 For this euphoric event post pandemic and this is terrific to see everyone in in person and thank you for doing this. I also want to give a shout out to Jay Precourt. And Jay has been the, I don't know how many words I can use for his support for energy issues at Stanford and even broadly. So thank you Jay for your support over decades to Stanford and greatly appreciate it. So before we start I just also want to give you an offer some hot off the press news. This is really hot off the press and this is very important news. Is the fact that each way our own fearless leader just got elected today to the National Academy of Sciences. The highest honor. It is the highest honor a scientist can get. There are only a couple of thousand people around the country and around the world actually. It's part of the NAS of the United States National Academy of Sciences. Something that was created by President Abraham Lincoln right in the middle of a civil war. And so that's the history of science in the United States and just congratulations again. Okay, so now we have a very important panel. Let me invite the panelists to come on board Dan Mateo Andrew Fong and Clyde. And this is on long duration storage, which is probably one of the most important things for our electricity system. And if there's a clicker out here, you want me to sit in the middle? That's fine. We'll figure this out. There you go. Musical chairs. We're going to do this each round. So this is a panel on long duration storage. I would say one of the really heavy and demanding issues that we face today and we are likely to face even more in the future. Just to set the discussion, here's just a couple of slides. I'm going to borrow a slide from Bits and Watts, which is Thursday and Friday. And say a little bit about the grid because this is about the grid. So this is what the grid has been for the last 100 years. It's the architecture that was developed by Tesla and Edison. The architecture has not really changed. And the paradigm is you've got turbo machinery generating electricity with lots of inertia in the system, going through transmission lines to substation to distribution network and to your customer. And so one way traffic and the balancing is done in real time with frequency as your proxy for imbalance. And it is still used that way today. We are now, and given this paradigm, the jurisdictional sort of boundaries were created around this paradigm to say that this is now, at least in the United States, the wholesale markets of trading electricity and all is regulated federally because they exchange electricity across state lines. By the grid system, this is the Federal Energy Regulatory Commission. So that's on the transmission side, on the wholesale side. On the retail side, like PG&E, and we have a member from PG&E out here. It is state regulated electricity utility. And then, of course, the electricity is delivered and it's a one-way traffic. That's been the paradigm for 100 years. It is changing and it's changing dramatically. Now we are trying to inject more than 50%. People are talking about 80%. And some people even talk about 100% of solar and wind and storage. And we'll see why. And the grid was never designed for fluctuating generation. And you also get these things like duck curves, which is essentially have so much generation in the middle of the day that you have to curtail some other things or curtail the generation itself. And you can have all kinds of issues on the grid. And then in the evening, when the sun goes down, you get a massive ramp up. And so this is on a daily basis. And we're going to hear about it. And there are other issues that we'll hear about. This is on the generation side, on one end of the grid. On the other end, you've got network in a thermostats now that can be turned on and off remotely by just changing the set point and you have EVs coming on whose loads are about ten times more than a single home. A single EV when you're charging in level two is about five to ten times that of a single home on an average. And you've got other things that are going on distributed generation from solar, et cetera, and some storage. And so now these are all connected to the cloud. And people are now with FERC, some of the things that have gone on the FERC, new orders, et cetera. You could connect them to provide services along with storage services on the grid. This is what is going on on the grid. And there's a lot of changes that are happening. And the regulatory environment, the technology are all kind of sort of coming along at the same time. So here's the grid was never designed for that much of solar and wind. A little bit of solar and wind is fine. Trying to get to 80% is a lot. And so here is, I'll just offer this to you. And this is a wonderful paper by Albertus et al. That you could take a look at it, but I'll just summarize it. That on the x-axis is the electricity, annual electricity from solar and wind on a regional grid, the percentage from solar and wind on an analyzed basis. Not a dynamic basis. We'll get into dynamics on the panel on an analyzed basis. And on the vertical right axis is the maximum amount of storage that you may need. And you can see it's a log scale. On the y-axis on the left-hand side is the maximum storage capital cost in dollars per kilowatt hour that you would need. That's kind of the target. And so as you go to 80% penetration, you will have some occasions where you'll need 100 hours or more of storage at the same time. And then maybe it happens about 10 or 15 times a year. And if you go to deeper penetration, you'll get even more. And then there's seasonal storage. And I'm not even talking about that. And so if you look at where the cost metrics are, this is kind of roughly. These are hand drawn things, but roughly, this is where the cost structure is. So as you go to higher and higher penetration, the cost of storage has to come down. This is not used enough. This is where lithium ion is kind of today, but roughly $100 a kilowatt hour capital cost. The fully installed may be a little bit higher. And this is perhaps the bottom of lithium ion. Maybe there'll be innovations to bring it down even more. Pumped hydro, we're going to hear about pumped hydro compressed air storage. But you can see that there's a whole white space below that. How do you get to below $10 a kilowatt hour, which is what has been claimed by this paper and based on today's markets? Markets change. The performance requirements also change. So here is where we are today. This is the requirements. This is the need. Now what we're going to do is we're going to start off with the needs for storage. And I'm going to ask Clyde from California ISO to talk about some of the storage issues that he is facing. And then we'll get into the technology, or you already got it. So good afternoon. It's really difficult to start off after lunch. We had two very, very interesting conversations or panels this morning. So I'm going to bore you a little and shift the conversation now to reliability and how we operate the grid. So this first slide here shows you the need for storage, multiple day storage. When you look at it, California has today a little over 14,000 megawatts of grid connected solar. We've got about 12,000 megawatts of rooftop PV. So that's 26,000 and about 6,000 megawatts of wind. The load on a weekend, it's about 18,000 megawatts. So without doing the math, you can see the need for storage. Now, if you look at the left, it shows you we had multiple days where it was rainy. It was cloudy. We had very little solar, which has a very top-down yellow. We had very little wind. So we had to rely a lot on the gas units. And if you look at January, January is not a lot of snow. Not a lot of hydro. So on the left, you can see you've got to use those thermal plants quite a bit. When you look at the right, the right is more mid-time frame. You've got a lot of snow melt, which is that blue area here, a lot of hydro. A lot of solar, a lot of wind. You've got to back the thermal fleet all the way down. Now, when you're backing your thermal fleet all the way down, you have a lot of challenges. You've got to worry about frequency control. Today, in 10 minutes with the amount of rooftop PV and the amount of grid connected solar that we see, grid changes by about plus or minus 1,000 to 2,000 megawatts in 10 minutes. That's a lot of variability. It's not like, you know, you look at OCOT, they have a lot more wind. I would wind geographic diversity minimizes variability. It's just the opposite with solar. So we see a lot of variability. The question is, how can you predict that? Because in the old days, remember, you had controllable supply. You had predictable demand, and it was easy to control. Today, we have variable supply and unpredictable demand. And it's a challenge. So from a grid perspective, it's very, very difficult when your system is changing, plus or minus 2,000 megawatts in 10 minutes to predict where you're going. So you see a lot of reaction today. You're kind of reactive, right? The next slide shows you, like on a daily basis, what we see. So this is a typical weekend. And a couple of weeks ago, we curtailed 5,000 megawatts, which is that red. 5,000 megawatts is equivalent to about 5 to 6 million homes. That's a lot of energy. Phong's going to go into why that's not a nice thing to do. So when you operate with a gas fleet, really, really low, downward flexibility is a problem. So if wind kicks up or solar kicks up, you have nothing to really back off. So that causes high frequency. It causes a lot of control problems. And one of the things a lot of folks do not realize is each balancing authority, like the ISO, we have an obligation to control the system frequency, and we got to do that every minute, right? We also got to control supply and demand every four seconds. That's not an easy thing to do. So when we look at storage, we need a lot of different types of storage, long-term storage, short-term storage, storage that can provide frequency control, and other things, right? Now, I know Arun had mentioned that we're striving by 2030 to serve 60% off the load in terms of energy. But I want to tell you, last Saturday, we served 99.8% off the load in California for two minutes with renewables. So renewables includes wind solar, geothermal biomass, biogas, these very small hydro less than 30 megawatts. That's a big accomplishment. It's almost 100% off your demand with renewables. So we can do it. For two minutes. But still, it's a challenge, right? So we know we can get there, but it takes a village. A lot of research has to go in. I heard a lot of encouraging discussions this morning, but there's a light at the NA tunnel where I think it's doable. So a lot of things we can get into. But I think we can make this happen. Thank you, Clyde, and for your sort of laying the groundwork from the grid operator side of balance, where you have to balance the grid on every four seconds. Let's move on to Phong from the point of view of a utility that is buying electricity and distributing. Go ahead. Good afternoon, everyone. Thank you for inviting me here. My name is Phong Wan. I've worked for PG&E for a long time, over 30 years. And I'd like to borrow a room from your slides a little bit if you could. Can we go back to this slide? Let me see how things used to be. I keep pressing this. It probably doesn't work. I mean, there's someone out there who is... Can you go backwards? More to my slides? Your slide. Yes. One more? One more? Well, we need some pictures. What? Thank you. This one? Okay. When I started PG&E, it looked exactly like this picture of Rune described. Except there wasn't really a wholesale market. That means I get to plan the entire generation portfolio and dispatch all the units exactly when the customers needed the energy. It's exactly what Rune said. I thought the job was very difficult, but in retrospect, it's not. It's because the electricity flow one way. And I could base load our renewables. And over the evening hours, I will use our gas units and dispatch them up as well as use our wonderful pump storage that I think Dan's going to cover later called Helms, 1200 megawatts. And how the world has changed is that we put in a California ISO, which my friend here represents. To put things in context, PG&E is about 45% of the ISO's load. So there's also two other major utilities in Southern California. What the ISO does is essentially the left-hand side. It runs the platform on power flow as well as energy markets. Let's go to the next slide. Where we sit today, PG&E said... One more? One more. Yeah. What we sit today is that... One more. One more. Okay, perfect. That's what I see today. And it's really challenging because we're still trying to take the wholesale power from that market on the left side. But I also have a lot of customers who we like to call them prosumers. I think you came up with that many years ago. And the consumers are using a lot of rooftop solar that we talked about earlier and putting the power back to PG&E and we actually put it back onto ISO to learn the middle of the day. That's really what the duck curve is. The reason is that these consumers are not using electricity over the peak of the day. And they're banking it with either PG&E or the ISO and they're drawing back the power at night when there's no longer any solar. And we also have lots of Tesla power packs that Mattel used to work on in his previous career. As well as we have a lot of electric load. As Roon said earlier, each EV, the way I think about it is about two air conditioners at your home. So that's about the load it takes. And every time they charge, I have to worry about big spikes. So we have a lot of interactions between the left side and the right side. And that's what PG&E is in the middle of trying to manage. My role over the last 15 years is that we probably bought more renewable energy than any of the other companies in America. And we are very fortunate as Californians because we have more renewable resources than anywhere I've seen in the world. We have great wind. Hopefully one day we'll be able to build offshore wind to complement the onshore wind. We have great solar resources, especially down in the desert as well as the Central Valley. We have a lot of geothermal. We have earthquake vaults. And then we also have a lot of biomass. Biomass made of either forestry waste, agricultural waste, or urban waste. We have more resources than anywhere. And that's why we say we are confident we can get the renewable numbers up. And the limit is not renewables. Renewable prices have come really down. Early in my career I bought some renewable contracts that were really expensive. When I look back they make me cry. But renewable prices have come down tremendously. We, California along with Germany, I would say we were responsible. And we contributed greatly to the drop in price that affordable takes. And everybody's benefiting from that today. Most of my renewable contracts were signed with fixed prices. Early in our purchase era, we were trying to get as much as we can. So if you produce a kilowatt hour or a megawatt hour of whatever unit you want to use, we will pay you according to that. Then pretty soon we began to figure out, hey, these renewables are not generating at the time we want. So we're building dispatchable features to turn down the power and try to get them to generate according to the low profile that you saw earlier in the duck curve. But the next step now is really we're facing a lot of renewable curtailments. That's what Kaly showed you earlier. For the contracts in which I don't have the ability to turn down the units, when we curtail them we still have to pay them. Because they were available to produce electricity. What Kaly has not shown you is that we also interact with the rest of the Western grid. We are shipping a lot of our excess renewable energy sometimes at very low prices, sometimes even at negative prices. So people would take the renewable energy off our hands during times when we don't need it. We don't consider that curtailment, but it's still pretty painful to pay someone out of one pocket and pay your neighbors another pocket in order to make the whole energy system balance. So we have been very involved in buying storage. I mentioned to a room of my fellow panelists that last year we brought 1000 megawatts of lithium ion battery online. And those are four hour durations. And this year PG&E will bring another 500 megawatts. And the question, natural question to set the rest of the discussion is should we buy more lithium ion? I would say maybe a little more, but it's time for us to switch to longer duration. Resources to complement the rest of the portfolio to make it work. What do we need? We need batteries more than four hours. I don't know if eight hours is the right answer. I know eventually we'll need it across days of the week. And that's because Saturdays, Sundays have much lower loads. And we will need it across the months. As you know, from June to September, that's our peak load. And then we'll need it across the season. So these two gentlemen are going to tell you how it's going to be done because I'll be sitting at the other end buying it from them one day. Okay, so now that you've heard about both the big challenges and frankly laying out the opportunities if someone can address this issue, now we'll go into different technologies that could potentially solve this problem, address these issues in a competitive way. So let's start with Dan Riker on pump diagram. I speak better standing up. So thank you Arun and Will. Congratulations, E, and pleased to be with you all today. I'm a senior scholar at Stanford Woods Institute. I was founding executive director of the Steyr-Taylor Center for Energy Policy and Finance here at Stanford. And also was U.S. Assistant Secretary for Energy Efficiency and Renewable Energy and spent several years at Google working on all these wonderful topics. I'm here to talk about the sleeping giant of long-duration electricity storage and that's pump storage. Some quick facts. Pump storage today provides almost 90% of U.S. electricity storage, 22,000 megawatts, 550,000 megawatt hours. Current U.S. pump storage projects, and this is interesting, they were built in the 70s and 80s to store excess nighttime electricity generated at nuclear power plants. The old pump storage is what we call open loop. It involved a dam on a river and a reservoir at the top of the hill. Modern new pump storage does not require a dam. Both reservoirs are built off-river, so this is called off-river closed-loop pump storage. It locates, and the wonderful attribute of off-river closed-loop pump storage is that the environmental community is increasingly supportive of it. The river conservation community because you're not building a dam. Pump storage, as we just heard, provides large quantities of long-duration storage measured in days or weeks with the capacity of individual plants generally measured in the hundreds or thousands of megawatts. The largest current pump storage project is over 3,000 megawatts and is a facility in the state of Virginia. Pump storage is getting a second look. The sleeping giant is waking up for several reasons. It's new low-impact off-river closed-loop configuration. Its ability to store large quantities of variable solar and wind, not excess nuclear, but excess solar and wind, and the many key attributes it provides to the grid. Long-duration peaking capacity, critical balancing services like ramping, operating reserves, regulation, energy arbitrage, and it's one of its least understood but very important is black start capability. We've had blackouts where it was pump storage that helped get the grid going again. There are about 90 proposed US pump storage projects representing about 80 gigawatts of capacity in the development pipeline. A small number of those actually have FERC licenses. The rest are in earlier development stages. I got interested in pump storage because of something we've been doing here at Stanford, which is called the Uncommon Dialogue. In October of 2020, we got the long, warring hydropower industry together with the environmental and river conservation community and said, let's see if we could actually find some common ground or, as I like to joke, some calm water. We negotiated for two and a half years and in 2020, October of 2020, reached a major agreement. New York Times wrote a feature story and that was how to advance the climate and energy benefits of hydropower, including pump storage, and the environmental and conservation benefits of healthy rivers. The core of the agreement was what we call the three Rs. Let's rehabilitate some of the 90,000 dams in the United States for safety. Let's retrofit some of them for power. It turns out only 3% of US dams make electricity and let's remove some for conservation as well as safety. We've got a number of working groups now moving forward. When President Biden was elected in November 2020, he said infrastructure, let's do infrastructure. We got back together again. We negotiated for another several months. We reached another agreement that said we need a lot of billions of dollars to implement the three Rs. at the 90,000 dams. We succeeded in getting $2.4 billion in the federal bipartisan infrastructure bill to begin a down payment on the many tens of billions of dollars. We need to do all of this. We went on and last month reached a third agreement on how to improve Federal Energy Regulatory Commission licensing and relicensing of dams and pump storage. And it's this funding, the $2.4 billion, and these improvements in licensing that we're proposing that has improved the prospects for US pump storage projects. In particular, closed loop off river pump storage. At the same time, pump storage has its own challenges. The need for water in a time frame where some US regions like here are facing drought. The good news on that is this off river closed loop recirculates the water and tends to lose very little. Pump storage projects do face major upfront capital costs. They can face extensive time frames for development and valuation issues like all storage in the face of electricity market uncertainty. And pump storage as a form of hydro power still is not fully accepted in the environmental and river conservation community, but is enjoying increasing support. So, but overall, I think the sleeping giant is awakening and found a more welcoming environment for US closed loop pump storage. And the bottom line for me, and I think I'd encourage all of you to think about, is the closed loop off river pump storage can and should be a significant tool in US long duration electricity storage. Thanks. And our agreement and all that we've been working on is featured in this month's Stanford magazine. It's the cover story so you can read all about it. Thanks very much. Thanks. Okay, so we heard about pump storage. Now we're going to hear from Mateo Haramio on the new type of electrochemical storage. Yeah, thank you. The pump storage? Yes. Do we do the Q&A now? Why don't you give a quick answer and then we'll come back to a more detailed discussion. I love pump storage and Dan's right. The issue here is actually the certainty of construction costs. This station has not, as far as I know, done a new pump storage in a long time. And the way the business world works is that if I was to sign up for a project, a big pump storage project, I would think it's at least $2,000 per kilowatt, which will put it into billions and billions. Right? I want to know cost certainty just like you would if you were to buy a house. But very few sellers are willing to give me a fixed price deal. And I cannot take that to my regulators or my customers with an unknown cost structure. That's the biggest problem. And no seller wants to eat the possible cost overrun. This reminds me of how this nation built nuclear plants three or four decades ago and now too. Lots of uncertainty to build very capital intensive projects with which almost no one is used to building and with a fixed cost. Fantastic. Well, the benefit is so great. I hope you'll sell it. I think if the state of California is willing to absorb cost and certainty, we can do it. Let's come back to that discussion. This is a very important discussion. I'll give you a follow-up when we come back. Okay. Mateo. It's a big problem. And I've been in batteries for almost 20 years now. Energy storage, probably speaking. In batteries, as Fong mentioned, I was at Tesla for about eight of those years. My good friend, Vinit, there in the back. And when I left Tesla about five years ago, that chart that Clyde shared is exactly the one that I started to take a close look at. What happens when you have, as we do have on the native California, multiple days of rain in the winter to the electric system? What steps in? What is providing the power in that case? And obviously overwhelmingly, it's the thermal resources, right, increasingly. And the thought experiment that I started out with was what kind of energy storage would allow you to replace the function of those thermal resources during that kind of event. So setting aside any technical approach, just describe generally what you would want out of that. What do you need? How many cycles do you need? Let's assume it's electrochemical. What is the efficiency you think you're going for? What is the duration that you really think you need? And sort of starting from that position, I've started to identify options that exist in the world. They may not have been commercialized, but they were options. And we zeroed in, a couple of co-founders, we zeroed in on exactly that crossing point between 100 hours and less than $20 per kilowatt hour. In other words, $2,000 per kilowatt, as Fong said. And it turns out there are some options that allow you to get there. Now, the question is, of course, electrochemically, what are you willing to trade off to get to that capex? There is no battery in the world. In fact, there is no form of energy storage in the world that does not involve trade-offs. There is no holy grail. It doesn't exist. Don't look for it. Try and find the right thing that fits the application that solves the problem that you know needs solving. And so for us, that was really this notion of multi-day storage. You have to cost-effectively provide that function into the system. And again, 100 hours, $20 per kilowatt hour. Of course, there are trade-offs to be made. And principally, I'll give you two examples. One of them is cycle life. If I am trying to bridge a week's long gap in the system and provide energy storage that is cost-effective over that duration, the very simple thought experiment, of course, tells you that theoretically, I could only get a maximum of 26 cycles per year, right? One week to charge, one week to discharge, 26 cycles. Therefore, over the life of a piece of infrastructure, let's say 25 or 30 years, what I'm talking about is in the low hundreds of cycles per year, maximum, theoretically, that I could get. Nowhere near the thousands of cycles that lithium mine is out there getting today, much less unlimited cycles as some battery chemistry would have you believe today. In other words, I can trade away cycle life in pursuit of my much lower capex costs, right? That's point one. The other point is efficiency. And we talked about the curtailment of the renewables there. We are spilling electricity all the time. In fact, probably right now in California, we're curtailing some solar, my guess. And that is fine. In fact, that's the economically rational thing to do, and will increasingly be the thing that we are deciding is okay. And the major signal there is that the input fuel costs are now, from renewables, are now the cheapest possible fuel costs to anything that you could hope to do. So setting aside the broad transition to electrification for that exact same reason, the same logic applies to electricity storage as well. You should be giving up the right amount of efficiency in pursuit of the low capex costs. And how do we know that's true? That's what the thermal industry did for 100 years, right? Coal and natural gas were the cheapest input fuels that you could possibly have, the efficiencies on those chemists, or on those chemistries, that's how I think. On those fuels is terrible, right? Thermal efficiencies are terrible when you're trying to make electricity out of it. And so applying that sort of mindset to a battery, we now can pursue something that doesn't look like anything that's ever been commercialized to date. It doesn't work for a car. It's never going to work in a laptop. But it is relevant for this moment that we're in the grid today and for the future of the grid that we're going to. 100% is just the beginning in terms of the penetration of renewables. I'm totally convinced it will get to 100%. There's 480,000 or so minutes left to solve, but we'll get there by the course of the year. So that's what my company does. That's what form energy does. We are commercializing this multi-day form of energy storage. It's an iron-based chemistry. It's an iron-air chemistry specifically. So we are reversibly oxidizing the iron. Our active materials are less than $1 per kilowatt hour, which is what you need to build a system that's less than $20 per kilowatt hour. And to give you a sense for the inversion of how these costs dominate, that active material cost in the end is less than 10% of the system cost, which means you have to be very intelligent about how you design your system and drive your inactive costs as low as possible because they dominate the system. But I'm happy to report that we are today very much along the path to commercialize exactly that, where a year away from our first utility project, two years away from our first material utility projects, these will be in the tens of megawatts, going to hundreds of megawatts by the middle of the decade. And all of this is in service of, to be very clear, a reliable, renewable, affordable electric system. Increasingly, that's where the attention is being put, is what do you need to enable your reliable system? And you're starting to hear this from PG&E, from California Public Utility Commission, other regional markets, where they are describing exactly this reliability imperative that has to go along with the decarbonization of the electric system. Our society simply is not willing to tolerate a less reliable electric grid just because it's green. That's a trade our society will never make. So that's the opportunity for us to step into, right? The alternatives are running the coal and the natural gas. Maybe we trade coal for natural gas increasingly as we've done, but there's a limit to what you can do that. And most importantly, the fundamentals of the renewables are there, that they are the cheapest marginal resource and we will figure out a way to store that. So we're deeply involved in the proceedings here in California, but really around the world, to think about this new kind of asset class, this multi-day storage and how it fits into the electric system and delivers that reliable, renewable, affordable electric grid. Great. So you heard two different, entirely different technologies that we'll have to compete at some point in the market. And now we're going to talk about an entirely different third, a different technology. Andrew? Yeah, hi everyone. It's really an honor to be here. Thanks so much. So I'm here representing Antora Energy and we are getting even more different than what we just heard as far, but still in the realm of long-duration energy storage. So our mission is really to decarbonize industry. And one of the things that's interesting about industry, besides the fact that it's over 30% of global emissions, is that more than twice as many of those emissions come from generating heat as from generating electricity. And so if you want to decarbonize industry, you really have to look at both halves of that problem. But what's interesting is the problem is very similar to the one we're just talking about purely on the electricity grid. If we think that the primary energy source of the future is going to be inexpensive electricity from wind and solar, then you have to have long-duration energy storage to turn variable renewables into consistent reliable electricity. And you also need long-duration storage to turn variable renewables into reliable heat. So you have to do both. It's the same problem, the same sort of durations matter, because really what you're solving for is weather. You're trying to solve for those multi-day events where you don't have that much wind or solar electricity available. So Antora is addressing both of those problems for industry, both the electricity and heat side, with a new type of thermal energy storage. So the way our system works is you take that inexpensive electricity when it's available from wind and solar, and you use that to resistively heat cheap carbon blocks until they're glowing white hot. You're storing a tremendous amount of energy in the sensible heat of carbon, and they're very, very hot blocks, and you put them within an insulated box so that heat isn't leaking to the environment, even if you want to store it for days or weeks. Now, to get that energy back out, we have a very different discharge mechanism than anything you'd see with electrochemistry or anything else. We actually open up an insulated shutter on the side of that box, which allows a beam of white light coming out of this ultra-high temperature furnace to exit the system. So once you have that beam of very, very bright light coming out of the side of this thing, you can do one of two things with that. The first thing you can do is you can use that to generate heat for industry, so you could use it to raise steam, for example. The other thing is a little bit more unique among thermal energy storage. So because that stored thermal energy is already coming out in the form of light, you can use a photovoltaic cell, very similar to a solar cell, to convert that light directly back into electricity, one step, no moving parts, very inexpensive and scalable. So that's sort of the unique aspect of Antora's system, a very inexpensive system that can absorb renewable electricity very, very quickly, store that for days or weeks, and then output reliable heat and electricity for industry. One thing that is slightly different than what Matteo's talking about, you mentioned kind of 100-hour charge, 100-hour discharge, because our charging mechanism is totally different from our discharging mechanism. It's just resistive heating. We typically spec our charge rate to be four or five times even faster than our discharge rate. So we also typically look on the discharge side at 100-hour type systems, but we can also charge within, say, 24 hours. And so in a system where you might only have a few hours in the day when you have a lot of wind or solar that would otherwise be spilled, we're sort of very able to absorb that power quickly off of the grid when it's available. The other thing I'll mention really briefly that's kind of interesting about being able to discharge both heat and electricity is a little bit of a solid point. So one thing Matteo mentioned was it's the rational thing to have some curtailment. So if you look at almost any analysis of the electricity system, as you get to high penetrations of renewables, you do two things. One thing is you overbuild your renewables to some extent. So you're going to have more wind and solar than you need a lot of the year in order to make sure you have enough in those weeks or months where it's more precious. The other thing you're going to have is you're going to have a lot of long-duration energy storage that's not being fully utilized. It needs to be there for the few days or weeks when production is in that high, but the rest of the time you're only using a small amount of its capacity, if at all. So one of the things that Antora is kind of uniquely able to do with a thermal system is take that same asset, the same storage, the same renewables, and during the, you know, let's say 80% of the year when you don't need every ounce of that for keeping the electricity system going, we can use that purely as a thermal system. So the same capital assets can be used to output heat for industry with essentially no additional cost versus what you already built for just an electricity storage system. So this is what makes the economics sort of unique. We're still targeting the same, you know, kind of 100 hours less than $20 per kilowatt hour, but we can offset some of those costs additionally by selling heat out of the same system without losing anything because that was otherwise curtailed or underutilized capacity. So this is a very, very different system, but I think it's so super interesting that it's the same problem of electricity storage just also applied to heat simultaneously within the system. Fantastic. So you have three different competing technologies which will have to compete with compressed air storage and hydrogen as we heard yesterday. So this is an amazing time to be living and sort of witnessing all of this going on. Let's go back to the conversation that was started earlier by Jay and, you know, you, Fong, you commented on some of the challenges. How would you respond? So I think Fong raises an important point, which is, you know, these are big, costly projects with long time frames. The good news now is given the urgency of electricity storage given this newer approach to pump storage, which is all off river, much stronger environmental support, is we now have big investors looking for these projects to buy. And in fact, Copenhagen Infrastructure Partners, the most successful investor in East Coast Offshore Wind, the leading bid just recently bought the two FERC licensed close loop off river pump storage projects, one at a thousand megawatts and another at 400 megawatts. So I think the difference may be, and time will tell, we have a much bigger investment community out there with a lot more money looking at a very large problem and I think pretty convinced that we've got a technology that can work. Still, still got to be proven. We've still got to show that, you know, you can get big stuff like this built. The other thing that I'd say, though, is we now have a Federal Energy Regulatory Commission that is very interested in simplifying their process for licensing new big facilities. The agreement we reached has been welcomed at the Federal Energy Regulatory Commission. The chair of the FERC called it a stupendous agreement. There'll be a hearing on Capitol Hill next week. It's got bipartisan support. So there could be, there's a lot of money available. There's a streamline permitting process that we could see put into effect either from a legislative or regulatory standpoint. So I'm relatively bullish that some of these projects will get built. All 80 will not, all 80,000 megawatts will not, but I think a decent number will. And I think it will be part of the larger package of storage. We need all these different ones. We cannot rely on batteries by themselves. We need multiple battery chemistries. And by the way, as you think about the capital cost of a pump storage project, it has a 40, 50, 60 year life. Obviously batteries have to be changed out quite frequently. So there's a lot of apples and oranges here that we got to get our arms around analytically. Dan, I completely agree that you need multiple technologies, multiple options depending on the need. Tell us a little bit about the risk. I mean, the technology seems, I mean, it's been around for a long time, right? There are incremental improvement, et cetera. But is it the cost? Is it the risk or the uncertainty? Is it the permitting issues? What is the major risk? I think there's elements of all of those. And so it's going to be smart developers in an improved regulatory process with sophisticated investors that I think are going to prove that this works or not. And the next couple of years are going to be very telling because we do have FERC licensed projects with large-scale investors behind them. And I'm hopeful we're going to have a FERC that says yes. And one more element of this. We have multiple billions of dollars coming out of the infrastructure bill ready to help buy down the cost of some of these initial projects. We have the loan guarantee program at DOE saying, bring us a closed-loop pump storage project. Don't bring us 50 of them. Bring us a couple because we want to be part of proving it out, too. So there's this wonderful intersection right now that I think will, and we'll prove it out one way or the other in a couple of years. Jake. I'm going to answer that question because I've been involved for a long time with investors in pump storage, and it is a fantastic project business-wise. In the past, the hold-up has been the environmentalists. A lot of environmentalists didn't want to see a dry reservoir at the long time of the day. And it was just a pushback in a lot of states from the environmentalists. Right. I know that the investor right now has got deep pockets, a lot of experience that would be dying to do some projects in California. But I can understand Fong's concern because he's with the corporation. They've got to be very careful about their risk management. It's a low return, lots of different rules that they have to combine by their private investor does not have to. So people are more risk-oriented. And I know people who are risk-oriented will be lapping up pump storage projects. Fantastic. Just to repeat what Jay's just said for everyone's sake, what he's saying is that the biggest challenge to this is in the past has been environmentalists sort of pushing back on this, even perhaps in the closed loop one, and maybe less so. Much less so because we haven't built any yet and they've basically put their names on a document which says they're going to give them much stronger support. And now I think if he can mitigate that risk, I think it could be, it's bankable in many ways for the investor. So I think that's a very, very important step. And one quick thing, as opposed to PG&E, these are going to be projects not typically done, often not done by utilities, done by private developers. That's what I was going to actually say. In the old days, utilities such as PG&E, we would own the generation and we would take on the capital cost risk and overrun risk, but they were generally pretty well-defined technology, except for nuclear and pump storage. Those are the two that were scary. But in the last 15 or 20 years, the way I've managed the business, I've long realized that we should leave innovation to others. Our role at PG&E is to spec out the business needs, write really good contracts, and let the capital market and innovation decide. And what I do, the way I structure these is you tell me what you can do, write it down in a contract, and you better perform on that contract. If you don't, I'm going to terminate you and take your collateral and good-faith money and it's to keep people on it, right? So same thing with pump storage. If you want to give me a great project, write it into the contract where you're going to deliver it to me for, because I think virtually all of you are our customers. I'm trying to spend your money prudently, rather than you won't believe how many great ideas I've heard in my career. And I've even signed a contract with a space solar company, which I still ask myself, why did I do that? But they promised me a very good price. And from my perspective, I'm giving them a chance to advance their technology, to advance their dream, because that's what this country is and that's where we all believe in. And if they can't do it, I don't lose any of your money or my money, because I don't pay them anything. So, you know, if Andrew Mateo come to me with great contracts, I said, fine, let's give it a try. That's how I look at it. A pump storage project is not one I would put PG&E's balance sheet behind and I do not want to lose a few billion dollars. And you don't have to, because he's saying, go bring me a power purchase agreement. If I like the power purchase agreement, you, Mr. Developer, and you, Mrs. Investor, you go do the project, right? That's right. And the thing I failed to mention is that in a power purchase agreement, I tell my employees, we make no money, because I charge you exactly what I pay for. So, our goal is to never lose any money because if we don't make any money in this deal, we'll have to be in a really bad situation and we'll lose shareholders' money in this type of proposition. So, the contracts are written very tight. But you would have supply if there was a builder of pump storage, you'd have more security supply of electricity during a high demand period. I think the way, it's true, but the way you should look at this is I wouldn't have a world of transition. PG&E started with 10% renewables, we're about 40% today. This is an average for the year. So, what that means is just like Clyde said, the highest percentages are actually during the shoulder season, like the spring. That's where we have very low demand and we still have lots of renewables. Those are the hardest to manage. But over time, we're going to get to 50, 60, 70, whatever percent. So, what we're trying to do is back down the traditional resources you saw in the first slide in a smart and manageable way while letting the new technologies advance slowly. So, we're facing out one set of portfolio to replace it with a different. That's what we're trying to do. Let me ask Matteo and Andrew. We heard about the risks from the, and some of the challenges and opportunities as well from the Hydro. Tell us about the risks that you feel are there in the technologies that you're developing and when will PG&E sign the contract with you guys? Or have they already signed? You don't have to talk. I'll go. So, we're commercializing a new battery chemistry and so the risks there are myriad. There are many risks. All the risks, in fact. And so what we're doing is, however, is not the way that we frame it up inside the company. We are not developing a battery chemistry. That is not the job. The job is to develop and commercialize a bankable asset. And with that mindset, you sort of apply the right standard for what it is that you are doing. And for those who aren't familiar with sort of that term of art in the industry, bankable asset is something bankable. It means, can you get low cost financing to back your project? That is the threshold for bankability. And the way to establish that something is bankability is to have data that backs it all up. And ideally, and for many years, by the way, the only way to be bankable was to have an asset that operated in the field for 20 years. That's the only way. By definition, that was what something bankable was. Now, of course, the imperative to bring these new technologies to the market is much larger. And so there are ways to establish new technologies as bankable. This is exactly the path that Lithy Mayan just went down over the last 10 years, let's say. And the way that we did that, since I was a part of it, is you have data which is extremely hygienic, and you are testing at the system level for the wide range of operating and environmental conditions that you expect to see in the field. And you're doing, of course, accelerated life testing and everything that goes with it. So you can present all the data to the independent engineering audit firms that do this, and they will give a stamp of approval. Yes, this is effectively a bankable type of technology. So that's the pathway that we are in the middle of. We've been doing this for the last four years. And from day one, when we started the company, that was always the outcome that we had in mind. And so in the lab, what it means is that we are, as I said, very hygienic. Every test that we do is thoroughly documented and retrievable in terms of its data. Was the left-hand used versus the right-hand used, the classic construction there. And that is how we have designed our company as a whole. So the risks are that we need to go from, basically, a technology readiness level one or two, maybe, all the way to seven. For those of you who don't know what that scale is, go familiarize yourself. And TRL7 is something that's been commercialized. It's out there. It's proven. And that is sort of a bankable type of technology. And we are going through all of those steps and being able to show all of the work that goes into it for anybody who would ask. FONG would sign a contract on something more speculative. But doing a one-off project is not the goal, right? Failing on delivering a contract to FONG is not the goal. The goal is to build an asset that gets deployed at large scale and has an impact in the gigatons of carbon production. That's the goal. So that's how we think about it. And the risks you have to break up through all those stages and in the end getting to a bankable asset. Andrew, and then I'm going to come back to Clyde after that. And then we'll open it up for questions. Perfect, perfect. And our situation is 90% overlap with what Moteo just said. New technology, need to show data. One thing that we really worked on early on is to look for industrial analogs. Obviously, there are not going to be antorabatteries that have lasted 20 years. But one of the things we really like yet, because we'd have to wait 20 years for that, but one of the things that really drew us to this technology is that there are already what are called graphitization furnaces that are out there in industry. That's how they make graphite, for example, for the steel industry for electric arc furnaces. So you can go all over the world and you can look at units that are already storing huge amounts of energy in carbon at the similar temperatures we're talking about with the similar materials. And so even though we won't be able to say, hey, here's one of our systems that's lasted 20 years and here's all the data from that, we're hoping to be able to shortcut a little bit of that process by being able to say, here are other systems that look almost identical in all of these ways that have lasted that. Now that isn't a replacement for doing all of our own testing to do as well, but that's very important from our perspective. The other thing I'd say is that, I'm not sure when we might sign a contract together, but our go-to-market strategy is a little bit different. We are going after industrial customers first, which has pros and cons. So on the one hand, there's a lot of flexibility. These are smaller players that can move quickly on a one, two, five, whatever megawatt system. So in that way, it can be easier to get to those early systems. On the other hand, to the bankability point, there's technology bankability and then for any project, there's off-take risk. And there's nothing better than the off-take risk that FON can provide for a project to say, hey, I'm going to be buying the electricity out of the system for a long time. So we do trade off something. In order to have that flexibility and speed to market using smaller industrial customers, behind-the-meter installations, we do have the flip side of it. We need to be very careful selecting commercial entities that aren't going to pile on so much commercial risk, so much off-take risk for these contracts that we heard our bankability more in that direction. So very, very similar situation with a couple of small differences. Let me come back to play. You used to be a startup that way. Right. Go ahead. So, you know, whatever technology it is, eventually, the ISO has got to operate. Although we operate, the first thing I look at is reliability. So costs, I do not factor in costs. I know how well I could meet the new control performance standards in real time. We got some very, very strict control performance standards. As a matter of fact, the U.S. has its strictest control standards in the world, right? So when we integrate renewables, are you here? Policymakers say, well, they're doing this in Europe. They're doing that in Latin America. When you come to the U.S., we, you know, the U.S. is the orderly place that I know it has. You have an obligation to support in the connection frequency like every minute, right? That's not a simple thing to do. Within a window, 0.035 Hertz. Well, pretty close. 0.036. Come out of room. So, yeah. So it's a challenge, right, to do that. So whatever it is, you know, we develop big pump storage. In the old days, you know, as Fong said, when they built the pump storage plant, it was still offset the excess energy from the two nuclear plants, the Avalon, you know, at night. Today, we tend to use that pump storage plant during the day. You know, especially on weekends during the day, we pump until 2 o'clock, 3 o'clock, which is something that we never did in the past. But in the pumping mode or in the charging mode, if we develop, you know, a lot of storage, we need three things, right? And a few years back, Nutt developed this North American Electric Relativity Council. They formed this group, and we had about 20 people, and they were looking at, well, what would it take to integrate higher and higher levels of renewables? So the guys back east, they said, voltage control. So I saw New England, they took out voltage control. And OCCOP said, Texas, well, frequency control is huge. So we need frequency control. And then I was there from the west, and I go, we need flexibility. Everybody looked and said, what's flexibility? Because we have our own song. One guy looked at me, and he said, California, you guys created this problem. Figure it out yourself. So those are the three things we need to control the grid. Voltage control, frequency control, and flexibility. Well, to make a long story short, about three years after the same guy, from the east coast, he called me and said, we got 1,300 megawatts of solar. Now I see the flexible capacity problems. Can I come out there and see all the studies you guys did? So it starts off in the west coast, propagates to the east. But whatever it is, or whatever technology we come up with, remember the system now is so variable. We need pump mode, generating mode, voltage control, frequency control, and flexible capacity. Typically with the pump storage, when you slap those pumps on, it's 300 megawatts. You drop it off at zero. Verbal speed pump, you can get controllability in the pump mode and in the jet mode, which is something, systems with a lot of renewables is what we need today. As I mentioned, we got 12,000 megawatts of rooftop PV. We have no controllability, no visibility. Yet, rooftop PV impacts system frequency like grid connected, solar wind or traditional synchronous machine. So now you have all of this stuff sitting on your rooftop. It offsets your high-cost energy, but we got to control that. You got to predict that. You have folks that can predict it, but they can predict it on an hourly average basis. I got to control every four seconds. So it's nobody that I know of today or no forecasting company could forecast rooftop PV on a minute-by-minute basis. We can say, like, next hour you might get 5,000 megawatts. The other big challenge that we need to think of, too, is in the old days, we had one variable, which was load, and load was temperature dependent. And you could predict what the temperature is, you know what that load is. When PG&E, when PG&E miss the temperature by one degree, when it was high like around 100 degrees or more, we missed our load by 1,000 megawatts. So, you know, that's how sensitive it was. Today you've got four big variables. You've got a load, and in California, a load is no longer predictable, because you've got a price-responsive load, energy efficiency, demand response. You name it, we got it. And then you've got other things, like wind, solar, rooftop PV. Predicting that for the next day is a challenge. Fantastic. I mean, this is amazing. I mean, you heard from Clyde that for two seconds, you were 99.9% renewable, you probably are holding your breath for two seconds, which is not too bad. Let's open it up for a question. This is a fascinating topic. Over there. And if you could introduce yourself with your affiliation, that'll be great. Hi. This is Eugene from Quinnu Energy. A question for Fong, actually. How has safety, particularly fire safety, affected your decision-making in citing long-duration energy storage projects? Well, safety in general is incredibly important to us. And for obvious reasons, we haven't had many long-duration projects. I should probably say that, but we've had a lot of long-duration projects in terms of our own health storage. And, however, we do, in all the PPAs I mentioned earlier, we actually asked the seller to give us a safety plan for construction and during operations, and we actually verified that. Because even though legally our responsibility in all safety events on the other side is theirs, we would like to do our due diligence on that. But it is very, very difficult, and if my colleagues will probably tell you, I was a chemical engineer as an undergrad, and I'm fascinated with chemical batteries, thermal energy, but these are actually pretty dangerous things to work with, you know. We'll have to start with that. Yeah, I was just asking because recently, you know, the Moss Landing, for example, and if we do more lithium-ion, we'll probably see more of that. Moss Landing, this gentleman's referring to, is a 300-megawatt lithium-ion battery. The time that it came online was the world's largest battery farm, and it's a PPA under contract PG&E. It started delivering last June. We don't really know exactly what happened there, and whenever something goes wrong, there's going to be a legal dispute as to who's responsible. These gentlemen are going to find out whether it was the equipment maker, whether it was the EPC, or whether this one could be the fire suppressant system. We'll find out exactly what went wrong after the dispute is resolved. They're going to all see each other, right? That's the way this thing works. And in the meantime, when they don't run, I don't pay them. That's what we're performing. All right, next question, and there's a line forming out there I can see with each way and others questions. If you want to ask a question, maybe you want to line up behind E. Go ahead. Hi, Gao Liu from Lawrence Berkeley Lab. Thank you. That's a really wonderful panel. We learned a lot of things. Recently, in Berkeley Lab, we look into climate resilience. Right? So we also look into a lot of these climate effect lately, especially the drought. If you look at the West Coast, we know that along the Colorado River, Lake Powell is dried up, Lake Mead is almost there. So now we talk about this storage using a pump hydro that you mentioned. I think we may actually have really good opportunity there but on the one hand, we run out of water. So we want to see where we can find water, right? And if you look at the climate pattern, right, the East Coast has a lot of rains, right? And the West Coast is going to be dry, and that's the prediction. And can we actually bring the water from the East Coast and pump it all the way using the clean energy and actually build up all the energy all the way to the Colorado River. And then we can using the Lake Powell and Lake Mead as the storage place. And then we can have really consistent baseline, you know, energy supplies in the next, I mean, that's obviously a huge party but that's really providing the way that we can actually have a huge impact, right, for these kind of parties. That's an interesting... Interesting systems idea instead of transmission lines to build water pipelines and then have hydro. What do you think? I think one of the projects you mentioned, I think I know what it is. They're not using fresh water. They're actually using underground brackish water. That's how you get around the water issue. No one's going to let you touch pristine fresh water, not in California. E? Maybe using ocean. I appreciate the panel. Very exciting discussion. I want to come back to flexibility. You know, second to second minute to minute frequency voltage, all of this. Four hours and then we talk about 10 hours then the whole day, right, just keep going. So what that makes sense to plan out in the whole system level for different timescale. If I look at me as an individual homeowner, I say, do I do how many hours? Would that make sense? I don't want to say install a technology say these can do six hours and then I could do another one can do what, two days, right? It would not make sense. So finding a system that can be so flexible to cross many timescales. I understand there's no silver bullet material completely. There's no, these one technology can do all. But crossing different timescale field to make it look attractive, that's one. And also coming back to the long-term scale, I want to pick the material brain a little bit. You know, my lab, I really want to work on long-duration system storage. But I think about if every year I use one or two cycles how do I make money out of that? Very hard to sell, right? It would be good to think about business side. I think two questions right here, not just one. Any responses? Sure, I'll jump in on the use case. So batteries exist as part of the system and you have to be able to model them in that system and look at how they are dispatched economically in conjunction with everything else that's going on. When we do the modeling in our system, before we had hardware we had software that we're doing exactly these, what we call fast expansion modeling or production cost models. And what it shows is that when you have batteries which are available for 100 hours, let's say, you have between 10 and 15 cycle throughput equivalents per year but you're only doing a full, deep discharge maybe two or three times a year. Because exactly that case you only have those events two or three times a year whether that's a polar vortex in the Upper Midwest or that's a week of rain here in California. So what it's doing the rest of the year is all that you would expect from a function in the afternoon and time shifting during the day and it works with the other stuff. And to be clear, it is this type of multi-disk storage that we're developing is not to the exclusion of let's say lithium ion or other shorter duration, higher efficiency, higher cost systems. Those work together. In fact, that's how you co-optimize the system and that's what we did with natural gas, for example. We have combined cycle plants and we have simple cycle plants and we have about the same amount of gigawatts of each because they're doing very different things so the same will apply to storage and that's what we see born out in all the modeling that we do. I think you hit the that's the key point of flexibility and as I was saying earlier when I started my career the only flexibility I had to worry about was that fossil plant how it can follow the load in the evening when each of my customers needed. The flexibility I'm looking for is actually across all technologies between supply and demand. So I'm looking for flexibility for renewable resources. You feather the wind blades or you turn the inverters off on the solar or by batteries or by pump storage or I go to the customer side and say can you not use our electricity during certain hours how much do I have to pay you for you to not use electricity. So we're constantly looking across the whole spectrum and looking for the economic sweet spot and it's always changing. It's actually a pretty fascinating exercise of making the whole market bid against the flexibility you're looking for. Can I just quickly say I think you will chew other people in storage X ought to be doing this in fact this kind of work very different time scales very different technologies cuts across engineering law business policy this whole thing we all are kind of in our little boxes I'm a battery person and I might look at a couple of different battery technologies I want to do thermal, I want to do pump storage we need it all and I think what Stanford Caduce is doing great work but if it could do it in a more integrated way across all these different needs with the great capabilities across the whole university I think that would be a huge contribution. Any other questions comments Will go ahead. I'm from Chevron great discussion this afternoon we talked a lot about distributed power generation in the future what are your thoughts around distributed storage versus centralized storage in terms of providing efficient solutions for storage. That's great. Home, substation or entirely big, what is the kind of distribution any thoughts? I'll start with that one question the engineering looks at the economy of scale and efficiency first so typically it's very difficult for distributed resources to overcome large scale I'll start with that however our grid is actually changing we have pockets of growth we also have pockets of wildfire risk that actually makes sense in certain areas to deploy distributed generation and storage as well as distributed customer demand response so it does make sense. I'm from a high level From a high level when you look at distribution versus transmission you've got to keep in mind it's like a drive in a car you can only have one drive at a time to see at a car so if the distribution folks they go up 10 megawatts we've got to note out on the grid side sometimes if the grid is going to control and say we need 10 megawatts out they can only go 5 we still need to work out that control system back and forth between transmission and distribution what do you control on the distribution what the grid can control on the grid side you cannot fight each other one minute left, last question goes to Wilchu and we'll finish it up let me add my thanks as well great discussion my question is calling look at the contrast between manufacturing and scale for something like batteries and solar cells you make billions and hundreds of millions and that's how we have achieved significant scale and cost but now I think looking at the new technologies coming online or old technology then it's one much larger installation how does that play into the risk and the opportunities for scaling how will we realize the cost learning curve and how would it be different compared to lithium-ion batteries and solar cells quick answers I'll go super fast so we're still making cells which turn into modules which goes into the packs which get deployed into the field in fact we're ramping up our manufacturing facility now they will be producing initially thousands of units and we're going to make many of those going to hundreds of thousands of millions so exact same concept because that's been proven many, many times I would just say we need both systems we've got to do both and we've got to drive cost down and scale up and increase the speed at which we can get this stuff deployed and again this whole panel I think is about all that we need to be doing across all these many broad ranges I'll just say we're very similar our power generation is actually everything looks the same as solar we have cells that are series connected into modules put together in a system in fact we even use the same electrical balance of plants so solar inverters, solar transformers it all looks exactly the same as utility scale solar plant exactly for that reason we want to be able to scale cells module system quick answers from the clients I've got one last one which is if you could sell me a dream a dream that Mateo has done to me over the years which is if I can get these units done the first hundred units would be this price the next hundred would be this price so on down if you can sell me a dream explain to me how efficiency and cost reduction can be done and we will buy into the dream with you but you've got to have a legitimate dream Clyde we start with you we're going to end with you whatever technology it is that you develop please make sure it's something that I can control you wanted to work on that high note especially Andrew and I have to give a speech I have a dream let's give a big round of applause for the subject