 Good morning from Stanford University. My name is Will Chu. I'd like to welcome everyone to today's Storage X seminar. It's been a month and a half since our last seminar. I hope those of you in the Northern Hemisphere is enjoying your summer. It is really a great pleasure for me and Professor Itui, the director of the Precourt Institute, to host today's seminar. We have a very exciting topic and two excellent speakers to talk about thermal storage. This is a topic that we have covered quite a few times in past seminars for, as a matter of fact. We had the pleasure of hearing from two of our Stanford colleagues and Nobel laureates, Steve Chu and Bob Loughlin, who spoke about thermal energy storage and conversion at some detail. And this is a very important topic for not only conventional but also for advanced thermal storage. And today's talk is really going to touch on the second aspect of novel storage mechanisms involving heat. And this is not only important for energy storage, but also important for industrial decarbonization as well. And to get us started on this topic, let me introduce our two speakers briefly. We have Yorick Parchash, who is currently the co-founder and CTO of Redox Blocks, a company that is developing high-temperature thermal storage technology. And we also have Professor Ashingen Henry from MIT who has been working on high-temperature energy storage for about a decade and a half. So Yorick will be our first speaker. So let me ask Yorick to come to the stage. Yorick, I will keep the video off since my internet connection is not that great. OK, no problem. Well, Yorick, do you want to just flash your image so we know what you look like? And then you can turn off. OK, there's Yorick, so now you know what he looks like. Yorick has been working in the area of high-temperature thermal chemistry and storage for more than 25 years, starting with his PhD work at ETH Zurich. And he's currently the professor of mechanical engineering at Michigan State University. He's on leave to start Redox Blocks, and which started just this past year to commercialize thermal energy storage systems. And Yorick will be introducing to us his company. I believe this is one of the first times in which Redox Blocks is talking about the technology. We're really pleased to have your work share with us all the details. And Yorick, we're really excited and looking forward to hearing more. Thank you very much, Will. So I will talk about the Redox Blocks technology, which is a thermal chemical storage technology. So we're adding a chemical component to this. It operates at very high temperatures. And I'll talk both on our applications for industrial heat as well as for grid-scale energy storage. Next slide, please. So my talk, I'll start with a brief introduction on thermal chemical storage via Redox thermal chemical systems and then introduce our technology concepts both for high-temperature industrial heat and then for really large-scale grid integration. I will then spend some time on the secret sauce, the material at the heart of our technology, which is a magnesium manganese oxide. And I will talk a little bit about its cyclical thermal chemical stability, the chemical thermodynamics. This is work that we've done in our lab at Michigan State with one of my PhD students, Alexander Boe. I think he's actually in attendance today. I will probably, for time's sake, skip over the heat transfer aspects and then give an overview on what we have been doing on prototype being we have built and tested and now this assembled 100 watt hour system. We're currently running a 10 kilowatt hour system and we're currently constructing 100 kilowatt hour system and developing some bigger systems, which I'll briefly touch upon in the next slide, please. So in very simple terms, Redox thermal chemical storage is something like reversible combustion. So what we are doing with our technology is we're taking something that may sound from a thermodynamic perspective not very beneficial. At first, we take cheap renewable electricity and convert it to high temperature heat to drive a thermal chemical reduction. And we're doing this with the material. It's a mixed oxide consistent of magnesium manganese oxide. And if you heat it above temperatures around 13, 1350 degrees C, it will start reducing just thermally, releasing oxygen. We remove, we pull, basically pull out the oxygen out of the system and then keep the material at that high temperatures in an insulated vessel. And it just sits there until the energy is needed. And we get the energy back by passing air, regular air, over a packed bed of the material. The oxygen in the air reacts with the system. The reduction reaction that we just drove is reversed. We have an oxidation. And the air is in direct contact with the packed bed. Gets heated up to temperatures between 13 and 1500 degrees C. And then we can use that high temperature air that leaves the system either directly in a range of high temperature industrial applications or to drive a turbo generator. If you go to the next slide, please, I'll explain the industrial heat application first. This is sort of the simpler application. And the reason I took leave is to set up an office here in Europe, where currently, as you all are aware of, there is a severe natural gas shortage. And our technology is basically a drop-in replacement for a wide range of natural gas-fired industrial processes. So in this case, as you can see, we have an internally insulated vessel. It doesn't have to be a pressure vessel. It just has to be airtight. And during the reduction reaction, what happens is the rotary air pump at the bottom left, or sort of bottom middle, is insulated from the system with a valve. And the system is heated by passing an electrical current through the packed bed. Magnesium manganese oxide is a semi-conductor. So it becomes electrically somewhat conductive at temperatures above approximately 800 degrees C. And then oxygen evolves as the system heats up, and the oxygen is removed with a simple industrial blower. Now, once we want the heat back, air is pumped through the system, through the packed bed with the blower or air pump. And the oxygen reacts with the magnesium manganese oxide bed. Not all the oxygen is used up, depending on the state of charge, between 50% and around, towards the end, around 25% of the oxygen in the air is removed from the air. And so we have somewhat oxygen depleted air that leaves the system at high temperatures, typically between 1,300 and 1,500 degrees C. And that can be used in a range of industrial processes, ranging from simple things like high-temperature steam generation to a range of metal smeltering applications, and so on. Why would we want to do this? I have not gone into the importance of storage, but I assume that for the current audience, it is obvious that, of course, you could build a high-temperature air hinder and just run it. However, if we want to transition to a renewable energy system, we have somehow we have to buffer the production or generation of electricity relative to when it is used. And in industrial applications, we typically see that the load is relatively constant over either 18 or 24 hours, depending on whether a given installation runs in two or three shifts. But it is typically a very constant load, as opposed to renewable generation, which, especially for in sun-rich areas, is pretty much centered around noon plus or minus three hours. So we have about three hours of production. And with a system like this, we would charge it during those six hours. Typically, you would run at least two systems in parallel, charge them rapidly, which is especially simple with this system, because a lot of the heat transfer limitations that typically are associated with peck beds, particularly in sensible heat applications, we don't have that kind of limitations, because basically what we're doing is we're dual heating the whole system, so it's like one big resistor. And due to the fact that there's simply a volumetric source, there is literally no heat transfer limitation. Next slide, please. This is our somewhat long-range vision of, and that is how we started developing this technology. This was our first idea, where we would basically take a pressure vessel insulated internally, fill it with a peck bed of magnesium manganese oxide. Heat it just as in the system that we showed before. But now it goes in place of the combustor of a gas turbine system. And here I'm only showing a simple turbine system, but of course, this is applicable to combined cycles as well. So in an ideal case where we're running on a very efficient combined cycle, the storage efficiency is about 55%. Otherwise, the basic idea is very much the same as I showed before during times of high electricity generation from renewables. We charge the system. We remove the oxygen. That oxygen can be used as a secondary source of revenue. It's not what we're focusing on at the moment, but we're generating a significant amount of oxygen during that reduction step. And then during oxidation, the oxygen is taken out of the air and put back into the storage material. Next slide, please. I now want to hone in a little bit and give a little more detail on the magnesium manganese oxide redox system. Next slide, please. So this is a material that was originally developed at the University of Florida by my long-term collaborator, James Cloudsner and his group, James and I, have been working on redox systems together for, I think we met 15 years ago. So we've been collaborating on and off. And this was developed at the University of Florida about four or five years ago. The raw materials are very cheap and abundant, sort of at the large scale, about $600 per ton process cost. Magnesium is the seventh and manganese is the 12th most abundant element in the Earth's crust. The performance is outstanding. We have an energy density of around 2,400 megatron per cubic meter. This is combining sensible heat storage between 1,000 and 1,500 degrees C and the chemical aspect, so entropy of reaction. It's about 2 thirds chemical and 1 third sensible heat. This is about three times the energy density of molten salt just for reference in comparison. Temperature operation is between 1,000 and 1,500 degrees C. However, most of the reaction, approximately 80%, proceed in the range between 1,300 and 1,450 degrees C. So the actual temperature range that we're running most of the time is somewhat more narrow, which is beneficial, especially if you want to run on gas turbines. This is equivalent to current efficiencies between 77 and 83%. The material is very robust. We have shown over 100 cycles in actual prototypes, 2,500 hours and counting. We did not observe loss in capacity in the sense that the changes that we see are within the measurement errors of our devices. The materials are safe. They're non-toxic, non-corrosive, non-combustible, except for the fact that they oxidize at high temperatures. And the material is 100% recyclable. And then we have patents, both on the material as well as on the application. Next slide, please. So now let's dive into some of the chemical properties of the material. Next slide, please. So the material, what has been holding back redox cycling or redox thermochemical storage is cyclical thermochemical stability. Most of those materials are geared towards high temperature operations. And that is beneficial from an efficiency point of view, of course. However, they also tend to center. So what you typically see in a lot of metal oxide redox systems is that the first couple of cycles are wonderful, high energy density, very repeatable for five, six, seven cycles, where everything is still in the powder form, and then slowly, slowly the capacity decays. And after 50 cycles, not a lot of your initial ability to exchange oxygen is left. And the amazing properties of this mixed oxide is that it maintains its porosity and more specifically its specific surface area through cycling. So what you see in the bottom left is a CT microcomputer tomography image of a pressed pellet. And then on the right-hand side is the same pellet after 50 cycles. And you can actually see that a pore structure evolves. The envelope stays relatively the same. But the internal pore structure evolves. And you can see that a large surface area is maintained. And this is beneficial for oxygen exchange, which is key to the performance. What you see in the graph on the bottom right is thermochemical cycling data in a TGA, in a thermo-grammatic analyzer. And at the bottom, you see almost perfectly overlapping red and blue curves that show the mass change, the relative mass change. So it's approximately 5% mass change due to the oxygen uptake and release during oxidation and reduction. So we're cycling between 1,000 and 1,500 degrees C and releasing and taking up oxygen. The red curve is in 50 initial cycles. And then you see the blue curve. That is the same pellet after a 500-hour dwell at 1,500 degrees C. You can actually see that initially, performance is slightly better. And then it converges back to the converged state. Next slide, please. Next slide. So what is fascinating about the material from a more fundamental point of view is how this pore structure evolves. What you see here is SCM images of a pellet that spent a couple of thousand hours inside a prototype reactor. And you see that it forms a very open pore, reticulous porous ceramic foam structure. And this is what we think is behind the excellent cyclability. Next slide, please. And then a little closer up, a microcomputer tomography image of a cross-section of a fresh pellet. Next slide. This is then after 20 cycles. You see there is some sintering happening, some compaction. But then after 70 cycles. Next slide. You see this almost radial structure due to oxygen exchange. And you have a nice open pore structure. And then we let it dwell for a while. Next slide. And you can see that you see some sintering. However, if you restart the cycling, you'll go back to a structure that's much closer to the previous slide. Next slide, please. This is just to illustrate that the actual composition, the relative magnesium to manganese ratio, affect whether the material shrinks overall, like in its envelope, or stays the same in its envelope density. And this is something we're still working on to get the exact formulation for as stable envelope volumes as possible. Next slide, please. Next slide. What you see here is the chemical equilibrium composition. So the chemical equation is almost a cartoon of the actual reaction. And y signifies the excess oxygen as opposed to the monoxide in the oxidized state. And what you see in the graph at the bottom left is the temperature and oxygen particle pressure dependent value of y. And what you can see is that at high temperatures and low oxygen pressure pressures, you're almost at the monoxide. And then as you go to lower temperatures and higher partial pressures, you're adding more and more oxygen to the equilibrium. Next slide, please. We've built a couple of prototypes. Next slide, please. So the very first one, and of course, by prototype, I mean a system that is relatively close to a potential product or application in the sense of it runs on regular compressed air. It's inside a pressure vessel. So our 100 watt hour prototype that we built about, I think we started constructing that about two and a half years ago or so, that's the smaller vessel that you see in the background in the middle. So it's basically an internally insulated system that has about 200 milliliters of the material, or a little less than that, at its center. We cycled this system. This was the very first trial between 0.2 and 11 bar absolute in the temperature range between 20 and 1,500 degrees C. The actual cycling proceeded between 1,000 and 1,500 degrees C. Power is about the kilowatt. Most of that power goes actually to keeping the temperature. So this is not really a storage device. It's just a device to cycle the material in the peg bed. Next slide, please. So this was sort of the first proof of concept where we did five cycles reaching the 2,500 megajoule cubic meter as predicted in five cycles in a real life situation of using actual compressed air. No above gases, a real vacuum pump to suck out the oxygen during reduction. Next slide, please. Then we went to, and this is the vessel you saw in the foreground in the previous picture. This is our 10 kilowatt hour prototype. Very similar, actually the same test stand. Just a little more elaborate. Next slide, please. And here we've been cycling this for 1,800 hours. Approximately, we currently have a downtime because we've experienced some issues with the heating system. What's interesting is that initially we were a little worried because we saw this decay of energy stored. This turns out to be an artifact of drift in the measurement of the flow meters in the measurement of the airflow to the system. So after C65 sort of towards the right of the graph, we replaced one of the flow meters and we're pretty much back to the original energy density. Next slide, please. So currently we're building a 100 kilowatt hour system. It's under construction and we're planning to have that operational this fall, basically scaling up the system to a level where the losses relative to the power input become much smaller on the order of 10% to 20% over half a day or so of storage. So now this is going to be the first real storage device that we're going to build. And the diameter is about a meter. Height is almost 2 meters. Next slide, please. And we have an ongoing project with collaborating with Siemens Energy and Martin Energy, funded by RPE, where we're getting this technology towards the first product running on a micro gas turbine, 2 megawatt hour storage. And that project is scheduled to run for another three years. And we're also in product development mode on the pure heat application, which obviously is a little simpler to do. And that's all I have to say today. I would like to thank my co-founders, my collaborators. And I'm ready for questions. I hope I'm on time. Jorak, thank you very much for the introduction to Redox Box and the underlying material science and heat transfer and technology. Really, really appreciate the deep dive. So we have a number of questions from our audience. Let me start with a higher level question on the technology itself. So Jorak, you showed a slide that talks about the chemical storage via chemistry versus storage via heat. So the question just want to clarify. So when you mentioned heat, that is referring to the heat capacity of the material? Yes, that's sensible heat. That's sensible heat. So that's just simply because you're going from 1,000 degrees to 1,500 degrees C. There's a significant amount of energy stored as sensible heat. And then the chemistry. And then there's the added effect of the chemistry, which is between 60, around 60, 60 to 70% of the overall storage is actually in the chemistry, in the oxidation and reduction. And I think for many of our audience, they can appreciate that the chemistry part of the heat storage can be quenched. So if you cool it down, the chemistry is retained. So can you talk a little bit about the various operational modes? If you want to get the most out of it in terms of heat, you have to keep it hot. But if you want to have a long duration storage, just cool it down. Let it keep cold. Just talk a little bit about that. So I don't have this on this slide. And this is a little, how shall I say it? This is a little higher, a little further up in the pipeline. We have also at Michigan State in a separate project that was funded through CITO, the OE Solar Energy. We developed a technology where we were actually recouping all the sensible heat. So this is basically a falling bed or falling particle tube-like situation with a heated zone and an inert gas in counterflow. And so basically, all you have to do is you have to match the m.cp of the material coming down and the counterflow inert. It actually can contain some oxygen, just not too much. You just have to look at the thermodynamics, what's sort of the optimal point here. But you can actually drop the whole thing down all the way to ambient. So we have this system where it's basically a tube with a heated zone in the center, air entering, or sorry, an inert typically nitrogen entering at the bottom at ambient, the material entering at the top at ambient. And then they exchange heat in a counterflow, which is really nice with actually published. It's just a little bit beyond what I wanted to show here. But you can actually recoup all the heat and then do long-duration storage on this. Thank you, Jorga. To me, this is very exciting because then there will be no self-discharge whatsoever in the fully That's right. And then if you can also recover all the heat during the cooling and prevent the oxidation of the material, then it would be the perfect battery for long-duration storage. So I think this is really exciting. So would you say that the technical challenge there is to prevent the re-oxidation of the material on cooling to make sure it doesn't lose any charge when you crunch it down? Actually, the main challenge that we had to overcome in that project was the flowability of the material. At 1500 degrees C, everything becomes a little sticky. There is some sintering tendency, not a very strong one, but we experienced logging from time to time and really getting the pellet size, the particle size right. We're working with a different pellet geometry on the storage on our re-dox block technology where we're mostly working with cylindrical pellets because it gives us somewhat better packing as opposed to what we call it, the SOFU technology. That is spherical particles, they just have better flowability. You have to get the flow rates right. I would say flowability has been a challenge that we have pretty much overcome now. Once it's at ambient, of course, theoretically, it's not thermodynamically stable, right? It would want to oxidize, but you can actually let it sit in the charge state for half a year and you will not have measurable oxidations. The kinetics are far too slow, so it will oxidize but in a million years. Well, that sounds terrific. This is the part that really excites me and I think one aspect of this is one aspect of this low self-discharge rate, I'm trying to understand how this could impact the system economics so maybe we can move on to the second question, which is, can you tell us a little bit about the technical economics here? How advantages contribute to the economics of this process? So the storage system cost, and I'm talking about the re-dox technology here, is depending on whether you're going for the grid scale application where you have a turbine and you need a pressure vessel, then at very large scales, the estimated costs are around $10 per kilowatt hour. Once you go to somewhat smaller systems, we're talking more like $20 per kilowatt hour for a pure heat application, because the pressure vessel is about 40% of the cost of the system and then there's sealing and other stuff that you don't have to do in a pure heat application. We're down to $4 or $5 a kilowatt hour, which I think is very competitive and the main reason is just the low cost of the underlying storage materials and the storage materials make more than 80% of the overall system mass. And you're just to confirm, so this is the total system cost? So this is the storage, this is the cost for the storage system, yes, without the turbine, of course. And you would see that if you had an integrated system, just to give you a rough idea for 10 hours of storage, the storage would be depending on the size, but between 10 and 15% of the overall system cost and the bulk of or the main contribution to capital cost, which is be the turbine. And how, the fact that during charging, at least that you have to go to this high temperature and requiring more specialized materials for your enclosure and reactor, how much does that add to the cost? So this is actually, it's not as bad as it sounds because basically it's fire break, right? So it's like, the construction is in many ways similar to a blast furnace, right? Which is, there's a metal casing, which is just a mild carbon steel. And then there's ceramic insulation and what is essentially fire break? You can actually use the magnesium manganese oxide itself as the innermost protective layer because it's a high temperature material. It won't melt up to 2000 degrees C. So it's, and that is cheap in itself, right? Great. Let's get two more questions. So these are a little bit more specific. So we have a question on the heat up of the material from room temperature on a cold start. Yes. So schematic illustrates conduction through the material. Electrical connection through the material as the way of transfer heat from the city. So how does a cold start look like? So that is an excellent question. Because you have to somehow bootstrap it. And the way we're currently doing this is by basically running hot air over the system. In a large scale system, you would inject a combustible gas, be it natural gas or be it hydrogen and combust that flamelessly in the pecked bed. We've actually done experiments showing that that's compatible and there's no problems with doing that. So you basically have a large porous burner and you only have to do that once, right? You only do that when you start up the system. And then maintaining the system at temperature is something you would do with a high temperature thermal system. If you can, anyways, because high temperature ceramics, fire breaks don't like to be cycled. It's something that in a blast furnace or similar application, it's something you don't wanna do anyways. And you basically have to, and also to use your capital well, you want to cycle the system relatively often, let's say on a daily basis, right? So you can trickle charge the system at approximately 1% of the capacity per day, which is not terrible. So for a, let's say, one megawatt, there's nominally one, it's a couple kilowatt for a several megawatt hour system. So it's not, but it's still something we have to be aware. It does not affect the economics of the system very much, but it's a very good question. Thank you, Yorick. I think E will have the last question. E, go ahead, please. Yeah, thanks, Will. Hi, Yorick, really interesting technology, very nice talk. What's the long trip efficiency here, considering for storage, let's say you have electricity coming in, right? You heat up the system, you oxidize that, and then oxygen release, because you heat it up, and then let's say you store this energy over a month, this whole thing cool down, and then you come back, you discharge. What's the long trip efficiency? So in the, just to clarify, in the application here, we would not let it sit over a month. We would do that in the system where we actually quench it down all down to ambient. You would do that, you would do something like seasonal storage. In this case, we're more talking about daily or weekly storage, right? So to overcome either your typical time scales associated with solar energy or with wind, right? With solar energy, it's basically a day. With wind, it's the sort of, the variance can be a little longer, it can be up to a week, but that's sort of what we're seeing here. The rounded efficiency is, if you charge and discharge once a day, is basically dominated by the efficiency of your thermal cycle. So if you have a simple gas turbine cycle, a simple braking cycle, between 30 and 40%, and then the storage efficiencies is between 95 and 98%, depending on the size of the system, right? The larger you make it, the better it is. 95% is if you do not recoup any of the heat that's in the oxygen that you suck out. Oxygen is about 5% of the mass. So if you lose all that heat, which in a large system you wouldn't do, you would be able to recoup some of it. And we've even built that in into our 100 kilowatt hour system. We have sort of like a, basically a ceramic sensible heat storage that uses some of the heat of the oxygen that sucked out. So it's 95 to 98% times the efficiency of your cycle, you know, 40 to 60%. Yeah, yeah, thank you. York, thanks again for this wonderful introduction. Personally, I find the technology very interesting because I was working on this process as also about 20 years ago for a different purpose. I'm glad that it has found a new life in terms of application. I think the most exciting part is just that there's so many use cases. You can keep it hot, keep it warm, you can keep it cold. And really the duration of storage is entirely flexible using the various operation modes. So very excited to follow the progress of UDOT's blocks in the coming months and years. York, thank you so much again. So now, thank you, York. And we'll come back to you for our panel discussion. So say it right there. Let me ask Ashley now to come to the stage. So for our second talk, again, very much building on this concept of thermal storage and thermal batteries. We're really delighted to have Ashid Henry from the mechanical engineering department at MIT. I've known Ashid for quite a while and I think the best way to introduce him is that he is a theoretician turned to crazy experimentalist. I think that's a pretty fair statement. Starting with just working with Adams and the computer to building these crazy high temperature devices and looking at very unusual implementation for thermodynamic cycles. And I think Ashid was exposed to the technology aspect when he was in the first cohort of ARPA-E Fellows many years ago, where he worked very closely with now our Dean of our new School of Sensibility, Aruma Jumdar. And I believe that really influences thinking and sort of migrate him slightly from pure theory to crazy experiments as well. And Ashid has been working on many applications of heat transfer and thermal storage and also just fundamentals of heat transport in materials. So Ashid, we're really, really delighted to hear about the latest and greatest coming out of your lap at MIT. Ashid, floor is yours. So thank you. Thank you. Thank you for the introduction and another piece of connection. Will and I got to know each other. We started on a project also on redox cycles. And so that's how we started together. All right, so I'm gonna talk about a technology, we call thermal batteries that we've been developing in my lab for, I'll call it about 10 years. Actually kind of the genesis was this project that I worked on with Will. And essentially what has happened over 10 years is there have been three very significant technological steps forward. And I'll review what they are. They have ultimately resulted in the recent foundation of a new startup company. I founded a company called Thermal Battery Corporation. We are hiring like mad. So if you're interested in the technology, if you're a clever, creative, high energy engineer committed to mitigating climate change, please do not hesitate to send an email to info at thermobattery.com. We are very interested in hiring the best and brightest. So I'm gonna review the concept itself, talk about some of the economics and experimental data, things that we've done today to walk you through these three big steps forward in terms of kind of breakthroughs in the technology itself. All right, so before I dig into that, I'll just show one slide of background and context. I think I presume the audience is familiar with the storage problem. However, I will mention three pieces of terminology that I'll use throughout this particular talk. I'm gonna talk about costs. So real quick, the slide on the right is basically just showing a composite image of composite data taken really from the three papers cited on the bottom left. Essentially what it says is that a couple of independent studies here have shown that one of the most important quantities in the energy storage problem is what we call the cost per unit energy. We labeled it CPE. What it is is the CAPEX, the capital expenditure associated with all the components in an energy storage system that scale with the amount of storage you have. So in our case with thermal batteries, and is the case for many different technologies, particularly ones that are not just regular electrochemical batteries, you can have it where the amount of energy you store is a separable decision you make, separable design parameter from the charging rate and the discharging rate. And so we separate the costs because you can scale these things independently. And so what the studies have identified is that the cost per unit energy, specifically the amount that it costs to store more energy is one of the most critical and important parameters. What's shown here is what's estimated as the amount of penetration we can see of renewables onto the grid as a function of this cost per unit energy for the storage technology that you deploy to store the energy. And there are of course a number of assumptions that go into this. Right now we're at about 25% or so globally, but in order to get to the high 90s and up to 100% we need essentially more than an order of magnitude decreasing costs compared to lithium ions. So I think one thing that is clear to many of us that work in this space is that lithium ion won't get us there, won't be enough. And that we need some essentially radically new or very different type of approach to storage in order to get the cost down. The second important parameter is what we call the cost per unit power. We label it CPP. This is the capital expenditure associated with all the components that scale with how fast you charge it or discharge it. And so power and energy are different. You could in theory have a thermal battery with lots of energy stored, hundreds of hours of storage and only discharges at 10 megawatts. You can also make one that charges at 100 megawatts and discharges at 10 megawatts. So each of these things is separable but at the end of the day we label all the capital expenditure associated with how fast we can charge and discharge the equipment that facilitates that we label that as the CPP. And then you have the round trip efficiency or RTE which is total electricity that comes back out of the battery after you charge it. So you charge it up. We spend some amount of kilowatt hours doing that. You discharge it a certain number of kilowatt hours come out. The output divided by input is what we call round trip efficiency. And then the other key parameter in the economics is the lifetime, how long it lives, how many times you can cycle it. We usually represent that in years. And in our case, we expect that thermal battery should be able to last 30 years or more. So these studies have essentially concluded a couple of things. They all come to similar conclusions which is that CPE is the most important. One of the most interesting things which is what really sparked us to think about moving in this direction is that the round trip efficiency has to be above roughly 35% or so in order to get money to make some profit from more arbitrage but it does not have to be super high. So batteries of course, pumped hydro all are up in the 80 to 90% range in terms of round trip efficiency but you can sacrifice efficiency if it buys you much lower costs. So that's the direction that much of the field is moving in. And specifically, we need to try to get to costs below $20 a kilowatt hour. So let me talk to you about how the technology itself works. It's going to sound. If you've taken a thermodynamics class, it's gonna sound absolutely idiotic. We take electricity, we convert that to heat and store it as heat only to take another thermodynamic penalty, converting it back to electricity again. But there is good reason to do so which is exactly what I mentioned on the prior slide. We can take, it is advantageous economically to take a hit on efficiency of let's say 50% if it buys you an order of magnitude reduction in cost. The order of magnitude is more important than the factor of two. And so the way we do this in our concept is we can take electricity from any source. We are specifically interested in focused on it being a renewable resource but is essentially technology is agnostic. And what we do with that electricity is we run resistive heaters. Now these resistive heaters are operated at extremely high temperature, essentially the same temperature you have inside of an incandescent light bulb which is about 2,500 degrees Celsius, much higher than most things that you could ever find in any technology used at an industrial scale. So extremely high temperature. We run those heating elements so hot so that they can heat something else super hot. And what we use to heat up is we actually use a liquid metal as a heat transfer fluid. Its purpose is simply to move heat from one location to another. It is not our actual storage medium. And so what we have is graphite tubing. Graphite is nice because it is able to be used up to 3,000 degrees C or more. We're not going that hot. We actually are peak temperature is 2,400 degrees Celsius but we use liquid tin as the liquid metal. And the reason we chose liquid tin is it's very safe, non-toxic, low cost compared to something else like gallium, low melting point, high boiling point of 2,600 C, melted to 232 C. And the most special thing about tin is that it does not chemically interact with carbon. So we don't have any corrosion in our system and we're able to pump liquid tin inside of a graphite infrastructure without experiencing corrosion. So that's the main reason for that choice of materials and that simplifies our life. So the challenge for us was then trying to make a system entirely out of graphite. Our entire system is essentially made out of one material. It's all carbon. It's different types of carbon but at the end of the day it's essentially one element. Another key point is that our entire system is housed inside of an inert environment. So there are warehouses that they build for fruit storage like apples. You can hold apples in an argon or nitrogen environment and keep them fresh for like six months or more. There are facilities, there are companies that make those kinds of storage rooms and facilities. We envision putting our entire system inside of one, building an entire inert warehouse that'll keep all of our components from oxidizing. And so this is a key piece of our technology is we keep oxygen out of the system. So our entire system operates in an inert environment. That gives us great flexibility, long life, great flexibility in the materials that we choose because we don't think about whether or not it's gonna oxidize. So what happens is you run these resistive heaters. The liquid metal comes in at about 1,900 degrees Celsius which is already glowing white hot. And we heated up with these resistive heaters all the way up to 2,400 degrees Celsius. And we are mechanically pumping it. We actually have pumps that operate at these temperatures and we can mechanically pump it over to our storage unit which is shown here on the left. The storage unit consists of a different type of carbon, very inexpensive carbon, which are just graphite blocks costing the range of 50 cents to a dollar kilogram. And we have the tin contained in dense walled graphite tubing that runs in between the blocks and as the tin is pumped through here, transfers the heat to the blocks, heating the blocks up to that peak temperature of 2,400 C. So when all the blocks reach 2,400 C, that's when the thermal battery is fully charged. Then later, when you now want electricity, you pump the liquid metal through to retrieve that heat and move it over to the power block. The power block in our case is a big departure from what's normally done. Typically people would use a turbine, turbines are the most efficient heat engines on earth. There's some good reasons why we don't go with the turbine that I'll review, but suffice it to say our major departure here is that we actually use photovoltaics to convert the energy, the heat back to electricity. And how we do it is by converting the light coming off of the piping. So we pushed the temperature so high so that the light source is super intense and it allows us to use very efficient, more expensive PV cells. So technically these are what are called thermal photovoltaic cells. Sometimes we also call them MPV cells because they're generally multi-junction cells. So we try to boost the efficiency as much as possible so we're almost wholly performance driven. These PV cells are not normal cells. These cells have mirrors on the back. And so what we do is we don't try to convert the majority of the light, we only try to convert the highest frequency light, the visible light that we can convert very efficiently. The rest of the light goes through the cell because the cell is transparent to it. It reflects off of the mirror and goes back to the hot infrastructure and it gets preserved inside the system. In a sense, it's like containing the sun inside of a box. And so it's like we make our own terrestrial sun and then we contain it inside of a box of insulation. And so any energy that we don't convert, we then send right back to it to help keep it hot. These cells are not running at high temperature. These cells are water cooled, so they stay cold near room temperature. And that waste heat is then dissipated in a dry cooling unit outside. Now, one of the most important features, one thing that'll come up a little bit later as I talk about, you'll see here, it looks like metal fins that are in here. We don't really need fins as much as we do need this metal layer here. So this is the one part of this hot side of the system that's not made from carbon. This is actually tungsten here. And so we use tungsten foil to act as the surface that's radiating the light to the PV. The reason for that is that carbon's vapor pressure at these temperatures is quite significant. And if you put something that hot next to, just a couple inches away from something at room temperature, what you've basically made as an evaporator and you'll gradually move carbon from the hot side and deposit it on the PV. The PV will get coated in carbon and you won't be able to get the light in. And so instead we use tungsten foil. Tungsten's got about four order of magnitude lower vapor pressure than carbon does. And that's the reason it's used in incandescent light bulbs. So this helps us. It's a very important aspect of the system. And the other piece is that these PV cells are mounted on an actuator. So this water cooled heat sink can be actuated in and out of the light, similar to a control rod in a nuclear power plant. So we can ramp from fully off to fully on very quickly, as quickly as we can actuate the PV cells into the light. This in particular has very, very large advantages and is very much of interest for utilities to be able to ramp that quickly, to be able to do load following and to be able to provide emergency services as needed. The cost of our technology is estimated to be a little bit below $10 a kilowatt hour. I'll show you that on the next slide. To understand why the cost is that low, best to dig into why the source of the low cost is ultimately that the storage medium itself is very low cost. So this is like the lower bound. Of course, there's a bunch of other costs to add to this to build a system around it. But the starting point helps you see why the cost is so low in the first place. So let's talk about this cost per unit energy. How you can compute that is you just simple estimate, take this cost of the graphite, the cheap graphite, 50 cents a kilogram divided by the amount of energy stored in every kilogram, which is the heat capacity, 2,000 Joules per kilogram per Kelvin, multiplied by the temperature swing. As I mentioned before, we're swinging the temperature from 1,900 C up to 2,400 C. So that's a 500 degree C window, so 500 Kelvin here. And our estimated round trip efficiency is 50%. And so we take another hit. So what we're storing is what we want is electricity. So we have to pay another penalty that thermodynamic penalty shows up here. And this comes out to 10 to the minus $6 per Joule, which is essentially $3.6 per kilowatt hour. And you can see by comparison to lithium ion, much, much lower. This gets us in that range of one to two order of magnitudes lower. And this is what the starting point is. And it's the reason we think about going and committing this like thermodynamic crime of going from electricity back to heat, only to go to electricity again. Now, when you start adding up the rest of the costs, you add up this inert containment, the pumps, the piping, the cooling, the insulation, construction, all these different things, of course, add to the cost. And they are, in fact, more expensive than the storage medium itself. But even once you add all that, it still comes out a little bit less than $10 a kilowatt hour, which is what makes it very, very exciting. It's one of the only technologies we're aware of that gets to this low cost range. In addition to, I guess, is what I've just heard with the Redux box as well. And then the other portion of the cost is the cost per unit power. The cost of a turbine is close to like a dollar a watt. And so we see some major cost savings by using the PV. Because the light source is so intense, because we've gone so high in temperature, the PV cost is close to like 10 cents a watt. But again, you've got to add all kinds of other things, you need an inverter, you need that dry cooling unit. Those are two of the big costs that actually affect our total cost. But it still comes out to less than about 35 cents a watt. So this cost of about a third that of a turbine is what really has driven us to using PV as opposed to a turbine. This system does have the option to one day use a turbine instead. We could turn the temperature down. The graphite heat capacity doesn't change that much. We could conceivably decrease the temperature to the range that would be more suitable for a turbine and use a turbine. But it seems as though PV is actually cheaper. And so we're focused on that. It's also something we think we can commercialize faster. One of the big key things to recognize with this is the systemal storage, the bigger the better. What drives our cost to want to go very large in scale is the cost of that insulation. So one of the only option you have if you want to insulate something with conductive insulation above, let's say about 1700 C, all the oxides will center and fall apart. The only option is really carbon fiber based insulation. That insulation is quite expensive. It is not too expensive for our system. But as you can see, as we go bigger, the skin, the surface area is where you lose the heat. That starts to shrink in comparison to the total volume of energy that you're storing inside as you go bigger. And so this is one of the most important costs for us is the cost of the insulation. And that's one of the biggest drivers for us to go to very large scale. Now the second thing you might think it sounds absolutely insane is I mentioned that we're pumping liquid metal at 2000 degrees Celsius. As of about six or seven years ago, that was not considered possible. This was the first, I would say technological breakthrough that really set off a chain of events when we were able to do this. This was actually in the project that I was involved with that I led that Will was a part of. We initially were thinking about doing a thermal chemical reactor that involved liquid metal pumping. We then moved to concentrated solar power, doing liquid metals and then eventually to thermal batteries. But at the end of the day, the core shift in thinking. And this is, as Will mentioned, started as a theorist. And from a theoretical perspective, heat used at extremely high temperatures can be similarly valuable to electricity itself, has the advantage that it can be stored much more easily or at much lower cost. But the challenge is that we don't really have a thermal infrastructure for moving heat at much higher temperatures. Once you go above about six, 700 degrees C, most of the thermal infrastructures are made out of metals like steel. Nickel alloy is more expensive. But we do have materials like ceramics that can go much hotter, but it had generally been assumed and believed that you can't really use those kinds of materials to build a mechanical system with seals and moving parts. But we challenged that assumption and this is why it was a significant breakthrough. This here is a little excerpt in a video from our first time pumping liquid tin at 1400 degrees C. The materials and the loop could have gone hotter than that, but we actually stopped at 1400 C because we ran out of heater power and we ran out of heater power because we cut a hole in the insulation so we could see it pumping. At this time, when we first did this, we didn't have flow meters that operated at 1400 C. But we've since invented flow meters that now work so we don't have to look at it to know that it's flowing and to measure the flow rate. But this was the first big breakthrough. This pump was a gear pump made out of aluminum nitride. And we've since moved on to doing pumps exclusively made out of carbon, made out of graphite that are centrifugal sump pumps that operate even better than this and can go all the way to 2400 C as we're interested in going in thermal batteries. I'll skip through a lot of this more quickly just to move ahead a bit faster. One of the main reasons that we use PV is because we can get similar efficiencies to what we can get with the turbine. This slide walks you through the basic energy balance. You can see why 50% is not unreasonable. This is using realistic properties of PV cells. So what's shown on the left is if we take the nominal intermediate temperature for us so we're swinging between 1900 C, 2400 C so halfway in the middle is 2150. So everything shown here on this slide on the left here is actually associated with the 2150 spectrum. The spectrum doesn't shift all that much over this 500 degree C window but the intensity of the lights shifts significantly so you get more power density at 2400 C than 1900 C even though the spectrum is rather similar. But what you see here is black body radiation shown in black. What's radiating to our PV is actually tungsten. So it's a shiny metal. So we actually get much less light coming out of the tungsten. So that's shown here in gray. And then what we're trying to convert is the stuff in orange and red. So we try to convert what's to the left of the peak of the spectrum. So that's about 30% of the light. The rest of this infrared is what goes through the cell. Some of it does get parasitically absorbed. We're trying to reduce that down to about 2%. If we get to about 2% absorption below the band gap then we will get to our 50% efficiency right now. That number is more like six or 7%. So that's one of the things we're working on is increasing the efficiency by reducing the parasitic absorption in the cells. But to show it to you here, this is tungsten's emission. We are only about 213 kilowatts per square meter is what's above the band gap. If you look at black body it's about two megawatts per square meter. So extremely intense light once you get above 2000 C. This is what makes our system very power dense and also keeps the cost low. Most of the slide here, 70% or so of the light that you can see goes through the cell just gets reflected by the back surface reflector that gold mirror on the back. A little bit of that is absorbed parasitically. So you get some absorption. This is waste heat. This is also waste heat QGEN inside the cell because we're operating at pretty high current. And the output power is about 100 kilowatts per square meter. So that's about the power density of what's coming out. Up here, if you now calculate the efficiency so you can take the 123 divided by the 123 coming out plus the heat generation, the parasitic absorption. And this 4.6 is what you lose due to convection. So we are not running our system in vacuum. We actually have argon gas in there and that argon gas will gradually convect heat from the hot side to the cold side. That's about another 4.6 kilowatts per square meter. Ultimately it comes out to about 50%. It's a little bit above 50% efficiency. And so the other key thing I wanna show on this slide is why we use multi-junction. You can see if we use one junction versus two you get about a 5% overall boost in efficiency. And so this is not a large increment in cost. Once you grow one cell junction using MOCVD, growing a second cell junction is not that much more expensive yet you get significant benefits on the efficiency. Another, the second key breakthrough so to speak or a significant advance we've had in our work has been of late. We had a paper come out of nature earlier this year where we set a new record for the thermofotable tag efficiency. So Dick Swanson at Stanford set a record back in the 1980s right around 30%. And if you look at the TPV literature over the last 40 years or so, most of the efficiencies were actually below his work. And just recently in the last year or so got to about 32%. We did this approach that I mentioned with multi-junction cells, a high quality mirror on the back and push the temperature of the emitter up to above 2000 C and then we get to a peak efficiency of 41%. This efficiency is not just at the peak temperature you can see we can get close to 40% efficiency over the temperature range of interest. We've got a couple of different kinds of cells that we've made. This has been in collaboration with Miles Steiner and Dan Friedman at NREL and Kevin Schulte. We've been in a great collaboration over the last few years and I'm really excited that they've been able to demonstrate cells that have a higher efficiency than what we've seen previously. And this is really what got me prompted to go ahead and found a startup company is once I knew that the cells would work and get to a range that's actually higher efficiency than what's needed to make money from arbitrage, then it became interesting to look at starting a company because I know that we've kind of crossed the last checkbox of what the system needs to work. What's also intriguing about this, again, previously TPV efficiency is around 32%. The average efficiency of a turbine in the United States is about 34%. And so this is the first time to my knowledge that a solid state heat engine has actually clipped the efficiency of an average turbine. And so turbines have always been the go-to choice for converting heat to electricity, largely because of their cost effectiveness and their high efficiency. Now TPV should enter that discussion and in some cases where it may make sense, TPV should be something that others look at and it's definitely something we're focused on. One of the other reasons I wanna point out that we're focused on TPV is specifically because of the lower barrier to deployment. So developing, if we wanted a turbine that's gonna operate in an inert environment at these kinds of temperatures, even if we go down to 1400 C, that's a $100 million plus for R&D effort. Conversely, there are companies in this space already that are very interested in working with us to develop the multi-junction PV that we need, much lower cost and we're getting a lot of good response in that field. This is just a conceptual depiction. So you can get a sense for what a system would look like. This is a gigawatt hour battery here. So 100 megawatts times 10 hours of storage. Each sub-component or subsystem in the entire system can have its own inert warehouse so that, for example, the pumps and valves could be serviced more frequently, have their own mini storage unit or mini inert environment. The storage can be its own place, the power block and the heater in their own place. The biggest portion of the system in terms of footprint is actually the dry cooling unit. Now let me briefly review some of the experiments and things that we've been doing in my lab over the last 10 years to kind of lead up to this technology and the founding of the company. So we've built a test rig that operates inside a vacuum chamber. We pull vacuum just to be able to get the gas out very quickly, but then we backfill with argons. We run at about one atmosphere of argon. We've got a system where we've got a graphite here surrounded with carbon insulation and also some oxide insulation. Initially we ran with, we've got the tungsten line and then we've got an actuated heat sink that we put PV cells on that we can put in the cavity, take measurements and demonstrate aspects of the technology. I'll move through this rather quickly. You can see here, heating elements, we make our own custom heating elements, get the machined out of graphite. This is the carbon fiber insulation you see here that the system rests on. You can get rigid insulation that can support the weight of different objects. This is us building the cavity. This is a cavity at 2,000 C. Again, the 2,000 degree C cavity in this image is about a foot back. So what you're seeing is just the light that makes it out of the cavity coming out here. Extremely bright, but this is us using our actuator with our water cooled heat sink that can go inside and carry itself. We can take measurements. Now let me get to this point about the deposition problem. So if you operate at 2,150 C, you put a heat sink inside of a cavity. Doesn't take but a few minutes. You put it in. If you don't have the tungsten liner and you don't do something I'm gonna show you that we do to protect the cells or to protect the heat sink, within minutes you'll pull it out and it'll actually be coated in carbon. And you'll see lots of black stuff. This would, of course, coat the PV and keep it from being able to operate. And this was one of the key problems that has existed with this technology and one of the reasons why I think it hasn't existed previously people going to these high temperatures. Most TPV work is done around 1,400, 1,500 C or even lower and largely to avoid this issue. What we did is we figured out that what you could do instead and we have a pattern on this is to instead, since I teach heat and match transfer an obvious solution is you actually blow gas. You need something that's transparent to the light that can physically move atoms out of the way without impeding the light moving through it. And so why not use a gas? So instead, what we do is we blow gas over the surface of the PV, of the PV, keep a thin layer of gas that sweeps away anything that would deposit so we can keep the PV clean and we just recycle it and filter it. So this is a nice kind of CAD image. You can see we have on our heat sink this approach of the sweeping noble gas curtain, the SNGC integrated into our heat sink. We bring in cooling water, have that return and also bring in cooling gas that we pull from the ambient environment blow through here to blows over out this gas cap over the cells and comes back in and gets sucked back in and pulled out. And so we were able to do this on all four sides and we have this all fully integrated in our system. We also use tungsten foil. This again gives us a four order of magnitude shift down and vapor pressure. This also helps suppress this effect dramatically. And we have some initial test data here showing. So we can put our PV cells in the cavity at over 2000 C for over six hours. We see very little impact, if any, on the PV. So this is us doing a measurement, same way we did the measurement of the PV efficiency. This is an IV curve for PV cell done before and after sitting in the cavity for six hours. So we were able to prevent this deposition issue and this ultimately allows us to operate at these extreme temperatures. So I'll pretty much stop here. What we're doing now in my lab at MIT, we're building a fully integrated prototype between one to 10 kilowatt hours with graphite storage and all these various components with pumping that can go through the full cycles and show the emitter deposition protection. And we'll do some long-term testing with this and hope to get another high profile paper out of this that'll showcase that the whole thing works when you put it together. At the same time, I founded a company, Thermal Battery Corporation. We're building a one megawatt hour pilot demonstration. It'll consist of a single repeat unit as you may have seen on the slide showing the full-scale depiction. You've got blocks. We're essentially gonna show and demonstrate one block that is then like a repeat unit with the appropriately sized pumps and heaters and PV sticks that go down into the light. So we'll show a set of graphite blocks that can then be repeated to build up to larger and larger systems without having to re-qualify the hardware. So I will stop there and be excited to take any questions that you may have and just one more time if anyone is interested in learning more about the technology, feel free to contact us at infoatthermalbattery.com and particularly if anyone's interested in possibly joining the company, do not hesitate to reach out. Ashay, thank you so much for that presentation and for the deep dive. Always interesting to see how you piece together different part of your research to come up with a grand technology. I think we have time for one quick question. So I think what is really novel here is the addition of the high temperature thermal transfer fluid here. And you touch upon some of the advantages, but I was wondering if you can also expand on other advantages of this decoupling beyond what you ever talked about. So if you separate the power block from the energy storage block spatially and also to some extent chemically, what advantages can you have that would not be possible if the TPV was inside your energy storage block? Yeah, so I would say the biggest advantage is that it allows us to independently size each subsystem so we could operate at 100 megawatts in one hour storage so we can compete directly with Lithium ion in the one hour storage regime. We also have the ability to go out to 100 hours of storage so we can explore the full range and resizing our system. It also allows us to decouple the charging and discharging rate. So we could, for example, design a thermal battery to have a, you know, five megawatt discharge rate and a 500 megawatt charging rate. There are some initial studies we've done show that actually there's some really huge economic advantages to being able to do that because particularly early on when we see more renewable penetration you'll have very few batteries on the grid and so when you have lots of solar that's overproducing you have the ability to charge and take up all this electricity that essentially would have a negative price, use it and store it and then trickle it back to the grid at a lower rate later when the grid needs it. So there's great advantages to being able to separate out charging and discharging. The third big advantage has to do with servicing the system. So by having separate buildings, separate containment vessels allows us to separately go ahead and shut down portions of the system. So the beautiful thing about tin is it doesn't, it melts as low as 232 degrees Celsius. Now the reason that's special, that's about the same temperature as molten salt. You can melt the liquid tin with a heat gun. You can do a variety of things that allow us to actually use the liquid metal to pull the temperature of a portion of the system down to where you can operate or where a human being can go and touch it with just some gloves. The liquid, you could even keep it in liquid state there or you can drain a section and put it back without having to go to really, really high temperatures to get it reheated and get it put back in and there's some nice overlap between the temperature where you melt tin which is 232 C, it's like solder and the temperature where like the peak temperature you can use like heat transfer oils, which is like 400 C. So you could actually use some oils to actually preheat the system, get it up to temperature then bring in liquid metal and do the rest of the preheating up to temperature and then bring it down. You can do all kinds of things by having liquid metal as a transport fluid. Ashley, this is very exciting. So in terms of economics, obviously you have to pay a price in CapEx to add this transfer fluid. What determines the tipping point when this becomes worthwhile? What are the benefits that you have to get out of it to increase the complexity of the system? That's an interesting way of framing it. So in order to, so I would say it's at the outset it makes sense. I guess the only place where there's a tipping point is if it turns out that for whatever reason the liquid metal infrastructure is not long lived, right? So it's cost, let me maybe pull up a slide. Actually, let me go back to the slide that had the economics. You can see the cost of the 10 infrastructure is actually not very significant. It's kind of negligible, let me see where we are. So it's really the lifetime of the infrastructure? Yeah, I mean, if for whatever reason so where we had transfer fluids, you can see here the transfer fluid is pretty small by comparison to everything else. So it's not a big deal for us to use 10 and gives us great advantages, great economic advantages, servicing advantages, decoupling advantages. So that's, so what you'd save is I guess an extra 50 cents a kilowatt hour, right? To not have the 10, but what it buys you is a lot of flexibility. And that's I think the one unanimous request and feedback we've always heard from the utility industry is they want flexibility. They want the ability to do different things. And being able to pull out, I guess I'll say this, the other thing is being able to pull the PV out of the storage, I guess by comparison to what I understand Antora's doing also allows us to have the very fast ramp rate. So we can fully pull our PV out, have our liquid metal running and our heat source is ready. And then when we're ready, we can dip in and get very, very quick ramp up of output. So there's a lot of another advantage with that is the liquid metal allows us a second degree of flexibility in how we discharge. So we've actually done some studies, we're gonna pull up publishing papers soon showing you could actually, so when you think about sensible storage, right? This thing is gonna gradually cool off. And what you would think is that you're gonna get a big drop off and power density as the storage cools off. But with liquid metal, you can discharge portions of the system in series so that you can actually have a constant output power almost the entire discharge. And so you can actually save the last few blocks that are at the peak temperature and have them discharged last. And you can have a very nice square looking profile for the power output. And the PV dipping in actually helps you compensate for that. And so you can vary the liquid metal flow rate to achieve that. So a lot of flexibilities, I think, what the liquid metal buys us. Excellent, Ashay, I think, you know, I've seen a lot of different types of energy storage solutions, but I think this is the closest it gets to a thermal flow battery. So this is really a thermal flow battery. This is very... And it's the opposite extreme of power density. It's extreme power density, right? And you have the true decoupling afforded by the flowing aspect. But yeah, we will have your colleague, Vic Bruchette, speak in a couple of weeks about electrochemical flow batteries. So this is, I think I'm really glad to see the intersection of these two ideas. Well, thank you very much, Ashay. If I can also ask York to rejoin us. We have about 20 minutes for a discussion amongst the three of us. And typically in this section, we try to weave the two presentations together and identify some common points. So maybe I'll start. I think what is really exciting in today's talks is that we are looking at the intersection of three to four main pieces, right? Electrical, thermal, optical. And I think this is sort of where energy storage is going. We have to go beyond just, for example, electrochemical. And also, sorry, I forgot to say chemistry as well in York's cases. To me, sort of the question I wanna start with is in both of your cases, you are increasing the complexity of the system, right? By adding additional things from those four, list of four, electrical, thermal, optical, and chemical. But you have seen that there are advantages of increasing the complexity of it. Do you think this will be a challenge in scaling up because you're embracing more and more elements of energy transfer mechanisms compared to simpler systems? Like batteries, for example, where it's only electrical and chemical. Maybe I can ask York to comment on this first. Okay, so it is true that we're adding some complexity. In our case, I think it's mostly the chemistry, right? That's added to it. We're, otherwise, we're trying to keep the system as simple as possible, right? So for example, particularly in that if you're only interested in the heat aspect of it, that your heat transfer fluid is air, there's actually no heat exchanger because there's direct heat exchange with the air. That's also the reactant. So we're actually trying to keep it as simple as possible. And in terms of scaling, most thermal technologies, and I think ours and Auschwitz is not an exception here, scale well, right? They get better when they get bigger. And I can tell you, I actually, I showed in my talk that the 100 watt hour system, we barely made five cycles. The 10 kilowatt hour system where you actually have some significant storage and your losses are only 50%, not 95% anymore, things actually get a lot easier. We were able to more or less automatically cycle. This thing has been running for 18, almost 2000 hours. In the lab and like I'm not, you know, fortunately I'm past that age where I sleep in the lab. So it's, and neither do my grad students. So it's actually pretty much, I mean, we watch it, but it runs on its own. And so I think scaling is actually a good thing. Yeah, I guess my opinion about this, I share your perspective on scaling kind of actually helps us in the sense that it's actually kind of harder to do it when it's small, because we're space constrained, particularly inside of our vacuum chamber. Like once we have more space to move around and we're not so worried about things touching and all these kinds of things, it actually gets easier. The other thing I would point out in terms of complexity, I think that my perspective about this is actually was heavily informed by a very early trip I took when I was an ARPA E-fellow, I got a chance to go to one of Austin's turbine manufacturing plants. And when I got to see, I mean, you learn as an undergrad, you know, you see the picture of a turbine, you draw a little turbine on a board, but until you realize how complex, I mean it's one of some of the most complex machines on planet and they, and they, and they comprise like 90% of the entire electricity infrastructure. Like this is where electricity comes from. It's super reliable and it's super complicated. But that level of complexity is managed. This isn't a household product that people get to like an iPhone and people get to drop and do whatever they want with. This is a heavily managed, there's a team of people that surround this thing and watch it and take care of it. And that is the way the power industry works is they want something very reliable. And at the end of the day, if you think about it, you think about how many heat engines exist. I mean, it's all kinds of Ericsson cycles, it's a sterling engines and all, it's all kinds of cycles, but one, one, one out, one of them beat everything out and became the dominant technology that's used and proliferated around the entire world. And that is because it meets the needs from a performance and cost standpoint. It doesn't matter that it's complex, right? It's actually, it's embraced, right? It's like, as long as it does what we needed to do. So if there was another technology that happened to be simple, like a sterling engine is simpler than a turbine, but it doesn't win on cost effectiveness and efficiency. And so we use the more complex technology because it works better because it actually, it meets our cost targets better and our efficiency there. And so in this sense, I really encourage my team to really embrace the complexity. It's okay for it to be complex. As long as we understand what's happening and we understand how it works, it's all fine. Simplicity, I think, is nice from the standpoint that it's easy for people to understand and easy for people to follow when you're trying to explain it to folks, but at the end of the day, I don't know that simplicity really buys us anything unless it's like making it easier to manage and operate. But maintenance costs don't usually drive these things. These things are driven by capex. So better to make it complex and it works and it lasts longer than to try to make the maintenance costs, which is only 10% of the cost anyway, a little bit less. Thank you, Ashi, and you're giving me also maybe at my two cents very much resonant with what you said. I think where you have complexity, but you're able to optimize the individual blocks separately, which I think it's a point that both of you mentioned today is really key. For example, in the material science field, we often see this all in one device. And I think that's where the problem comes in, because when it's all in one, then you can't really optimize it separately. For example, solar cell plus electrolyzer is a really good way to make hydrogen from water and you keep it separate and you can optimize it independently. But when you put it together, it gets much more difficult. It is simpler when you combine the two into for example, a photoelectric chemical. So from an engineering perspective, it is actually not a better approach. So I think I really resonate with this. I think it's complex, but can be optimized as unit operations and achieve their maximum cost and performance tradeoffs between the two. Really completely. And maybe to add the, in both technologies we saw today, you have a separation essentially of the power block and the end and the storage, right? And we can scale the storage just to the size that is required in a given application. And finally, to complexity, there's technological complexity, but there's also psychological complexity, right? And Ashley, you said it, right? Like all electricity comes from like today, right? It's all, there's a little bit of hydropower and there's a little bit of PV, but it's all thermal, right? So utilities know this kinds of systems, right? They integrate well with what's there, right? With the infrastructure that's there. So speaking of integrations and dropping in your solutions into the grid, certainly on a system level, this is a no brainer, right? This is when the technology works, this will be easy to drop in. The one aspect I think that does stand out from adopters of technology such as yours is that you're going fundamentally to a considerably higher temperature, especially in Ashley's case, to really derive the better Carnot efficiency. Do you expect this to be a bottleneck in terms of stability, in terms of customer adoption, going to 2,000 degree C or 1,400 degree C, which is considerably beyond some of the current thermo cycles? How have you been interacting with prospective customers in this area? Are there any psychological issues there? I can go first. This is a question that we've been asked, right? Because if you talk to a turbine guy or girl, they'll say, oh, 1,500 degree C, right? No, we can't do that because it's all nickel alloys. They're all like 1,260-ish degrees. Now, we don't see this as like for our technology specifically, we don't see it as an issue because the way we do this is we have a bypass valve that, and I have not shown it in the slides, not to make things too complex, but actually there's a bypass valve where over the discharge of the system, we're bypassing a varying amount of air to keep the turbine inlet temperature constant throughout the discharge, right? So what the turbine actually sees is pretty much the same temperature as it would see in the case of a natural gas turbine, right? So in our case, that's something that has come up and it requires a little bit of explanation and education and the 1,500 degree C in our case are actually necessary to have an essentially constant 1,200 degree C or 1,100 degree C, right? Even with the chemistry that you still need a little bit of drum to control that. And we're not as far off. So an engineer usually gets it pretty quickly. Ashi? Yeah. I don't, I think what puts people at rest or at ease is two things. One, when you realize that tin's oxidation is very weak, it's not like one of these alkali metals that will explode when it gets water on it or something like that. So it's a very weak oxidation. It's like solder, you can melt it in air, it'll form gradually form a little oxide crust. It's not dangerous in that respect. The second fear people usually have is you get something this hot, imagine it like spills out all over the floor kind of thing. That's actually kind of hard for it to do. We've had spills before in my lab. The reason is because it's shrouded in a big, thick layer of insulation. So as it's working its way through the insulation, the insulation temperature is decreasing and it cools and it freezes. So it actually freezes usually inside the system. It's very difficult for it to get out and actually pour out anywhere. I guess this isn't necessarily put people at ease, but it's I guess an important reality to think about. If you think about someone like being burned by it somehow, right? If someone drops 2000 degree C metal on your foot, it's the same basic outcome as if they drop 600 degree C salt on your foot. I mean, it's super hot, it's gonna burn you and it's the same basic outcome. So the safety issues are not really all that different than molten salt and there's a hundred of those plants around the world that operate pretty safely and are now considered essentially a proven technology. So we're able to really piggyback off of a lot of the safety infrastructure procedures and constraints associated with molten salt technology. And actually just to build on that, in terms of regulatory compliance when dealing with high temperature heat and material, I presume it won't be very different from the steel making industry. I think York mentioned blast furnaces. It says it's already come up in terms of how this could get approved as a project. Yeah, I think it's the same as you mentioned but we have an extra layer of protection which is that ours is in a warehouse with inert gas. So this isn't something someone's just gonna walk up and touch, right? Like in order to even go in there you need an oxygen tank. So there's so many layers of safety between us and the actual device, the actual system that most of these things are mitigated but you're right. I mean, you say 2400 degrees Celsius and it surprises people but then when you start to walk through the advantages then they start to realize, I mean, you could right now go turn on an incandescent light bulb and you can bring your hand within two inches of something at 2400 degrees Celsius, right? I mean, people have this in their homes. It's not as though these temperatures are never achieved anywhere. It's just that people are comfortable with that. Right, so Ashley, we have a couple of minutes left. I'd like to maybe ask one last question and get your thoughts on this. Both of you stress the importance of diversity in the most of operation, whether it's the duration of storage and other characteristics of your battery. When you speak to your potential customers and customers can you give us a sense of exactly how, what application can they enable when they have a storage that can go from days to season, right? Because that doesn't really exist readily elsewhere. And then just for our audience too, this is your regular spider plot of energy storage where you want your low cost, you want wrong trip efficiency, you want diversity of operations. It's really hard to get one technology that does it all but for more we're hearing today, Ashley and York, I think you check a lot of the boxes there especially when you can deploy at a scale. So I wanna get a sense of what your customers and downstream, what can they do with your technology that they can't do would say something that only checks a few of those boxes and cannot be used as diversely as yours. Maybe I can ask York. Right, so I'm actually at the moment in conversation with a range of more small and medium sized companies that require process heat. A lot of them is high pressure steam. So we're talking a few hundred degrees C that is currently gas-fired. And just to give you an idea, currently a spot market natural gas prices in Europe are somewhere between a hundred and 200 euros slash dollars per megawatt hours, which is crazy. I don't know what's that in the MMVQ, but it's a factor between five or more of what prices in the US are currently. Electricity is actually quite a bit cheaper, quite a bit, like factors. And they are just worried about if you have energy intense, heat intense processes in your company. And we're talking to packaging, food, dairy, glass, although that's a little harder. A lot of metals, melting, like aluminum smeltering, ring melting of metals. Pop and paper, a lot of it is, I would say about 50% of the people who are currently, or the companies who are currently talking about to steam, high pressure steam, a couple hundred degrees C steam on the heat side. And then the rest of it is a little bit more diverse. A lot of just gas heated processes that where you basically need a stream of hot air slash hot combustion gases. And this varies a little more drying processes, for example, where you would use it just directly, the flue gases. So it's a range and then a little further down our pipeline, we're talking to utilities where it's basically, it's gas turbines. So that's something they're very comfortable with and very familiar with. But in terms of the product pipeline, it's just a heat application where we're just the closest to actually getting up. So it's a range and then a little further down our pipeline where we're just the closest to actually getting a product into the market. Ashay. Yeah, I think that, I think it's going to take a while to do essentially customer education. And I think that, when you talk about the utilities, talking about the central background of the entire modern world and what allows the modern world to be what it is. And it is one of the most risk averse industries that exists takes them a long time to get used to an idea before it's like widely adopted and accepted. So the time scale here for adoption is like 50 years, takes a long time for a particular idea to percolate and really catch on and really get widely deployed. I think that one of our approaches and my company that we're trying to take is to try to simplify things for the customer. I think one thing that, like we don't want them to have to go create a new storage market. Like that's like asking the customer to do too much work. Instead, what we're trying to do is look at pairing the PV or the renewables with the battery. Because one thing that many utilities have done over the last decade or two is they have gotten in the habit or gotten used to creating a mechanism for them to put some form of renewables on their grid. And what we want to market it as is we will sell you renewables. They just, you get to tell us when you want it to turn on. And we handle the complexity rather than forcing them to handle the complexity. So I think that that's part of our approach in trying to ease the transition. But I think it's gonna take a while before they actually realize all the things you can actually do with this technology. There's so many features to the flexibility. I remember being at a meeting some years ago and there was a guy who was a grid operator and he was giving a talk and he said, if anyone ever wants to know what it's like to be a grid operator, it's like trying to drive on the highway at 80 miles an hour and you're blindfolded and you only get to open your eyes every three seconds for a split second. And that's what it's like trying to manage the grid. Like all of a sudden a car could show up out of somewhere and you have to very quickly adjust and they're constantly trying to balance the load. And so now realizing that they may have a resource that could ramp at gigawatt scale. You could get an extra gigawatt or two or 10 gigawatts all of a sudden in a matter of seconds if you deploy it, I don't think that's not a capability they've ever had. And so teaching them that that's a possibility is gonna take a while. Thank you, Ashe. Thank you, York. This is truly exciting, very unified, I think presentations from the both of you conveying some of the same messages with slightly varied versions of the thermal energy storage technologies. Great, best of luck to you both in your academic turn entrepreneurial ventures. Really excited to continue to follow both. And as Ashe and York mentioned, they're hiring like crazy. So I'm sure there are many students and recent graduates were looking for positions. Please do reach out to the both of them. Like to thank you both again, Kaylee and Evan Capenda closing slides please. So this is the first of our summer presentations for StorageX. We have three more in two weeks, four weeks and six weeks talking about flow batteries, aqueous energy storage and also material informatics for accelerating energy storage R&D. So please come back and attend these talks of your scheduled permits. And as a reminder, please stay connected with us. And if you're interested on learning more about energy storage and other energy transformation technology, Stanford also offers online education and you can find more details on the link on the right there. Once more York and Ashe, thanks so much and we'll see you in person soon sometime in the future.