead to success in California. And that's why I keep going back to policy is at the heart and successful policy is what's gonna help California help us get to our goals. Right, so that's actually, I mean, a really good segue. The question, you know, what is the future of the policy in California? What, you know, given that, you know, maybe costs are not coming down as fast as we'd like them to, although there are many things that are very promising. And, you know, there are so many other options out there. You know, there's so many other technical approaches to this thing. You know, what is the state of California gonna do? What should the state of California do? Interesting, you're looking at me in the reference of the state of California. So I'm not gonna represent the state of California, but I just wanna point out a couple of things, right? So actually, it's actually a good thing that we have some diversity and lots of pilots and trials, right? Because if basically picking a technological winner, you know, that's not the United States, that's definitely not California, okay? That's a defunct system, okay? But the question is what can we do to nudge things along in certain kinds of ways? What can we do to promote things? And, you know, there are examples of good policy and bad policy, okay? So trying to pick a technological winner, that's gonna be bad policy. If we decided in 2006 that, you know, linear solar thermal was the winner, it's like we never would have gotten this. You know, we never would have gotten what we have today in telecommunications. So the question is, how do we have lots of small bets, right, and lots of small losses? And then we actually learn a lot, right? So you think about this as a dynamic learning problem, you know, how do we structure that? That's really hard in the policy space to structure things so that there are lots of small bets and a lot of learning before big bets are plunked down. You know, where's the ribbon cutting for that tiny pilot? Ribbon cuttings are for big, you know, massive programs. And the challenge is when, you know, keep going in a certain direction when the signals point elsewhere. The issue is actually cutting off the losers early enough. It's not that they're losers, it's about learning about it and cutting off those losers early enough and moving to places that seem more promising kind of on the margin. So I try to think of this as, we do lots of small, frequent, incremental things to learn. That's really hard in the policy space. So from the industry point of view, what I see that if we can offer more value to the customer, we can assure the wide scale adaptation. What kind of, I mean, there are several ways that you can create revenue from a storage. There are a few of them that makes no sense. It's paper looks good. But one of the things that was another panel earlier talking about the micro grids and basically UPS and providing the power for the essential resources. This definitely gives more value. I mean, you can consider that you save money because of your bill. You can participate in the eliminating some peakers or helping the dock shrinking. And the top of it when you have a power outage which is fortunately we don't have very often in the PG and the territory. I'm glad you guys have experienced good service. But it's basically like buying an insurance policy. I mean, we heard a few examples how much economic damage can cause by a few hours of power outage or a hospital or a fire station and things like that. If you can provide, the storage provides the power for the few hours that has a tremendous value and that against just can justify the investments but the customers have to put it in. And this is again, it becomes a challenging sociological problem. Who here owns a house? Raise your hand, okay? Who here has earthquake insurance on your house? Keep your hand up, okay? So maybe a fifth, you know? And so, why do you have earthquake insurance or why you do not, right? Why does one install a micro grid or not, right? And what are those preferences and what is one assuming about what happens in that kind of catastrophe? And what can we learn from the experiences in California but also from Superstorm Sandy or what's happened to Puerto Rico, right? In terms of the resilience needed. And so I talked about, you know, safe and reliable. And so, you know, when we think about climate change, resilience really becomes the dimensionality which that shows up in the 21st century. And we're really trying to rethink that. What does that mean for customers? What does that mean for the grid? And so again, that's one where lots of promise for storage and we're trying to check it out. It goes back to the end of the day. What are the benefits as well as what does it cost? You know, a multiple set of benefits. I appreciate, you know, you're really careful and not double counting certain things. And, you know, there's been a lot of work out there on benefits of storage and a varying quality, shall we say. And so, you know, getting a good quality curated work. Actually, that is an opportunity for those of you interested, you know, to kind of demonstrate that and really test out what are those benefits and how, and under different market circumstances, they get greatly different value. We need to recognize that. Roses cost more close to Valentine's, just before Valentine's Day, compared to just after Valentine's Day, you know. And so a power plant that's providing energy is really valuable towards the height of that duck on the neck, but really it's actually negatively priced. You should pay to stay on the grid. And folks actually do, in the middle of the day, about 5% of ours in the first six months of 2017 were negatively priced. Okay, so it's like people are paying to stay on the grid with their power plants. And so storage has a real opportunity in that space to kind of provide a balancing. And the question is, does it pan out those benefits and the prices relative to the cost? Great, I'm gonna open it up for any audience questions that might be out there. If you want to come up to the mic and just ask a question, then we'll cover it in orderly way, please. John Mashi, actually two quick questions. One is, what effects do you see from the 2020 net zero housing rules that are coming? And the second is, I don't know if people have visited the central energy facility over here at Stanford with giant tanks of hot and cold water and very sophisticated computer modeling. I'm just curious if you're seeing that, the use of a storage that's heat and cold storage showing up in other places. It, I can't besides things industry or in homes. I want to take the second question. I'm gonna take a stab at the net zero home. So I think you're mostly speaking about title 24 changes. So as an industry, we're processing that, honestly. To my knowledge, the primary thing that we're seeing is that as we start to work with home builders, they are an interesting lot. They are extremely cost-conscious. Believe it or not, the profit margin on building a house in California is very good. So their desire is hyper-economic. They are going to do exactly what they need in order to meet the title and they probably aren't gonna do more unless homeowners demand it and are willing to pay additional. So what that means for us in the near term is that we're seeing smaller solar. So all the houses will probably have some sort of solar system on them, but they'll be smaller. The requirements around combining that with storage, however, I don't see any in there unless somebody has read farther than I have and can point to it. So right now you have a situation where title 24 is gonna require essentially more PV, but there's not a requirement for storage to go all over with it, which is from my point of view, a little unfortunate. You do get credits for combined heating and cooling. So there's a desire for heat pump, water heaters combined with their conditioning and those are all good things. But from my point of view, it could go a little bit farther. And one of the key elements is zero net over what time period, right? And so this goes back to the role of the grid, transmission and distribution in integrating this because basically the more houses will see like this, will actually see more two-way flow, more customers with two-way flow, sometimes sending power onto the grid and whatnot. And so that just becomes, again, it wasn't originally designed that way in the 20th century, but it becomes a 21st century challenge that we're facing now and that Hawaii is definitely faced and that we're trying to integrate these issues and these resources into the grid. So I haven't seen the Stanford, I would love to see it. I heard the first week that you went to the lab. And, but there are other storage technologies that use by like grocery stores or they using pre-chilling and creating ice at night and basically pumping into the air conditioner and things like that. Hi, Bennett Meyers, Stanford Slack National Lab. Andres, you talked about one of the important factors being the increase in battery lifetime in terms of number of cycles. We've gone from 1,000 to 2,000 to 3,000 over the last few years. So I'm wondering if you could talk a little bit about the challenges of dealing with battery lifetime in like an energy management system context with renewables where you're not doing these nice charges up and down. Like how should we think about lifetime, particularly when that's kind of the published data is these number of cycles, we may not actually ever see cycles like that. That's a good question. You know, I used to have dark hair a year ago. It's very challenging because, and basically what I'm dealing with is the warranty issues, negotiating and dealing with the manufacturers. Manufacturers accept that you keep the battery in 50% SOC and say you make one charge, a full cycle, which of course shouldn't be 100%, maybe between 20 and 80 or let's say 10 and 90 and after keep it in 50% and they give you 3,000 cycles, right? That's not reality. And the other thing that they're asking us that please provide us a typical profile. Okay, we have 500 installations and we have 501 profiles. So what basically, what we try to do is in the last five or six years, we collected a huge amount of data and introducing the machine learning and try to basically control the battery and doing, it's overused smart algorithms that maximizing the battery life. For example, it's no point to charge up the battery to 100% SOC at night if you know that you're not going to discharge it till 3 p.m. Right, so keep it in 50% SOC, which is the optimum and you know that you can charge it up in maybe two hours so you start to charge it at one o'clock actually which helps the dock, right? Now it's, we still have issues that usually the one o'clock the rates and things like that but that kind of algorithm to introducing and that can extend the battery life. And the other thing is that the keeping tab on that how much you used for the total energy throughput and when we're talking about the virtual power plant, right, we have many properties that we can take a look that this battery already had a thousand cycles, that battery had only 800 cycles. So let's use more of that battery than this one. Yeah, I mean, I just wanna follow up on that as a great example because now we had a strategy and analytics team and so some several days to scientists, my team kind of looked at some historical data and operations. I wouldn't say with STEM but with a comparable company in that space and we found that could be way better in terms of optimizing the usage given the signals than what was happening with the customer deployment and the customer battery. So again, it goes back to technology and then kind of we can get even better lower costs with permitting and so there are things we do to squeeze out efficiencies using machine learning techniques and so forth and it's actually very promising and exciting. Hi, Amber Kerr, UC Berkeley. I hope I didn't miss it, correct me if I'm wrong but I don't think you mentioned the California Public Utilities Commission mandate for certain storage capacities in the investor-owned utilities. I was alluding to that when I talked about California policy. Oh, you were alluding to it. Okay, so my understanding is that currently CPUC has mandated that utilities have, what is it, 1,325 megawatts of storage and they just added a new requirement for 500 megawatts of quote unquote behind the meter storage. Yes, PG's requirement is 747 megawatts total from those two programs, that's right. 700. For PG&E. Oh, PG&E, right, thank you, okay. So I wanted to ask all of you, especially Mr. Strauss, do you believe that CPUC is on the right track in making this mandatory? Do you feel it's a useful way to put the right pressure on the system or do you feel that it is premature given the states of the technologies involved? And we already established there's no commissioners here so speak freely. Let me answer, yes. And so it goes back to what is California trying to do with climate change, right? California's about 1.5% of the world's emissions, right? And so it's basically saving some greenhouse gases and avoiding it in California. Does that make a difference? And so yes, but it's also what we do sets examples for China and India and the world and that's California's ambition. And it's not just Jerry Brown's ambition, that's California's ambition, okay? And so when we think about storage it's California's ambition to basically transform the market. How to do that, what's the best way to do that? As a matter of general policy, PG&E tries to avoid placing mandates on our procurement and activities, right? That looks like another constraint which basically doesn't necessarily yield the optimal in the eight dimensional space, right? In terms of efficiency as well. So they're probably better other ways to do it, but given that policy, we're out there trying to figure out how best to meet it in a low cost way but also in a great learning way. In fact, one of the challenges is the mandate is in action in terms of megawatts. Not kilowatt hours, which is the key else. So even the metric units for how this mandate was set up, I was like, there's some other things that could be even better, right? But overall it's like, let's take a step in that direction. What's the right size step? And so it's 747 megawatts for PG&E by 2024. Is that too big a step? It goes back to rough estimate might be for every three kilowatts, megawatts, I'll keep it utility scale of solar PV in the system, maybe might need a megawatt of battery, okay? Might be other kinds of storage. But let's say roughly three to one. So if you think about the kinds of quantities we're talking about, we're talking thousands and thousands, right? Of additional megawatts of batteries. So what's the right step? So that's a judgment call. At this point, I accept the policy. PG&E will comply and we're trying to do it in a way that makes a lot of sense for the learning as well as the cost. Just a quick comment on that. One of the things that fascinates me is how you go about mandating and causing these things to happen behind the meter. And right now, the primary thing that we're seeing is that California is by far the number one consumer of storage, as Todd mentioned, and it's because of the S-GIP program. So right now there's a very, very positive benefit to installing behind the meter storage and people are responding to it very, very clearly. How that fits in with what the CPUC intends for us is a little arcane, but it is what is happening right now. So we're at the end of our time, unfortunately. There's one more question. It'll also be a quick one and it'll give you a chance to speculate. With all the electrification of transportation and all the talk about vehicle to grid storage, I wanna hear you comment on the economic model of using vehicles for this kind of load storage given that someone who buys a car and buys the battery may have a different economic incentive than a utility operator who wants to use it for storage and cycling. So I hope you'll think about what models make sense for this. And I'll give them a pause if you can think about it for a moment, but clearly it's like there's a marketplace model, right, in which basically you have a variety of kinds of actors buying and selling voluntarily to do that exchange. And so what might that look like when folks are not mandated to provide that battery storage energy at times what price signals are needed there. Or again, to the extent we're talking about technologies, what codes and standards or technological devices. So there are a variety kinds of policy levers that might be able to be done there, but that clearly is a huge potential, right? When we think about two-way bat flow from storage and transportation, you know, huge, huge potential. The question is how to tap it technologically, how to tap it kind of sociologically, institutionally, and how to kind of make it work for customers in a market system. So yeah, so there's from my point of view, there's two dimensions to that. The first one is actually very easy to solve conceptually, but it's going to take a while, which is creating standards so that the utility can interact with the car, right? So the reality is that bolts don't speak the same language as Teslas, don't speak the same language as Nissan's. Getting all of that sorted out, this is what organizations like the IEEE do for breakfast, lunch, and dinner. So we will get that sorted out. The next problem is actually a more interesting one from my point of view. Remember, I was saying that to get into smart energy, you have to make it very seamless for the consumer. So the very practical use case is, you know, I have my Tesla, I go to work in the morning, I come back, there's some excess energy in my car. Am I willing to forego the opportunity to go out at 10 o'clock for whatever reason, I would go out at 10 o'clock at night in order to power my house because I've been incented by a very, very powerful time of use rate structure that says that it's going to be a lot cheaper for me to discharge my car than it is for me to buy it off of the grid or better yet, even to take that energy that's in my car and give it back to the grid. Trying to come up with a way to understand the use patterns and people's tolerance is going to take some time, but it's a very obvious use of this vehicle to grid and we just have to figure it out. Anders, any final thoughts? My point is theoretically, it's a great idea in 2018. I don't see that it's going to have a broad, widely accepted application in the next four to five years. Great, well, thank you very much, everyone, for staying even a little bit longer. Glad to be here. Please thank the panel.