 Hey, good morning. Good afternoon. Good evening, everybody. Welcome back to Astrology X symposium again. My name is Yi Chui. I'm faculty co-director of Astrology X Initiative at Stanford. I also serve as the director of preco-institute for energy. On behalf of my co-director, will chair, I would like to welcome all of you back again. So today we have three very exciting speakers who will do a deep dive of thermal energy storage. These three speakers are Ravi Pressure, lab associate director of Lawrence Berkeley Lab. I will mention his background a little bit more later. From Google X, now it's Alphabet X, Moonshock factory and Brock Forrest, the CTO of NetPower. We will introduce the latter two speakers. Now let me invite Ravi to the stage. I want to give you a little bit introduction about Ravi. Ravi is very well known in the energy field. He's the associate lab directors of energy technology areas and also the interim division director of Cycletron World at Lawrence Berkeley Lab. He's also an adjunct professor in department of mechanical engineering at the UC Berkeley. He has a lot of experience. He used to be an entail before and later he was recruited to US Department of Energy for the RPE. He was one of the first program directors and while he was there, he just started a few really exciting programs related to thermal, related to energy efficiency. One is beaten, the other is heat. So Ravi is a very dynamic person to interact if you know him. I myself just enjoy so much of his intellectual horsepower as well as his leadership for the energy research field. So with these short introduction, Ravi, let me invite you to give us a presentation. We'll do 20 minutes presentation, then we'll move on to the other two speakers at the end we'll have a panel discussion as well. Ravi, please. Thank you very much. So, first of all, thank you for the kind introduction he both he and well, thank you for inviting me for this store x symposium. I've been a big fan of it and try to attend as many as possible. So I'm going to talk about a topic or dynamic and tunable thermal storage and transport. I'll touch upon the transport part a little bit. There have been few other, I believe talks related to thermal energy storage in the past. My talk is a little different in sense that the other two talk that I attended at least one of them. And it is basically I believe where thermal energy storage was used to actually finally store electricity so you have electricity as the input and electricity is the output. Whereas here, what I'm going to talk about is electricity is the input but the output is thermal energy. All right, so before I get going, let me acknowledge my collaborators. Because the work that I present this has been done in collaboration with these people here. Professor Chris James from UC Berkeley. Dr. someone called Dr. Galu and Dr. Anwar Jam from LBNL and Dr. Roger Jackson from Israel. And more or less all the funding for the work that I'm presenting here has come from the building technology office of us duty. So, why is it important, why is thermal storage important. Particularly if you're going to store it as the output from that storage is going to be thermal. The reason is that if you look at, if you look at the building sector, let's say starts in the United States, buildings consume roughly 75% electricity today. Right, I mean, of course electric vehicles are going, come on going to come on board but the storage is already built in the car. Building cooling is on the rise in the whole world, significantly. In fact, here I'm showing you in this graph, peak electricity demand for air conditioning. Just in two countries, India and China, and as you can see there's a dramatic increase already today roughly one terawatts of power just goes in supplying air conditioning door for India and China. The world electricity production today 70 terawatts. And if you add rest of the world including including United States, significant amount of peak power just goes in supporting cooling. Building heating will get electrified because primarily the building heating today is supplied using natural gas, almost all over the world. So it will get electrified using mostly heat pumps or maybe some resistive heat, right. That cooling and heating is going to be more than 50% of electrical load in buildings. So now let's look at some scenarios in the future. Today, we have mostly as we all know the peak electricity load on the grid is primarily driven by cooling. And you have mostly the peak happens during the summertime, almost all across the country you can see here. But in 2050, as I mentioned earlier, expectation is that almost all the heating will get electrified. Then we will start to see a lot of peaking electricity demand even in the winter time the blue is the wind winter month and the red is the summer. And, and as you know if there is peaking and registers can make a very, very big difference. And then, apart from peaking in the load side, there is going to be a mismatch because it is expected that significant amount of the energy will come from the enable sources. So, so that could be significant mismatch between where the demand is and where the suppliers. And that's why energy storage is very, very important. The question is that how much of whatever the stories we do in the future, there's a lot of different technologies are being developed as we all know, and it gets covered in different toxins stores X symposium. Question one of the questions we recently asked is that doesn't matter where you put the storage whether it is on the grid level or it is distributed, or it is a community level doesn't matter what kind of storage it is. So that storage will go in supporting building thermal loads, right, a simple question. So here I'm showing you some some analysis of this is we did many different analysis but I'm just showing you two examples here. Let's assume for the time being that this is for 2050, the year 2050. Let's assume that 50% solar and 50% bin is how we are getting electricity. This is a typical summer month and 24 hour period. The black is the supply curve for the electricity and and the red and the gold are the demand that is the thermal load and the gold is the non thermal load, right. And you can clearly see here that these are sold our month hours, where do we need storage because you just don't have any electricity available, and you have access supply here. So that's true for the winter month as well. So now we can calculate overall storage requirements just to support heating and cooling. What we found out for various scenarios all the way from 100% solar and 0% vent to the extreme of 0% solar and 100% vent across whole country. Yes, almost all the stores will be needed in building sector to support the thermal load right there will be some stores needed for the non thermal load, for example running computers and washers and dryers and other things. Roughly I mean it changes all the way it can the number changes all the way from one terawatt hours to five terawatt hours, but average is roughly 2.5 terawatt hours right less than less than, and then we call it short edition less than 10 hours. Okay. So less than 10 hours of storage just to supply thermal load will be roughly 2,500 gigawatt hours or 2.5 terawatt hours. This is equivalent to roughly 50 plus million miles, the millions of electric vehicles. So that's not a small number. We are talking a fairly significant number in terms of the amount of stores needed to support the thermal load in buildings. And then if you will simply long duration storage which we are not going to talk about today, greater than 10 hours then that number is more than 110 hours right. So now the question is the question that we started asking is that does it make sense to store the energy in the form of the supply, which is electricity here. We are assuming that both the cooling and the heating is coming from a heat pump. So now the question is whether we want to store the electricity in the form of supply, or we want to store the energy in terms of the demand itself right. So if you're storing in terms of supply that basically you have electricity coming in and storing in a battery, either at home like a Tesla Powerwall, or you're storing it somewhere on the grid, and then running a heat pump, right, and then you get the you know, energy from the cooling and heating whenever you need it. Other one is that you have electricity coming in you run a heat pump here, and then you store that output is thermal with a cooling or heating load energy, and you store it in thermal battery or thermal storage. There are two ways you can do it either you can just have a box integrated with a heat pump or you can distribute the storage in the envelope of the building. So the second scenario, which is why I'm focused on this in this talk is electricity coming in, and you store something here in thermal energy storage and output is heat. Because of that, you know, things become a little interesting when you want to compare the cost. One other thing we started looking into the can I do an apples to apples comparison on the levelized cost of storage. And to our surprise, very surprising that that study did not exist. And it is a little interesting is a little complicated is because you have electricity coming in a heat going out so you got to apply the second law of thermodynamics, right. And that comes into picture when you are comparing the levelized cost of storage. So that comes in the form of COP coefficient of performance, right. The coefficient of performance is depends on the ambient temperature, because you are basically pumping the heat from the ambient temperature. And then then there are other factors as well the discounting rate utilization which is basically depends on the depth of discharge and number of cycles. These are similar for thermal and electrical for people who do this so much of study on levelized cost of electrical that this should be obvious to people who are doing that electrical. So we do the same thing for the thermal, except for the COP. One big differences, the same the COP depends on ambient temperature, you can really tap into the diagonal temperature swing. Okay, because what you can do is, you can do the storage when the ambient temperature is low. So let's say you're pumping like energy to do cooling, you can use the low ambient temperature you can effectively use, increase the COP and store the energy in the cold storage when the ambient temperature is low say in the night. And then you use that stored energy, the thermal energy during the daytime when you really need the cooling. So that offers you some unique advantages in terms of thermal storage, just tapping into the diagonal temperature swing. So here are some results, I'm not going to bore you with most of these charts here, we analyze various scenarios. One is one, but very important factor is that this idea of tapping into the diagonal temperature swing. For example, Denver, Colorado has significant diagonal temperature swing, whereas San Diego, California does not have significant diagonal temperature swing. So the results differ, and storage becomes a little bit more attractive for Denver, Colorado, for example. But what we found out that the state of the art thermal energy storage, the levelized cost is significantly lower than state of the art lithium and battery. And you will also see that utilization here for thermal energy storage is smaller than utilization for the lithium and batteries, right, because challenge with thermal energy storage is right now the way the technology today is, either you can store heat, or you can store cold, right. That means you won't be using it for throughout the year, whether that's the beauty of lithium and batteries, or any electrical battery that you can use it every day, right. So utilization is significantly higher. But still, the levelized cost of storage is cheaper, primarily in most of the cases, not all the cases, is because of increased lifetime so thermal energy storage can survive for 30 years. There are a lot of field data available also. It has much cheaper capital cost. And also it can be potentially used to reduce the size of the heat pump itself, heat pumps are very expensive. So you can tap into the diagonal temperature swing and you can actually make this heat pump smaller. It's almost like a hybrid car, where you can have a smaller engine combined with with the battery, right. And then further reduction is possible if you can increase the utilization of the thermal energy storage and you bring it from here to here, right. So the question is, if it is so simple and it looks so economically attractive, why isn't why don't we see it everywhere right. That is where the research becomes very interesting. And the question is very interesting because there are some fundamental materials and chemistry question that that that's the reason, you know, it is not ubiquitous. Since this is a symposium organized by Stanford University. I have to tell you maybe a lot of you are aware of it that Stanford already uses thermal energy storage to supply the thermal load of in the campus. So we have this is a hot water tank, these are these two tanks keep chilled water, they run a heat pump and and when the electricity cheap, and then use that energy to store hot hot and cold and then use that to supply all the thermal load, I believe almost 100% of the load is catered through this by this, these right. That's great. Sanford is a has a beautiful campus it has a lot of land, a lot of area and you can afford to have very gigantic tanks. But that's not the luxury that we have in all the houses as well as apartments right. And the reason is that fundamental reason the energy density of thermal energy storage is low. For what I'm showing you here is a thermo physical storage which basically means just like these tanks here is a heating or cooling right a material or you're also melting a material I'll talk about that later. So that is face change based storage. You can also. So here you have the face change or the or the thermal physical store is a pretty good volumetric energy density but very very low gravity, gravity energy density. Whereas, thermo chemical storage where you can store the energy and chemical bonds have very good gravimetric energy density but very low volumetric energy density. The reason is because most of the chemical storage uses gas phase so solid and a gas phase okay. This is at the material level six somewhere here. So that's a very big question. How do we increase the energy density. So there are a few potential strategies. One is if you can distribute the stores in the envelope of the building. I'll talk about that in the next few slides. Then you can solve the problem, or you can develop new molecules for high energy density. I'll touch upon that a little bit. And also maybe you can come up with some innovative system design. So let me start with the distributed stores and envelope. Any building has a lot of wall area available as we know that right. So imagine that if you can embed in this wall or thin slab of a face change material. Okay, and then you can store the energy by melting and solidifying this material right. We did some math, a typical 2000 square feet residential building can store roughly 150 kilowatt hour thermal thermal energies in it right. This is equivalent to nine hours of putting in a typical building. Okay. However, there's another challenge here that you need, you know, when you store the energy, the energy will start to leak. You know, because at them, there is no perfect phononic insulator, the phonons leak, the heat will leak. So we also need some kind of a thermal switch. Okay, and that's where the tunable transport becomes very, very important. So now, and the idea of putting a storage inside the building is not new people have an envelope but idea of combining with a thermal switch, I believe is fairly new. And that's what we have proposed and we got some funding to do the research on that. But let me give you a quick example of, of show that how the switch helps. This is from our collaborator and well. So you have, you know, the typical wall, we have insulation and other material in the wall, you just put a face change material. First of all, if you don't put the face change material, one day cooling load thermal load in this scenario is, you know, 1.52 for example. And then you put the face and material it reduces the load by almost 15%, right, but see not jaggy significant reduction. So the thermal switch here, it dramatically reduces the load, right, because the switch lets you really control the flow of energy, right, and that really helps dramatically so switch has a big impact to play. As I said, the idea of putting storage in the wall is not new. You know, people have tried it. And I think there are startups as well. However, the field data actually is, is not very promising. Here's a few data from one study, where you see that some of the walls in that that building where the field data was collected was 0% active right that means facial material never changed phase. Okay. Other times we're only partially active right. Why what happened here is, is I'm showing you something called a patient indicator. It's almost equivalent to the depth of discharge in a battery. Okay. Whatever means is completely solid state, and one means it is completely melted all the material is melted in between means that partially melted partially solidified right so that's like, just like lithium and butter you have a depth of discharge in less than one. And you can see, it's most of the time. Nothing that it doesn't change the phase at all, it just as stays stagnant, right. So that means the materials is the system is just inactive, you have a dead capital sitting there. So traditional approaches has not worked, and one of the main reasons is that ambient temperature fluctuates. So what happens is that your facial material fixed melting temperature and I mean temperature fluctuating dramatically. And it is not changing his face because depending on the ambient temperature it's not changing its own melting temperature and behavior. So there is a study which was done, which shows that if you do not have if you have a traditional facial material, whether with a fixed up facial temperature. It is just most of the time it is either in a liquid state or or in solid state. But if you could design a thermal storage material where you could change dynamically tune the face change temperature matching some of the changes in ambient temperature. Then you can increase the utilization as you can see back and forth is a lot of charging discharging happening. And then it increases utilization of the store material by 20x. Continuously transition temperature ability to store both hot and cold can dramatically increase the overall utilization of thermal energy storage material particularly if you put it in an envelope by building. So here's the idea, you know, you basically here's the ambient temperature variation and you have a flat traditional material as a flat melting temperature but if you could match this profile that will be great. And that's what we recently just published our first paper, where actually we learned from the, you know, the little chemical community that and this is we know already that if you put ions and salts it changes the melting temperature. The transition is going to be can you dynamically tune it so basically, can you put ions in and take ions out. So, you're basically you can change the transition the pressure from, you know, 18 degrees for example in this case to 25 degrees depending on how many ions are inside this transition material, in this case it's a peg. And so here is fast data that this paper just came out a few days ago, where we have shown that we can dynamically tuned by applying voltage, the patient temperature by almost seven degrees. And, and so and that's one aspect. Other one is that this also shows some promise so that we can combine thermal energy storage and electrical energy storage in one device that can hopefully increase the utilization of that device, but the very preliminary but it looks very very promising at this point. Let me move on change gears a little bit I talked about some other ways of storing energy. One of them is in the form of fluids, which is always used used for both the heat transfer and storage micro electronics this is where I started my career, significant interest in doing liquid cooling. This is the patent or Tesla patent where the all of us know that Tesla uses liquid cooling for to cool the batteries transformers use liquid cooling, you know, in terms of mean oils. And then there's a lot of interest right now there are quite a few startups I just took one, I think it's company called harvest thermal, they're using again heat pump and a storage tank to store the thermal energy. So if you're using fluid than the heat transfer rate and the storage density is directly proportional to heat capacity. So, so, so the question is, how can we increase the heat capacity of fluids. And that's what that this become interesting. Most of the thermal fluids artificially designed thermal fluids are basically vendor was bonded molecules, right. And, and the bond energy as we know is pretty low for less than five kJ per mole. That's basically translates into low heat capacity. Water has the highest capacity, but 70% of the heat capacity in water comes from hydrogen bond breaking, which we is like 10 to 30 kJ per mole. We use covalent bond breaking, which has very high bond energy to beat water or achieve high heat capacity and also one of the challenge with water is that as you know it freezes zero degrees Celsius and balls at 100 degrees Celsius. So, question is, you know, can we also expand the temperature window, right. So there is a class of reaction which is very well known in the chemistry community called the also the reactions of it was actually proposed in late 70s and early 80s. And the reason is that it has a very high reaction enthalpy, which is great for the heat capacity, but it also has a very high entropy of reaction. That's very important, because if you have a low entropy reaction then the turning temperature will be very very high. High entropy reaction allow the turning temperature to be moderate. Many times this reactions happen in liquid phase. So that makes it very interesting from the point of view of designing thermal fluids. We just got our first data where on some well known molecules. Here's the heat capacity, blue is the data of this thermochemical fluid. And this is water here right here in a narrow temperature in now we are able to beat water, right. Water is never hardly used pure in heat transfer fluid directly you also mix it with anti freeze like I think like a propylene glycol. And the propylene glycol and three glycol heat capacity is here much much lower. Right. And the mineral oil is another very well known heat transfer fluid and stones medium. It is right here much lower right so you can see that the heat capacity much much higher than most of the other liquids and also a little higher than that of water. That's one viscosity is also very very important. Most of the liquids have very very high viscosity except water. This one has also a pretty decent viscosity, much much lower than some of the liquids. So that also makes it very interesting. And importantly, we also used DFT to see whether we can predict the performance. So DFT can predict the enthalpy reaction, then we use the airing equation for rate of reaction, where I can DFT can predict the rate of reaction kinetics. And then the data that the black is the DFT result and, and the color is the data and it's not fairly well. So that allows us now to start investigating that can be designed the molecules So we are starting doing some work where we are starting putting functional groups on the different molecules and you can see that we can tune depending on what the functional group is, we can tune the transition temperature. Like for example in this case the peak temp heat capacity is happening around 70 degrees but in another molecule the peak heat capacity is happening around 110 degrees. So, again that that work is going on now because now we have confidence that DFT does a pretty decent job in predicting. Very quickly I said systems innovation can also lead to some dramatic improvements. This was a project that I had funded at MIT when I was at RPE as a program director, where this team from everyone's group they combined heating and cooling into one system. And actually they used an absorbent here. And, and instead of just using ice they use water vapor, except the vapor was not kept in a vapor state but it was absorbing and absorbent. That leads to significantly increase energy density. That's a really good example of a system level innovation and also, since you're storing both heat and cold in one system. That means utilization can go significantly higher because you can use it almost throughout the year. You can use it for heating during cold months and then you can use for cooling during the hot months. I want to quickly touch upon this in a minute or so then then I'll wrap up. As I mentioned earlier, that the thermal switch can make a very, very big difference in the performance and thermal switch has a lot of other applications as well. Actually, I want to highlight some very, very good work coming out of electrochemical thermal switches. On the right hand side you see this is what came actually came from Ken Goodson and he is group where Aditya Sood was a PhD student and, and he used they showed that using ion intercalation, they could change the thermal conductivity by almost an order of magnitude or the thermal conductance by on an order of magnitude, right, and you can basically have a switch is all it is, and then you can control the conductance. So, and, and, and the very first work actually came from David Cales group where he looked at lithium and battery and lithium cobalt oxide which is the basic material in the cathode site. And you could show that by intercalation that the lithium could change the thermal conductivity quite significantly. So anyway, so we feel that this is really interesting direction it has a lot of other applications. The last slide is that we are working on another method where we are just using shape memory alloy, which allows you to contact and decontact two surfaces together. And that can lead to a significantly higher switch ratio, and we are developing some devices for buildings application with that I'm just going to stop and leave this slide in the background. Thank you. I think you're, you're on mute. Yeah, yeah, yeah. Okay, got it. Hi, we thank you for the very interesting talk. So let's ask a few questions for now then we'll come back to panel because the later topics also related. When you show that picture, the Stanford's our hip arm, his storage cold storage, right. I was thinking. For single family house. You are certainly not for apartment for single family house. So you probably have done a calculation if you think use the same idea. Hot and cold storage right there. What kind of footprint size right expected whether that makes sense you're using water tank idea for single family I want to pick your brain a little bit on that. I mean I quite frankly I feel that that's going to require quite a bit of space okay like if you 50 gallon water tank is originally just to supply hot water for shower and stuff right. Imagine that now you use that water tank also to supply space heating, for example, right. That's going to increase the footprint a lot. One way to do that is also that you do underground right I mean you if you have you can just bury that thing inside the ground. That way at least you don't have to worry about the footprint, then then then you have to really look into the world cost of excavation and stuff as well which are not looking to right. So that's one second challenge is that, you know, if you put it outside the garage, right, you know, because you have that I can hear about freezing and stuff depending on which climate you are in. So, so, so I think one other thing which is very interesting is if you can look in a phase change material because the phase change has much with higher energy density, compared to just, you know, storing energy in liquids. But, but, but I think there's a complex question I would say that on one hand from our point of view resources point of view looking new molecules it would be exciting but from from practical point of view also ideas like bearing some of these tanks and background could be equally interesting too. Yeah, yeah. Thank you very. The second questions I have is, I really like your idea, going to a direction explore the kawailam bomb for storing energy right so and your initial exploration is really really interesting. There's really a lot of energy certainly now the question becomes whether you can find something that allow you to reversibly Tony back and forth, you know, and without too big energy barrier right with. So, it's very interesting, I want to ask, you know, in terms of bomb formation kawailam bomb is one type. Another bomb is I only bonding. Sometimes maybe I because kawailam bomb is so directional it's actually this quite a bit of challenge to find the right system but I only bonding could be system. Have you look into that that's I only bomb breaking in reformation, it's electrostatic, mostly with that nature. Would that be easier. This is this idea just come up listening to your talks I have no idea so that's I that's a very good that is why I put it here in the talk because I think I would say put it this way right. I think we need to go into molecular scale understanding to if you want to design new molecules and new ways so the game is on whether it's ionic or covalent or hydrogen bonds. And I think I would, I, we have not personally looked at it. Right, but on the other hand, I think that is very impressed, even even I would say one for the height height of the storage which which is which you use for electricity storage like you know you're going to, if you want to run a heat engine based on that storage. So, I think this field is very open in many ways, both from fundamental science as well as engineering point of view. So yes, I we have not looked into this. Yeah. So, right we let's take one more question that let's move on to the other talks. So one, one person asked, does the heat capacity of the phase change materials, when is it liquid state. Can not deliver some energy during hotter summers, because it's already a multi state and I mean, this is for clarification like it's doing hotter summer unique cooling then that's that will not work deliver some energy and that's I mean it can I mean I quite I didn't put it here you can also combine phase change and just sensible. For example, this is an example, you can make ice during the summer month and you can use that same thing to heat it up during the winter month, right. So goes water to all before it becomes steam. Yeah, but again there are some engineering challenges that somebody will need to be able to figure it out because when it becomes ice heat the thermal conductivity and heat transfer is very very different in the solid state than it is in the liquid Yeah, how do you manage the engineering and the heat exchanger and stuff will be different but but on the other hand potentially it can increase the same system can be used throughout the year for both heating and cooling. Yeah, so. But that will be some very interesting engineering challenges. Okay. Thanks very for the time consideration, a very just a few question and Q&A if you want to type in the answer to the those folks feel free to do so. Now let me invite my co director wheelchair to the stage introduce the other two speakers. Thank you very much, he and Ravi thank you for the great talk and setting the scene for thermal storage. I have many questions I'll ask you later as well. So if I can also invite Adrian and Brock to the stage please. Perfect. Wonderful. So you've heard quite a bit from Ravi on the materials development and introduction to thermal storage and Adrian and Brock will now take this into the systems level discussion. So let me introduce Adrian and Brock a bit more. So Adrian is currently with Google in the moonshot factory. And previously, she was systems technical lead and Malta, which is developing thermal energy storage. And she is also like Ravi spent time at RPE in the Department of Energy, where she was a fellow. And Brock for us that he is the CTO of net power. And he's an expert in designing thermodynamic cycles and rolling out large thermal and electrical thermal systems as scale. And he's been working on several hundreds of megawatts projects with net power, looking at supercritical CO2 cycles. Adrian and Brock were very much looking forward to the systems level discussion, which is a very crucial one for thermal storage. The floor is yours. Thank you for the intro. We're going to switch gears quite a bit here and go for something that is maybe quite a bit different from what is typically seen on this lecture series. What myself and Brock will be talking about are kind of the nuts and bolts practical challenges of taking something from the lab and building out a first thing in the in the field on a real customer site commissioning turning it on for the first time and operating it. You know, as, as mentioned previously, I saw a lot of technology ideas come across my desk at RPE. Since then, I've been involved with a few different startups directly and indirectly. And what I'm trying to convey here is my generalized observations of the early stage. Technology industry and basically convey those learnings to you so that if you are working on an early stage technology, if you're doing a startup, or even if you're at a slightly later stage where you're just about ready to turn your plant on. Maybe there will be some things here that will help guide you and push you in the right direction so that you're most likely to succeed in actually bringing your storage technology to market. So, in terms of what we'll be talking about here, we'll be not just talking about thermal energy storage. We won't even just be talking about energy storage. I'll be talking about anything that is basically an industrial scale facility that needs to be built and operated for the first time. So, I'll start off here with this image I found this is I found this at a hydro dam built by BC hydro out in British Columbia, and I was quite charmed by this image, but I think it's really perfect to explain kind of the scenario that we're in right now. You can see the little baby at the bottom which is electricity storage of force. Of course, this was in reference to the hydro dam, a different sort of energy storage. But we're looking at various other types as well. For example, thermal energy storage which can be converted to electricity. And so, one of the scenario we're in right now is we have a lot of baby ideas and a lot of baby plants for thermal energy storage or for energy storage. And we see the big giants on the fossil side, or on the process industry side who are kind of looking at us and wondering what's going to happen to that kind of new nascent technology and thermal energy storage. So, in terms of a little bit of the background of the talk here and kind of my background on where I'm coming from having, you know, learned various lessons in the thermal energy storage space and trying to convey them here. And Brock, if you just want to say a few words to introduce yourself really quick. Go ahead, and then I can start jumping into the meat of our presentation. Yeah, so thank you very much Adrian. So, I started working in the energy sector about close to a decade ago, focused predominantly on super critical CO2 power generation. And when I started in this field I was dealing with something that was little more than we'll say heat and mass balance that was not much better than chicken scratch on the back of an Afghan. And I worked, you know, very intensely with the chief engineer of the company I was working with Colbate Rivers at that time to take something from quote unquote chicken scratch all the way through to you know, it's really worth more than $100 million that's now an industrial scale experimental lab test facility for moving on to the next generation of clean fossil based power. And so, I've got to live the life cycle of, you know, you have this great idea, you got to convince people to give you money. You got to build it you got to design it you have to deal with will say the hiccups that come along the way, and you have to pivot in terms of your commercial commercial future. And I'll be talking a little bit about that the kind of the growing pains of brilliant idea to how do you actually get the thing to a point where you have an impact in the world. So, thank you very much Adrian. Thanks Brock, and we'll explain this a little bit more later but Brock's company net power is not strictly an energy storage system. However, it does have a lot of the same challenges in technology development and build a first of a kind of highly relevant to any large scale thermal energy storage facility. So, diving into it. Okay, great. So, okay, let's get serious about solving climate change. What exactly do we have to do to get to our goals of limiting global warming. So, I took a look at the IPCC report one from 2018 and you know they had a lot of data in there in terms of how much renewable electricity generation capacity we need to install to reach out to me. And so, by 2050 they're predicting under one, I'll take a sample scenario here by 2050 need to install nine terawatts of renewable electricity generation. So, I looked back at the envelope calculations and I was like, how much do we actually have to build to get to that level of, of like hardware installation. So if you take the three gorgeous dam in China is the largest hydro facility in the world 22.5 gigawatts. It took 18 years to build. We would have to waste all renewable, or if we did all this renewable capacity, just with hydro dams of this scale, we would have to build one to two of them a month every month for next 30 years. Similarly, if we did this with the world's largest nuclear plants it's a gigawatts in Japan, we'd have to build one of those 10 days, or sorry one of those every 10 days for the next 30 years. We would have to feel going in the ground and going in a lot faster than has ever been done before. Okay, let's take that a step further, we know we can't just put a ton of renewable energy generation. We also need storage. So let's say we have 50% renewable penetration, we may or may not need more than that depending on what energy mix we choose. But if we take the Tesla 100 megawatt lithium-ion battery plant that took 100 days to build, we need one to two of those a month every month for the next 30 years and that's just for the US. And then if we're doing carbon capture and sequestration, there's a whole bunch of other stuff we need to build there as well. So basically what I'm trying to convey here is that we need to build things faster. And we're running into a lot of roadblocks in that process of building faster. And so there are a lot of great ideas out there, including the one that Ravi just presented. And it's kind of like, okay, great. I want to use this. I want to build it. I want to get it out to customers. But again, this is incredibly slow. Like, why? Let's break this down. Why is it so hard to actually build these things? So I can give you a whole laundry list of reasons here. I'm not going to go through every individual thing on the list here. If you're interested, there's a book by Bill Bonvillian. He's, he's a player in the US manufacturing space, but he does a lot of thought leadership in terms of innovation when he calls legacy sectors. And I think it's just kind of funny that the cover of this book is, you know, a guy rolling a giant stone up the hill that's inevitably going to fall back to town. I feel like that sometimes. Anyways, he, he puts a lot of thought into like specifically what are the things that are stopping us things like perverse subsidies codes and standards are difficult for regular storage standpoint, little focus on implementation. But I'm going to zoom in on a few specific things that I've seen and try to explain them in a, in a way that I hope is a bit more digestible. So let's say theme number one is basically, there's a bottleneck to how many of these novel or early stage technologies, how many of them are able to get through that bottleneck and actually become operating plants. And if I were to generalize, I would say that the two main things limiting us, or that are creating this bottleneck are basically limited options for mobilizing capital, getting as the getting the quantity of money, we need, and getting it fast enough, and having patient capital that is willing to tolerate longer timelines, I'll go into a little bit more detail in a second. The other one is actually the construction process, very difficult to construct, especially first of the kind plants fast. Typically there are a lot of delays and complications as you're building something out for the first time. And it's also kind of an unpredictable murky process like how, as an early stage company or team, how do you correctly interface with the EPCs of the world the vendors and contractors as suppliers of the world, and bring them along in your process such that you have a smoother construction process and higher chance of success. And so I'll go into that specific one and more detail with that. But let's start with the limited options for mobilizing capital. So, if we look at kind of the, the funding map here. You know, you hear of things like the value of death, you know, trying to get over that value of death, typically sometimes is referred to as manufacturing value of death I'm talking more about like an implementation build out type of value of death. And really what's happening is that there's this kind of gap here that I've highlighted large scale demonstrations where there aren't a lot of places to get funds for these sorts of projects. No one, or I shouldn't say no one. It is especially challenging to get funding for your first of a kind plant. It is less challenging to get funding for your second third, fourth kind of a plant. One of the challenges here is that capital is typically, or rather the timelines of build construction and later build and later construction of the second third, fourth of a kind. Those timelines are just typically a bit too slow for the return rates that you typically see yet. Let's say your standard venture capital firm or your standard private investor, etc. So that's one issue. Another issue is that going to the construction issue, trying to break down like why exactly is construction so slow and challenging. And so a trend that I've noticed is that, even for nth of a kind plants, things that we know we've built a million times before. So let's say a natural gas turbine power plants, or something like that a mining facility, things that are big and heavy hardware sort of facilities but you know we've done it before. This here was a study from McKinsey where they basically looked through a database of timeline and spend for a range of these different kind of known assets. And basically, the conclusion that they came to was that 90% of these known projects incurred cost overruns, and on average they were slipping more than 20 months behind schedule. And so this is kind of like a systematic issue in the construction industry. You know, building fast, even when you know what you're doing is really hard. And so, okay, that's for nth of a kind plants that's for things that we know. First of a kind plants are especially difficult because in turn, you know, you have those kind of initial challenges of the construction industry being slow. And then on top of that you have these kind of other issues. With the first of a kind plants in, let's say we're doing thermal energy storage or energy storage for load shifting of electricity. Say that for now, even if you're load shifting heat or something like that. You're in the end competing on price. It's hard to differentiate your electron from the electron of your competitor, or your phone on from the phone on of your competitor. And so that makes it difficult for customers to be more selective and say like yes I want your product. The other main issue in the middle here is that as a small startup company you don't have a very big balance sheet. And so it's very, you can say as much as you want you can claim as much as you want as an early startup company about your metrics. But in the end, you don't have any capital kind of backup or provide warranty. And that makes people and that makes customers nervous. They're worried that you're either going to break their grid. And so you spend a lot of time getting a grid interconnection. Break their grid, or they're concerned that you're going to break yourself. And then they're going to lose all the capital investment that they made in you. And finally at the end there, the last kind of big challenge that's unique to first of the kind plants is, well, and second of the kind answer and honestly, is a very high risk financial environment. People want to know exactly how much money you're going to make. Well guess what the markets for energy storage of various durations aren't mature yet. And on top of that, if you're looking at pulling multiple revenue streams for your energy storage asset, those aren't priced yet. So it's very difficult to say conclusively how much money you're going to make your customer. So. Okay, yeah, and so if you kind of end up having this sort of challenge when you're in a energy storage startup it's like, I want to build a thing. I don't have the money all the money I want so I have to build a simplified version of it. But how do I build something that's kind of good enough, how do I build the skateboard so that I can live long enough to build the scooter, build the bike to build the motorcycle. You don't want to build a wheel to start off with that's completely useless and then just have grumpy customers that these interests. So you're balancing all these different things the juggling act and it's you're trying to make smart decisions here. You're trying to build out your tech. Okay, so I've done a lot of complaining. I pointed out a lot of issues, and I will make an attempt here to make some recommendations on how we can progress here as academics as non academics. When, when we're trying to really get serious about putting steel on the ground and moving the needle. And yeah, my hint here is basically, it's basically all in how you choose what to work on. So, let me explain that in a bit more detail. I've seen that often companies are optimizing for the wrong metric. When you talk to customers, especially in the large scale, thermal energy storage or energy storage space. Typically, the desire is to optimize first for reliability and safety. These are the things that get your foot in the door and convince them that you should operate on their land for the first time. They're often less concerned about efficiency and costs. That's very important. But that is not necessarily going to be the thing that you have to prove for them to approve your building on their land. The other way to think about this is typically we think about technology development from the bottom up start with the fundamental research and then go to the through the stages to make it more mature. I would argue, or rather my opinion is we should start thinking about doing this in the opposite way. And this is a lot of kind of customer outreach stuff, understanding your markets, things like that. But there's also a technical element to this. So if you go to your customer and you talk to them, they say, I want a system that has an uptime metric or reliability metric of 99% uptime. You might design that thermal energy storage system in a very different way with lots of redundancy, for example, and then your requirements trickle down, and then you say, okay, this fundamental research area makes sense, because it addresses all of these kind of customer customer requirements from day one. And then I would say, when you're choosing which technology to work on or when you're refining the one that you're already working on. So in terms of again just thinking about optimizing for efficiency or cost, there's a lot of other kind of product or system features that are actually incredibly important later on when you're trying to commercialize your technology. You know, I've listed a bunch here, the more you can satisfy with your technology the better. So to emphasize one or two here, for example operational flexibility, what's really unique to energy storage is that inherently it has to load follow it has the charge has the discharge. So looking at full load partial load. This is a very transient thermal system, and we are not. We're kind of in this new, let's say area of optimizing thermal systems to operate in transient ways. Previous industries have always bent over backwards themselves to always operate at steady state. And so this is like a fundamentally different way of thinking about thermal systems for thermal cycles. And this is necessary. If you want to target more markets as a thermal energy storage asset. To highlight is use existing ecosystems. This means things like supply chain, where's your labor going to come from. Can you use trained welders as they are or do you have to completely retrain them and something else. Think about how, like, who's actually going to build this and how and where are the parts and the bits and pieces going to come from. Do you have to reinvent the manufacturing wheel to make your thermal energy storage technology work, or can you use the existing manufacturing options out there. And then, yeah, embrace system mindset. So, you know, I've become a thermal systems engineer used to be at the component level but now I'm at the system level and often what happens is there's a lot of focus on let's say what I call the core So let's say if you're working in hydrogen you focus really a lot on the stack. But what ends up happening is that so much attention is put on the stack that you forget about the balance of plants or the balance of system. You forget about the purification steps you forget about the step afterwards for example there's a lot of other stuff that has to go into your product to make it work. And then often when you add these things and after the fact you find that your system costs. I don't know 10x more than you originally thought. And so the, the kind of elephant in the room as I say is that, you know, when you're evaluating a new technology like have you thought about the full system yet and what the cost implications and what the real operability implications are going to be. Okay, and so that's essentially the things I had to say again my generalizations from having seen the progress of many different startups in the energy space. So let's let's hear from Brock. Again, he doesn't that power is not an energy storage technology not thermal energy storage technology but does have a deal with a lot of these similar challenges of building out big industrial assets. So he'll present a little bit and explain what that power is all about, and kind of work in some of less of his lessons learned, since he's actually had a lot of boots on the ground at these facilities. So, Brock, I'll hand it off to you just let me know when you want me to switch slides. Yeah, well let's just move on to the next slide. Thank you everybody for for being here. Kind of just to give you a little bit of background why that power has relevance to this conversation. When you talk about like grid level thermal energy storage, you're still talking about EPC level concrete steel piping systems control systems, you know, training operators heat exchangers. You know whether or not you're going will say solar thermal to a steam plant or you're doing pebble bed heaters. You're still kind of like grappling with the same issues of how do I take something the industry's not used to building right now they've got the raw skill set they've got the hardware capabilities but you're trying to teach them how to build a new mouse trap. And that's kind of will say some of the lessons learned we've got with net power. You know this slide here is just kind of showing the credentials of the people who who really built net power it's a diverse group of people. As Adrian said earlier on the CTO of, or as I said earlier on the CTO of net power and working in power infrastructure for the last 10 years predominantly super critical co2. And I can tell you it's been a fun ride but it's not easy. It's something that really requires you to dig in and go from academic to the nuance of, you know, usher or taking care of your baby and showing people actually how to do something right from the process design standpoint and learning from them. So if we can go to the next slide Adrian please. So this is an example of the test facility that we built. I started working on this about nine and a half years ago. It's an industrial scale plant. So, like I said, we went from, you know, basic aspen plus heat and mass balance that you would ask, you know, a third year undergrad in chemical engineering design to building something that was well north of $100 million with the investment from strategic. So there's clearly, you know, been a lot of information we've garnered in terms of what it takes to go from the brilliant idea of Jeremy fed fed and Rodney Allen to planted actually is functioning and working and until about midnight last night I was actually at that facility so talk a little bit more about that later, but let's go on to the next slide. So, how is this relevant. Okay, so we saw that we needed a clean future. We saw that we needed a clean future now. The inventors of the power technology understood that the infrastructure around fossil fuels is ready. It may not be the answer everybody wants but how can we make it clear. So we started from the ground up redesigned an entire power system, assuming that thermal energy could be sourced from existing infrastructure and we could capture all the carbon. We will basically wanted to create low cost electricity, the fully integrated carbon capture, and that wasn't really possible with existing infrastructure. So we were dealing with a new mousetrap that really never been seen before. So we started talking about super critical CO2, you know, high pressure carbon dioxide, you know, 30 megapastals your north of 1100 C, you're dealing with chemistry that are typically in the oil and gas sector, not the power sector. You really are forcing people who build power plants to look at things in a new way. In reality you're hybridizing skill sets, and a lot of that's going to be comparable to grid level thermal energy storage because you're going to be asking people to build things that have actually been done at the pilot scale, or even will say smaller than that, you know, desktop scale and turn it into something that actually has will say bankability insurability safety standards, etc. Those are things you don't think about necessarily when you first start developing a project like, you know, do I have to think about installation on a piece of pipe, how does it affect the operations doesn't mess with my process design, and why am I doing this. And so I'll get a little bit into that later on but you know we started off we wanted to make clean energy exploiting existing infrastructure so we could move as quickly as possible, while allowing electricity prices to remain low cost so the world was not penalized and there was no business case against going clean. So next slide please. I mean, this kind of talks about, you know, the big will say mission of what we were trying to do. But the one thing I want to emphasize is that when we talk to customers and to Adrian's point earlier, you know kind of looking at like the end stage and then reverse engineering, how you should be developing the concept became critically efficient. One of the best things we learned was the capacity factor was probably the most important thing you should be looking at capacity factor in the face of levelized costs of electricity and this is going to be true for thermal energy storage. You can design the most efficient widget out there. And today it's going to be regulated utilities that have to serve, you know, hundreds of thousands of not millions of people. Electricity at rates that are dependable and output that are dependable that are really going to have the final say, you know, they don't mind it being a little bit more expensive because it's a little bit less efficient. If they get an availability and capacity factor that guarantees the lights never flicker. And so that was something you know we learned kind of will say mid stage was that we learned less about efficiency and more about reliability. And that's kind of something that was really hard to grasp my first started on this in this on this path was that you want to create the highest performing system. But that is not what turns into dollars at the end of the day. You know, think about utilities get penalized if for some reason they can't actually supply because they have these guaranteed contracts. It's a thermal energy storage system that only has a capacity factor of 70%, but when it's working it's you know, 90% efficient. That's great, but they're going to want you to be 8590% capacity factor and they'll take the decayed performance. And that's kind of something that we learned like so these mission statements were really what motivated us to, you know, build the technology, but on the back end we said, Okay, that's great. So what do you know the Dukes the TV is what do you know the smudge what do you know, Pacific gas and oil like what do they actually need, they need to make sure they have reliable power 24 seven. And this has always been kind of one of the biggest criticisms of renewables and this is actually why we find that, at least our technology has a home with renewables because we can provide the capacity factor that currently doesn't exist at a grid level storage. And hopefully we will get there in terms of alternative storage capabilities, but level life cost of electricity is king. And even in a world where there are carbon taxes that are forcing people to go lower CO2 emissions or no CO2 emissions, that still factors into the concept of level life cost of electricity. So you have to look at all the factors it's capital costs it's capacity factor it's tax regimes it's tax benefits and so that was one of the big lessons learned we had was, to create a business case it can't be what you want in terms of optimal performance as an engineer or scientists. It can't be what you want as an idealist in terms of climate change. You have to actually look at who's going to buy and build this now and today. And I told Adrian in a separate message, climate change requires that we have infrastructure built now, not tomorrow, not, you know, 100 years from now, we're not, we're not waiting for a prayer. To bring us, you know, the holy grail of something. We have to look at what we have in hand as far as resources, and as far as resources with respect to research and resources with respect to EPC. Next slide please. So I'll just kind of go through this real quickly because it's not really relevant to, you know what we're talking about today. So we developed a super critical CO2 cycle, we combust oxygen. In the presence of methane and recycled CO2 at 30 megapascals, anywhere between 900 and 1200 C, we expand through a turbine go down to about 30 to 40, three to four megapascals do recuperative heat transfer. Go to a water separation stage compress and pump the CO2 peel off a portion of the CO2 for sequestration or utilization. And then the rest goes back through the cycle. So it's an oxy fuel combustion system. It's basically like a high pressure steam cycle, but uses transcritical CO2 as the working fluid. And what this system allowed us to do was to generate electricity at a levelized cost, equal to that of CCGT today, without capture, while still allowing for sequestration utilization of CO2. So this was the value proposition of what we're working on. But the real point here is we're talking about, you know, air separation units, we're talking about advanced turbines, we're talking about heat exchangers, compressors, pumps, water separators, high pressure piping, high temperature piping. These are all kind of the buckets that any thermal energy storage system at grid scales can have to deal with. You're going to be talking to the same contractors that we do. In terms of being able to deploy these units and you need to understand how competent they are, how quickly they can move, what the supply chain looks like, etc. Because ironically enough, a lot of people don't realize when you're building projects like these, you're not necessarily in a vacuum. So when I want to buy my heat exchangers, I have to worry about floating offshore platforms, because they buy the same hardware. And so, going from lab to actually building something, there's this massive supply chain learning that has to occur. You really have to understand, you know, has China had an uptick in terms of stainless steel consumption. And so now everything across the board is more expensive because nickel prices went up. And so those factors, the things like levelized cost of electricity when you start looking at thermal energy storage, are using, you know, potential catalysts that are highly expensive. This kind of factors back into my efficiency doesn't matter as much as cost, confidence and cost of deployment, reliability, and then ultimately capacity factor, maybe a PebbleBed reactor is not as efficient as a molten salt storage system. But what are potassium prices? You know, where is NAC today in terms of cost? And so that's kind of what we learned in this system is that when I want to buy an air separation unit, I'm fighting against the steel sector. Like, is there an uptick in steel production, basic oxygen steel making in India and China, because that's where the largest growth happens to be. You know, is there a large scale methanol plant going online in Iran, and have they basically booked up the next two years of production. So these are kind of some of the hard lessons learned of like you've got a great idea, you've got all these components, but other people are going to be competing with you and if they've got basically faster timelines and better economics, you're going to be waiting. Next slide please. So we'll just keep going. We're going to kind of dive into some of the lessons learned. So one of the biggest and hardest lessons learned I think the United States acquired within the late 2000s. We had this giant clean tech boom, you know, VCs particularly in the West Coast, Northeast United States were pouring in tons of money. The problem is the timelines for returns were too short. The experience in terms of management was not necessarily there. And a lot of the technologies we're talking about when it comes to thermal energy stores or mine, you require strategic investors they're going to take five or 10 years to actually usher in the project. When you get back to kind of talking about supply chain, you want to order something like an air separation unit you want to order something like an advanced heat exchanger, you're talking 18 months to two years before it shows up on site. And that's after maybe a year worth of detailed engineering. That's after you got your funding. That's after you actually had a competent design to even show to an investor who use will say someone like Adrian or myself to critique your technology. So you're talking right off the bat, no matter what you're doing. You're doing something that requires as me code approval insurance operators long lead items, five years at a minimum to really build any of these projects. You really can't do it any faster than that. And that means traditional capital routes that you know have we see in the tech sector are not going to be acceptable for these kinds of projects. And that's not to say there aren't investors out there who are leading the way, but as a general rule of thumb, you're really going to have to start looking at people who are in it for the long haul and in it for big bucks. And that means looking at strategic for example like we had, you know, eight rivers I was part of we invented the technology. So that was self funded in terms of getting the, you know, the concept together. But then we started to look at like, all right, at the end of the day, what does it take to build this. We need somebody understood carbon dioxide, we need somebody who understood, you know, running plants, we need somebody somebody who understood building plants. And we had to fit into their strategic initiatives in terms of why they want to participate in technology like this. Now, big strategic energy companies do five to 10 year planning so you know they're looking at the writing on the wall in terms of carbon emissions etc. They're really kind of what we, I wouldn't say exploited but we went after like these firms all saw that the future of their business five to 10 years out require they had to have low carbon technologies in their portfolio. And so they're willing to put a lot of money on the table to develop a technology like ours, because they know that 10 years from now, they already have to have had that project in the planning cycle, because these types of project takes five years to build. It doesn't mean the money's easy to come by you, the level of due diligence if you're asking somebody for 50, 50 plus million dollar check. It's going to be a lot more than series a B or C, especially when you're talking about technologies that have to have reliability, high capacity factor and have human safety as a concern. So, the upfront lift we had a date rivers was a lot larger than you would find with say, than like a technology app, because there's no such thing as agile engineering when it comes to building something that that pressures and temperatures, they could could physically harm someone, or damage companies reputation because they couldn't get electricity back on the grid for a couple weeks, because they will suffer probably millions of dollars of penalties that are actually a large chunk of the investment they're going to make you in you in the first place. So, to go from lab to industrial prototype is in this field is going to require a lot more work and I think people are used to. You're going to have to get down to the level of actually developing preliminary P&IDs equipment lists, equipment data sheets, you're going to have to talk to people about transient operation as Adrian said earlier. I want to know that at least you started to think about the ecosystem they live in, and you've got the ball rolling. And that's generally enough for them to feel that their experts should start working with you to actually flush it out and build out the technology. They get to a point where they do their own diligence maybe six months to a year later and they're like, we're good to go. It makes sense. Let's invest. Let's keep building this. I'll keep beating this dead horse. You know, once you get the money, it just starts it's now five years from that point before anything ever gets built. So next slide please. So we kind of chat a little bit through this. Building industrial technologies is not faster. There's no way you can really accelerate this other than, you know, paying at least supply chain, you know, paying them large sums of money, it's basically push somebody to the back of the queue, and even that only guarantees you three to six months of building a plant. You have to go through, you know, feasibility studies you have to go to, you know, pre front end engineering and design, then front and engineering and design in those phases alone probably take a year and a half two years at best construction anywhere and a year to two years depending on the complexity of your technology, then you're going to go through commissioning and startup that's probably three months, and you've got a hand over and we're talking to best case scenario post investment post you actually getting your ducks in a row to get someone interested in pushing the hardware. So, you know, that's kind of this timeline shows that, you know, we started with the idea in 2020. We really only started doing the detailed engineering, just in 2013 2012 when I joined the company. And then, right now we're moving to, you know, the commercialization phase. When you when you talk about technologies are going to solve climate change you kind of have to get real about what it takes to build them. What does the end customer want and reverse engineer whatever widget or concept you have to kind of fit the current role of execution. It's not ideal it's not perfect. I can tell you I personally been frustrated a million times as to like why can't you understand this is potentially a better way to do it. But you're kind of, you know, yelling at a wall, simply because you have to deal with the talent and the resources that currently exist. That's not to say don't be shrewd and don't be picky about the people you work with. There are suppliers that, you know, are supporting your technology, but that's also going to take time to I personally got on a will say global trip to visit every printed circuit heat exchanger manufacturer in the world to look at who actually had the furnace capacity to support our technology and that's not something we ever thought about when we first started off is like, I'm in South Korea, I'm in China, I'm in Sacramento, I'm in the UK, like literally flying across the world to figure this out. I'm in a detailed process engineering which I was actually used to. And so it's it's it's those little things I would say little things it's those kinds of things that really make the ideation, I think ideation is maybe 1020% of a successful project. It's, it's the, it's the grind of figuring out actually how to do it. It's like many of the people on this call are probably smart enough to know how to build a house. But you got to go find who's going to sell you the wood one is it showing up, you know what kind of nails you're going to use are using galvanized nails not etc etc etc it's like so it's not about having a good idea it's not about being smart enough it's about really figuring out how to put the record set together in detail. Next slide. Okay, so I mean this is just kind of to show you that, you know, we, what we put together, you know, this is quote unquote a prototype but it's an industrial scale facility that was built to all applicable codes that an EPC firm would require. And the reason we did this was Adrian was saying you know you go from skateboard to scooter etc. This is effectively the skateboard. You can see how this goes from 50 megawatt scale to something that's maybe 500 million to a billion dollars in scale but there's a linear understanding with respect to their engineers very PC firms, you know their bankers etc. So you have to speak their language in terms of what you demonstrate. Otherwise, they're going to ask you to do it over again. So this, this plant, a lot of people come here I'm like wow this is a prototype it's like, not really it's a test facility that was built to all applicable as me and API standards. And so, yeah you spend more money you spend more time, but it saves you on the back end with respect to their being a linear understanding as to point a to point Z. So this is kind of where it gets fun for those who are into process and systems design so it takes you five to 10 years to build you can't stop inventing. So anybody who's basically going to invest you know hundreds of millions of dollars as a strategic investor because you're solving a problem of theirs they see in the future and they're going to be your first customers. They also want to know there's longevity to what they're investing in. So, that's the other part of this kind of Pandora's box is that you, you, you're doing all this other work, but at the same time you have to keep inventing you have to keep building optimization new pads etc. Because you want to give them a reason to find that this is something that has 1020 2030 years of life for the money that they're going to put in. And so, that's how you convince one of these larger strategic that makes sense is that you're solving their problem initially and you're thinking about the problems they haven't even thought of yet. And that's, that's kind of the value add for those, you know, who are our thinkers and future thought leaders of the energy sectors that don't stop inventing don't stop filing. So, as you move through the process of trying to go from ideation to execution you're going to learn a lot and write everything down, make sure you integrated into new patents, new will say copyrights new proprietary learnings. Because that is how you make sure your technology actually continues to have an impact. So on that, I am done. I'm going to move on to the next slide and leave it to questions. Thanks, Brock. And there's something I want to emphasize here on this slide is Brock this is, you have listed here 50 megawatt thermal that's what this is and that's 111 of the full scale. Some people in the audience might be thinking, you know, oh, you know my technology is more modular, you know, maybe it's one megawatt, maybe it's half a megawatt. I mean, my technology is something that's going to be in the home so I don't have to deal with these kinds of construction challenges. I would argue that even in those scenarios, you're still looking at a situation where you're going to need a factory somewhere that's going to be popping these things out like hotcakes if you're really looking for the impact you want. So that, so the construction and supply chain challenges than a lot of the other things we're running into here, they're just transferred to that manufacturing. Yeah, different different players same problem. Different scale but you're, you know, I've got a friend who makes jewelry and uses supply chain out of Malaysia and China and she has the same issues that I do you need to figure out lead times. You know, capacity, you know, lowest cost bitter quality control. It doesn't change. Cool. So, just one last slide here I'm going to leave homework for the audience. If you want to read a few books, or some things about kind of the importance of having both good technology but also good technology management. I'm going to be reading these things here. I found especially making the atomic bomb to be quite inspirational inspirational in the sense of the amazing technology development that's possible when you have good leadership. And so I'll just leave that up and we can go to questions. Thank you very much for the presentation. It's wonderful to see the connection between the two talks of very early stage technology development and deployment at scale. So we have a number of question but I thought I would just start with my own. If you can indulge me, you know, being a materials person I always try to draw analogy to system level challenges versus something smaller. And the example that I came up with as a semiconductor industry. So semester in the semiconductor industry, we can very effectively develop the device right based on known loss of physics and design rules. And the unknown there is the cost learning curve and the yield. So that's very difficult to predict ahead of time you can have a perfectly working device but it's just very low yield to produce. And the purpose of the piloting the mass production is to learn the cost curve and then to learn and improve the yield, and that ultimately determines the success of the product. In larger energy systems like thermal energy storage technologies. What is the thing you're trying to learn through the piloting. What is it that you can readily simulate and design on paper with very high fidelity. What is preventing us from getting a plus minus 10% estimation today at the design level and not be surprised by significant cost overrun in the actual piloting the process. In other words, you know why are we seeing technology with attributes and metric that's often over promised and under deliver rather than the other way around. So, is this a technical problem or this is something else love to hear your thoughts on this. I mean, you want to go first Adrian I can rant about this. Let's look at IGCC. Well established technology understood huge players in terms of who was supporting the hardware. But every single plant that was built was effectively treated as first of the kind. What was the issue here project management. Clearly you've got your issues when it comes to you know fuel stock variability you got to deal with you know the chemistry concerns when it came to gasification. I think a lot of these industrial technologies kind of fall into the void of people in silos just business as usual, and when you start asking them to do something that they're not accustomed to. They just apply the same metrics, and you end up with something that looks like Frankenstein's monster as opposed to the beautiful vision you once had, you know when you started off the project. So, it's, it's really being somewhat of a helicopter parent, and making sure you oversee that entire process from ideation to execution, such that you're talking to the people who are selecting the materials for example. You know, there are no shortage of materials I could use that come from, we'll say the turbine space and have the allowable stress profile that are required, but they're not as me code approved. They're going to have gone through 100,000 hours of process for stress analysis, in order to be insurable. So that means that I asked somebody who's used to either the power sector, or the chemical sector, they're going to use different as me codes, which means they'll select different materials. And this has nothing to do with whether or not they're suitable for the process, but it has a massive impact on costs and timing. So, that if you're not paying attention to the people who are building your system and what they're doing to build your system. It's going to get lost in the weeds and that's why I use IGCC as an example that's a technology that when coal prices were lower than gas prices. There's no reason it shouldn't have proliferated. There's no cost and cost overruns. Many brilliant people involved, but you have to have a maestro who's basically overseeing it at all times, if the sector, if the workforce is not used to building what you're asking to do, they're going to do things the way they're used to it and that often leads to inefficiencies. Let me answer the question with a different grant in a different direction. Let's go to the material science example. Just in their issues of, and, you know, I'll kind of roll my eyes when I say this unknown unknowns. So let's say we're trying to figure out if a given metal alloy corrodes in a certain environment, so that, you know, our parts don't fail after one year of use. So you think that adding the working fluid to that specific alloy in a specific materials test is enough to determine the lifetime of that piece. You might thermally cycle it. You might put it under stress or stress or tension to figure this out. And when you put it in the real plant, you find that these parts are failing anyways, even though you ran all of these tests. And for example, some of the unknown unknowns there might be. Oh, when we come when we built this system, these pipes were exposed to humid air with, you know, water content in it. And these water molecules created sensitive regions in the metal lattice that made them susceptible to corrosion at a later time. So it's, it's, there's a lot of really nuanced things to how you select your metal alloy. What do you do you have to keep things dry and under vacuum as you're building the plants when you turn it on for the first time are they're going to be oils from things upstream or downstream that are going to contaminate this thing temporarily and sensitize it. There's just so many different things to take into account. I think kind of where where you were going with your question is like kind of the digital twin question. Can we make a good enough digital twin of a plant. If we had, if we understood the physics well enough, like, can we do an Aspen model on steroids or something like that and really account for everything. In theory you could. But to be honest, I don't know if I would be able to predict all the important physics. I don't know if I would be able to know all the unknown unknowns, especially in the first of the kind plant. And it's kind of like, you know, when you're running experiments in academia it's like why do you run the experiment instead of just modeling it, because when you run the experiment you learn things. Bad stuff happens. And sometimes good stuff happens. But that's like a critical part to the learning and design process is actually making it run. So, so digital twins are very useful. And they get you, I don't know, 80% of the way, something like that, but you still need something a little bit more beyond that. Thank you, Adrian and Broggett. I've really resonated with that point as an experimentalist. You're always surprised by what you see good or bad. I have a follow up question on this and it will bring E and Rafi back for a discussion in the panel. Adrian, if I think about this digital twin concept right agree that you can only go so far but it can certainly help. And I just want to also point out in the battery. Long duration storage area we're facing a very similar challenge which is, you know, Brock spoke about insurability, bankability. The time of long duration storage isn't really proven and even accelerated evaluation is a problem right so can do we really know if, you know, even if lithium ions going to last 25 years, which is, say, that's what is needed to make financial sense. We will know until the real test comes back in, and even the acceleration itself is questionable. I know you're describing sounds like a related problem but even a larger scale. And, you know, your example of, you know, humid attack humidity attack on metals is a very good one so you don't know what you don't know until you try it. So I don't know if Adrian and Brock you can come a little bit on your thought, you know, for systems are supposed to last 25 and 50 years. How do we accelerate that process. So we can make bigger bets today rather than 25 years later until the first batch of plants have gone through it and, and I just want to point out consumer electronics is really being at the forefront of this is that's what really enabled the battery revolution for electric vehicles is it brought in that three to five year demonstration that was needed and then gave people confidence that battery can be safe and then that went on to, you know, 10 years for EVs and that will go on to enable 25 years increased storage. I mean, I think you kind of hit the nail on the head there you're talking about billions of dollars being effectively deployed for demonstration. That's kind of where you end up with these industrial scale technologies is not going to necessarily be billions of dollars but, you know, we do things in the oil and gas and energy sector like Ram analysis. And so you would do the same thing you look at failure points you would look at what are capital spares and you'd actually say that. Okay, I actually expect, you know, a heat exchanger to fail at least once in the lifetime of the plant so I'm going to carry, you know, an extra X millions of dollars amortized across the life of the facility and assume that's actually my life cost of electricity. And that's what I'm going to a banquet because you're going to expect that kind of risk premium to be integrated. And so if it doesn't fail or you get an extra rule say 510 years. So operating, you know, the op X was improved and the facility made more money, but you got to factor those things in from day one. If anyone can poke a hole in your technology and tell an investor, particularly a bank who's going to put in you will say hundreds of millions of dollars, they're going to shy away they want to know that at the very least, you know, your conservative return on investment is something that actually turns a profit. And that's kind of the cynical truth, you know that I've encountered is that you have to assume they're going to poke at everything that could go wrong. They're going to have their own engineers and scientists look at what you're doing. And if they can argue logically that you know it makes sense that you should have, you know, spares, you have to take that into consideration. So a quick follow up and then we'll move to the panel is there, you know, in terms of say, meantime before failures for all the components that would go to a large power plant or thermal storage plan. Do we already have very reliable numbers and you know is there a body that is estimating this empty bfs for new technology coming on new materials new chemistry. I'm guessing that if you know the empty df you can calculate the cost pretty readily, and that the cost of service in the system. The answer is yes and no, yes, it exists. Any of the large oil and gas companies any of the large EPC firms power firms, they have this data in house. It's stuff that they're used to doing a, you're not going to find it on the internet, you're going to actually have to talk to people who work in these sectors. It is annoying. It is hard. But it, you know, for example, a Ramco could tell you how often a fuel gas compressor fails a reciprocating fuel gas compressor fails. And what the capital spares are going to carry because given their environmental conditions, you know, within plus or minus two months, major overall has to be done. You know this, but when it comes to like new components that's where it gets hard. Nobody has the knowledge for example like my technology, we're developing, we're working with a novel turbine everything else is, is, you know, exists it's just being using a new way. And so you start having to do from like the ground up assessment risk assessments in terms of what is known how is it being put together in a novel way where the risk profile, and then they still add a margin on top of that. Adrian I defer to you on this because you've had a lot more work with novel hardware than I have. Yeah, I was actually going to answer this similarly, which was which is the say that as much as possible you have to pull it from what's already out there. You know, if you're designing a novel component, try as much as possible to pull from the past data and expertise building the same or similar component. Another potential way to get around this or to work with this is think about creative ways to make your tech more modular. The example that comes to mind is I think it's the Tesla Gigafactory that's actually modular factory. They brought it online in chunks. And so the very first one they bring online they start making parts they realize this, you know, things are breaking down not working how exactly how they like but then the next module that comes online. They're incorporating the learnings from the first one. And the third one is incorporating learnings from the second one, and then you know you kind of go from there and so if you're able to do that sort of thing you might be able to kind of knock out some of these longevity and reliability things a bit faster just because you have a technology iteration cycle that sets a bit faster. Thank you, Adrian. I can't help to also notice an analogy to the aviation industry, and that's where you know safety really matters. Iteration costs is very high the number of prototypes are very limited, say for passenger airplanes. So that's also something that I've been thinking a lot about what we can learn from the aviation industry. I see he has a question and then also I think Ravi we can have you also join as well we have about 23 minutes left, we can have a spirited discussion on all these topics, please. I'll just go ahead and ask my question or agent thank you so much I think I resonate very well your general common what's happening. Maybe this is this is more like a common sharing with you the experience now come back to the lithium ion batteries and this lifetime. The learning pace decades. 91 lithium ion batteries coming out you have what 150 cycles roughly 200 cycles that's about it. 2000 also. Now that's about 10 years later, right in the commercials paid place, people learn how to get to 500. That's about 500 2000 2005 range. Now 30 years fast forward I will just say we are roughly about maybe 3000 cycle except lithium ion phosphate, you know in the good cases, 10,000 cycle so 30 years. So let's look at the cycle life we learn. Well we from a two years lifetime calendar life, you know we, you know a year or two to now maybe about 10 years but this is still dependent on temperature. Agent I really like your comment about, well if you can leverage the learning from the past already accumulate in terms of reliability life, you want to utilize that. I want to resonate with that. And for brought I have a question for you I'm listening to your this new technology right in about decade long you achieve this progress is really amazing. Since you're starting in about 2012 to now. So, these CO2 generation burning oxygen used as a working food. So, eventually you're going to keep burning methane having CO2, I didn't quite catch right away just a lot of exciting information right, and then, and then you have a super critical one is working through eventually eventually you're going to have a lot of CO2 generated. Well, what do you what do you do with that CO2 again after you have that is injected under the ground that's the yeah. So, there's utilization either enhance oil recovery pure sequestration and saline aquifers or depleted oil and gas reservoirs. A lot of people don't realize there's 5500 miles of CO2 pipelines already in the United States that go into sequestration capable reservoirs so. Yeah, the goal is either use it for building materials value added products or sequestered but not in the atmosphere. Yeah, okay. So I want to make a comment on this whole idea of modularity and scale up and stuff okay. And I'll tell you I'll give me give let me give a story of a very good friend of mine who works in the Wall Street. They finally they control the money. Okay, he told me hey Ravi I have seen a lot of photovoltaics panels on lamppost and stuff I've never seen a big factory or or turbine based plants in my life. Right. So, so the people who control the money. Right. They see the batteries the CPVs right but large scale plants just not visible to them. Okay, you have to be much more sophisticated investor. That's one second is another I'm a mechanical engineer but the challenge with mechanical engineering machines is that look at a turbine. If you make a one kilowatt prototype that efficiency is going to be 5% and then if you pay 50 megawatt turbine that is going to be 50% okay. So as mechanical engineer researches all of us have advanced degrees and we can understand why I can model the heck out of it and I can show you 5% at five kilowatt, but it's going to be 50% when I build the whole 50 megawatt plant. I'm not buying that okay this I don't understand this. I, there's too much of risk in saying that from 5% to 50% these guys are telling me okay. That's another major reason that that large scale plants have financing challenges. So that I want to bring it back to research. If mechanical machines modularity can be achieved with similar efficiency high efficiency, but that's a fundamental that will lead to significant deployment acceleration and deployment. Right, that is by the way that is the one of the main is the CSP plant was a PV PV took over why, because a five watt PV and five megawatt PV almost will have the same efficiency you just stack them you add them in a you just multiply right there's no non linearity in the system. And some of the things which has a big impact on the financing of the products. And I think that long duration store the same thing will happen. You have a bunch of ideas right now based on turbines and mechanical machines. And I think again people will ask, show me at scale. Don't tell me that you can just do it on paper. And that's exactly why I said, if you're asking for these large amounts of money to take multiple years for deployment. You have to go to people who understand the problem you're trying to solve. It can't just be, as they say dumb money it has to be people who really like you're solving a pain, pain point for them, and they have the capital to actually to do something about it. And so it is hard, you can't just not everybody with a checkbook is going to be useful. And throw out a crazy idea that touches both on modularity as well as lifetime. What if we took a lesson from the consumer electronics industry. And we did planned obsolescence of components or systems I know Brock's going to roll his eyes and say no way you crazy PhD person. But I wonder if if there's something there. It may be better to know exactly when a component is going to fail, rather than than not knowing so if I can say in a predictable way that this component's going to last five years, and then I will replace it. It's probably better than like last seven years or seven eight nine 10 but it fails when you're not expecting it and your whole plant goes down and you lose a bunch of money. So, you know, is there a way to work that into our our thermal engineering system designs. That already exists. So I'm not rolling my eyes and actually agreeing with you. It's the LTS a model in the energy sector it's the whole idea that you reduce cap X, you push it on OpEx it really depends on who owns the plant who's building it who's financing it. For example, there's no issue with like gas turbines, having hot gas path components wore out on a regular basis, if the initial capital cost to build the plant is X. And the reason they, you know, that's something that's tolerables. Nobody knows what power prices going to be when they first build the plant, you know, 10 years from now, nobody knows what gas prices are going to be so the industrial sector does do this at scale. And that's based on what they call long term service agreements. And for example, a lot of some gas, some gas turbine vendors who shall not be named will actually sell you their gas turbines at a loss because they're intending to make up all their money on the service agreements. So, yeah, no, that's, that's right on the money in terms of thinking about your technology you just have to inform investors that you expect there's planned failures. And that is, I will not name company either but there are people that fundamental reliability challenges has not been sought. Then you come up with a modular swapable ideas. Okay. And then the thing fails, you just come in swap it with another one. Right. But fundamentally you have not made it highly reliable it is the same technology everybody knows there's a lot of reliability issues so that there are those models which has evolved. That's what I thought. Let's take our nuclear fission fleet. Getting quite old build a long time ago, I think they were built to have what 2030 years of life. Now they're being extended. A lot of them are being extended for another 1020 years. I might have those specific numbers incorrect but the idea. What are they doing. How are they modifying their plants to extend them by 20 years. Is there something we can learn there and apply them to our first of the client plants. I don't know what the answer is, but maybe it's worth talking to them. It's kind of a, well, I'll use the coal plants in this country as an example. It's kind of a myth when you say coal plants built for 20 to 30 years but she, but they're running 50 years later. There's nothing there is 50 years old, you know you're constantly recycling retrofitting components are being replaced, maybe the reactors are not being replaced but you're, it's like, I think there's a staff that the human body replaces its skeleton in terms of tissue 13 times over in the lifetime of the plant. It's the same thing you wouldn't say that you had 13 skeletons but realistically, the material recycle rates allows for that and that's kind of true of these large facilities. I mean that's how you've got their large oil and gas companies in Texas that have refineries that have been in operation for 100 years, nothing there is likely 100 years old besides maybe the original operator. So that's, that's, you know, that's the reality of these plants plan plan for plan for obsolescence plan for longevity. It just, you have you have to fit a financial model for whoever's going to be buying whatever you're selling. Let me have my two cents of this to I am really resonating with this discussion. And modularity I think has another big opportunity which is, it could be a proving ground for new proven technology so you don't have to build out the whole system. And again I pointed the consumer electronics as a very good example how he catalyzed the battery revolution is that every year we put in a slightly better battery as he pointed out and we keep learning keep iterating. In large system this is not possible but perhaps we can consider doing at a modular level, you know, maybe we try out a less proven system with some sort of a limited scope. I think, certainly in the entrepreneurial sense people want to build this whole entire new system that's completely revolutionary, or perhaps there's an opportunity for smaller scale development in conjunction with existing projects for you to pilot something out. So I don't know if this is something that is feasible or make sense, but just lower the entry to getting some field data as quickly as possible in order to do with the technology. I think we're seeing this very hugely in the battery field right you can now build a small component in a battery, and you can try it out, and you can see how it works and because that is where mechanical systems differ from a solid state system fundamentally the different. Okay, because mechanical systems you have a battery when you have one water battery or a mega water battery to large extent you can stack them together just multiply the number, right. And earlier, if you took a turbine you look at a chemical reactors whichever right you just small scale system is a much lower efficiency than a large scale system even the lab scale right when we do experiments in the lab. You see any paper, they'll say oh I did a small scale I have a lot of heat losses I here's a thermal model, and that at scale this will look like this right that's great PSD students and postdocs and professors can understand that. They will not understand that we'll say oh I'm not giving you millions of dollars based on a model that you telling me that it'll be really really good. So I think that's what I was saying right that if you can actually fundamentally solve that efficiency issue and mechanical systems and thermo mechanical thermal system at small scale, which is of course different research. That will move the system big time. Okay, I mean, but I have not. There are a lot of talks on this that how do I make it modular. Even chemical reactors, same thing, fish or drop you probably have heard right, oh fish or drop will solve everything I can have a scenic gas and I can kill it every kind of 30 billion dollars my friend one plant 30 billion dollars who's going to give you that money. So if you make a small scale fish or fish or drop system, man, I mean the yield is going to be very, really, really low. So, so Ravi, this is good point you are running a cyclotron row. So, what about what will be the mechanism and national lab and university we can do a little bit more. Instead of centimeter square, can we go to about meter square of kind of device right. We can go to, you know, no 5% somewhere in the middle not 50% efficient that cost too much but somewhere in the middle to prove out a little bit more what's the financing mechanism we can do this type of. I'm a Stanford right here and we try to change that by starting a new school and sustainability, the accelerator program this will be set up. We hope to be able to do some of those to build out the financing people's confidence of hey this is we are getting to next step now. So, any thoughts because you have been doing a cycle jungle. Very, very good question see I've had a lot of discussion debates internally as well. What is the minimum scale at which you have the list. Enough. Right, the people investor will see I can see it right. I think I do not know I somehow I feel based on everything that I've seen at least megawatt in mechanical systems is is the number Brock has direct field experience. Maybe at least you may need two data points, at least for a for an investor to say that it is the efficiency is increasing the size. I mean, you know, I just, I think every mechanical system that I've seen it starts with paper off five megawatt or something Brock you have any thoughts. I completely agree with you because, for example, what I'm doing, you know, we have a first of the kind of control system, but whether or not it's controlling a one megawatt or 500 megawatt facility. It's still the same control system so it's pretty easy and straightforward to deal with. In terms of things like compressors and pumps it's like is it in the same frame size family or the same technology family so you know as small as you can, but if someone had already demonstrated 10x larger and that's what we actually had to do in our facility. When it comes to mechanical systems and sometimes chemical systems it's really difficult. For example, you know all the talk about post combustion capture on CCGT. And chemical systems that have to be built have never actually been demonstrated in terms of mass transfer and the kinetic rates that would occur as a result of the sort of interacting with co2 because you're dealing with heat fluxes you're dealing with changes in the actual internal velocities as the fluid moves up and down. So to Robbie's point like you do have to prove it to the extent of market understands or investors, you know, point A to point Z. And I do that because the cost of doing is just too big for an academic institution or national lab. But that's where to Adrian's point you take what's actually been done in the field, and you start to kind of hybridize what you can do in the lab with computer simulation, trying to extend it based on practical real world experience I mean you think about it every refinery it's ever been built on the face of the earth is effectively a first of a kind because it's a different chemistry it's a different environment it's different hardware but they're extrapolating based on previous experience. And I'm using that because it's a chemical reactor system it's a mechanical system. You're dealing with things that don't have modularity per se. But modularize as much as you can as long as the physics don't change as long as the economics work. If you don't modularize or shoot yourself in the foot. I would say that even in carbon capture, almost all carbon capture plants are based on thermal energy, right. I mean you're basically using heat to dissolve the CO2 similar issues your scale matters there too. Right. So, so even the carbon capture if everybody very serious or carbon capture and looks like there's a broad agreement worldwide now that we'll have to do carbon capture, you have similar issues the financing some of the mega projects. Hi, we have five minutes left. I think this one question we absolutely need to take from the audience, we have a lot of students are listening. There's a question about, well, just to be brief, one minute per person right, how can current student PhD engineer scientists tune their studies to prepare more for careers and energy. One minute otherwise, who wants to take that first after this question will wrap up the day session. I mean I can go first. Let me just say one thing which I have also a lot of students what I tell them is that how do you fit your science into the broader you have you need to understand the scale of the problem in the energy sector. Appreciate the scale of the problem and also have some idea of techno economics, then then see how your science can really help move the needle in the big time. This is really a rapid API mindset, I think really good. Yeah. Okay, blog and agent. Yeah, agent first. I guess I would maybe say two concrete things that you can do as a student do a lot of do as many co ops as you can go work in industry for a little while see how people talk, see what skills are important see how people make decisions, see how they influence others. And the other firm recommendation is, if you can get into a good MBA program and a good university. This might be a way to kind of shortcut a lot of the early career learning so that you can be a more effective engineer with these kind of larger infrastructure projects. And for in Berkeley business schools are not too bad so. I would say, if you're just graduating look at what technologies realistically can be implemented next five to 10 years and find the companies that are going to do that, because those are the firms are going to help you right out of school, get the experience that's needed to kind of transition into the real world. In academia, that's a very different discussion if you want to go into the private sector, look at who's actually going to build the stuff that's capable being built five to 10 years from now. Because you might find that the polls either very large or very small, but if you if you want to have an impact. Do the sniff test of what's realistic in terms of what's going to be on the project list in the near future. It's a good thought you can learn about both about the R&B still having right there you can learn about scaling as well for you know the technology implement five to 10 years. We'll back to you. Wonderful, this is great learnings I think if I add everything up with what you all just said, it's really innovating and delivering at scale, this is really the challenge for all of us. No, I'm so glad to be joined by all of you doing this every day. So, again, thank you so much, Ravi, Adrian and Brock for this very well connected set of talks on how to get things from the lab to the real world. Thank you once more and if I can Kaylee have the exit sky slides please. We have two more talks this fall quarter. So in two weeks, we will have a session on circular economy of lithium batteries and we're very pleased to host leaders from north bold and bsf to talk about battery recycling and reuse. And then to end the quarter we're going to have Jessica transit from MIT who will also discuss like today system level considerations for generations storage and renewables. So with that I'd like to thank all of you again for tuning in today and please connect with us online. Thank you very much and have a great day.