Solar-Powered Data Centers with Exowatt

Nov 7

Hannan Parvizian is Co-founder and CEO of Exowatt. Exowatt turned heads earlier this year when it announced a $20 million Seed round with backers including Andreessen Horowitz, Atomic, and Sam Altman, the CEO of OpenAI. The company has developed a novel combined solution for energy generation and storage that concentrates heat from the sun, stores it as thermal energy, and converts it on demand to electricity.

They've primarily focused their go-to-market around the data center boom driven by AI and argue that their solution is purpose-built for the massive energy need of data centers. We were eager to hear from Hannan as the rapid increase in energy demand from hyperscalers and AI remains a huge topic in the energy space, especially as we work to transition away from fossil fuels while also meeting AI-fueled demand.

In this episode, we explore Hannan's insights on the limitations of "traditional" solar PV and lithium-ion battery storage in meeting these needs—and how Exowatt aims to bridge this gap.

Episode recorded on Oct 10, 2024 (Published on Nov 7, 2024)

In this episode, we cover:

[2:01] An overview of Exowatt
[4:56] Exowatt's technology and approach
[6:29] Reasons for pursuing a thermal energy storage solution
[14:46] Hannan's previous startup experience
[16:45] How Exowatt's Seed round came together
[20:07] Response to skepticism about Exowatt's technology
[24:41] Technical details on the Exowatt P3 system
[29:09] Geographical and land use considerations
[34:33] Comparison of Exowatt's approach to other thermal energy storage solutions
[38:59] Exowatt's manufacturing and deployment strategy
[41:48] Pricing and economics of the company's system
[43:10] Current deployments and future growth plans
[43:44] Hannan's perspectives on Exowatt's competitive advantages and challenges
[46:02] Where Exowatt needs help

Cody Simms:
Today on Inevitable, our guest is Hannan Parvizian, Co-founder and CEO of Exowatt. Exowatt turned heads earlier this year when it announced a $20 million Seed round with backers including Andreessen Horowitz, Atomic, and Sam Altman, the CEO of OpenAI. Exowatt has developed a novel combined solution for energy generation and storage that concentrates heat from the sun, stores it as thermal energy, and converts it on-demand to electricity.
They've primarily focused their go-to-market around the data center boom that's being driven by AI and argue that their solution is purpose-built for the massive energy need of data centers. I was interested to talk with Hannan as the massive increase in energy demand from the hyperscalers and AI is obviously a huge topic in the energy space as we collectively try to transition away from fossil fuels on the grid while simultaneously adding significant AI-fueled energy demand.
And I wanted to understand where Hannan and Exowatt believe that "traditional" solar PV and lithium-ion battery storage had shortcomings in solving for AI's energy need that they felt Exowatt could overcome.
From MCJ. I'm Cody Simms and this is Inevitable. Climate change is inevitable. It's already here, but so are the solutions shaping our future. Join us every week to learn from experts and entrepreneurs about the transition of energy and industry. Hannan, welcome to the show.
Hannan Parvizian:
Thanks for having me, Cody.
Cody Simms:
Let's jump right in. What is Exowatt?
Hannan Parvizian:
Exowatt is a developer of modular renewable energy solutions for commercial and industrial applications, specifically to address the growing energy demands on the grid driven by these energy intensive applications such as AI data centers.
Cody Simms:
And when I think of it, I think of the growth of what you just described, the renewable energy generator as it's going to be solar and some storage or wind and some storage, but you are designing essentially a new mousetrap, I think, like a new mechanism for generating power and storage. Can you maybe unpack that for us a little bit?
Hannan Parvizian:
Yeah, absolutely. So our flagship product is called the Exowatt P3. So it's a three in one modular renewable energy solution. And what that means is that we have renewable energy collection, storage, and generation packed into one modular unit. So the way it works is essentially we capture solar energy throughout the day using a custom-developed array of lenses that convert this energy to heat. And the reason we convert the sun's energy to heat is because heat is significantly cheaper to store as compared to electricity, which makes it an ideal medium for storage, for long duration storage, and for dispatchability throughout the day or over multiple days.
We store this heat in a sensible heat battery. What that means is that the battery cell itself stays solid throughout its service life. That doesn't go for any phase changes like molten salt. It doesn't have any chemical or electrochemical reactions happening in it like lithium ion or iron oxide or any other chemical reaction for that matter, and it doesn't have any mechanical moving parts. It's a very simple design in that sense, and this design and this way of storing energy allows us to store energy longer, cheaper, and essentially with no degradation.
And when we need to dispatch this energy essentially deliver it to the customer, we convert it from heat to electricity using a built-in heat engine and dispatch energy throughout the day up to 24 hours a day giving almost a baseload power output to the end use.
Cody Simms:
So when I think of what I think we would now refer to as traditional solar and storage, I think of mostly a chemical and electrical interplay, right? Where you're taking the sun's rays photovoltaics or converting it to electrons, you're then storing it via chemistry through lithium ion and then ultimately releasing it as either electricity or heat where it needs to be.
What I'm hearing from you is a much more of a mechanical solution where you're essentially taking large magnifying glasses like you would use as a kid, point the sun down at something on a sidewalk and then holding that heat as opposed to electrons and then using the heat to drive electricity as you need. Yes? Am I understanding correctly?
Hannan Parvizian:
Yeah, absolutely. The core principle here is around capturing and storing the sun's energy in the form of heat. And the reasons for that are essentially that it's just cheaper and easier to store heat as opposed to electricity. So as you said, traditional solar PV, the way it works is you capture the sun's energy photons, hit the solar PV panel, and using the semiconductor you built in, you convert that to electricity, normally DC output electricity, and you can then sort this electricity in an electrochemical battery like lithium ion battery or any other type of chemistry for that matter, and then dispatch it in form of electricity as well.
The downside of traditional solar PV plus battery electrical systems or BESS is that the cost of storing electricity is very high, even though it's a fairly dense storage medium. And this doesn't allow you to store electricity beyond two or four hours feasibly, maybe even up to six. But that basically means that for applications where you're looking for more than 10 or 12 hours of dispatch, daily dispatch, or you're looking to store energy over multiple days or dispatch it over multiple days, this way of capturing the sun's energy and storing it isn't feasible. And so that's why we've designed the Exowatt system and the module around thermal energy and around capturing and storing heat because of those benefits of being able to store longer, cheaper, and with really no degradation.
Cody Simms:
What helped you uncover the problem that you just laid out there? What was the "aha" or the reason for pursuing this in the first place?
Hannan Parvizian:
Yeah, absolutely. The reason we started a company is that Jack, who was the co-founder of Exowatt, also a founder and CEO of Atomic, the venture studio where we incubated Exowatt had this thesis that there are going to be the secondary order effects of AI, essentially AI boom. So not the application layer, but the infrastructure layer, the chips layer, the energy layer, and we really need to focus on those secondary effects and that's where the opportunity is essentially in the picks and shovels, not in the gold rush itself.
So when him and I got together, I'm a mechanical engineer by training. I've worked in the energy industry for a number of years at Siemens, at General Electric, at Tesla, and wanted to develop a solution to address this problem. The key insights that we had were, it's not so much about the physics of the technology, it's about really the approach you take in implementing the technology.
And so historically, the idea of capturing the sun's energy in form of heat and storing it in form of heat isn't a new concept. People have tried it before, even centuries before, but the main difference for us is that we really focused on the modularity of the unit. And the reason is that we had this key insight that technologies like solar PV have been able to go down the cost curve tremendously over the last couple of decades because they were modular technologies. For example, the cost of solar PV manufacturing has dropped by over 99.6% over the last couple of decades, which is truly remarkable.
And you're seeing the same trend with electrochemical batteries like lithium ion batteries where the cost is going down almost by 50 to a hundred percent year over year. And the reason is very simple because these are modular technologies. They can be built in a factory where it can really scale manufacturing, where you can have control over the costs and where you don't have to worry about on-site construction and essentially deployments that take very long. And that was the main insight that we had.
Cody Simms:
That would seemingly be an argument in favor of the status quo, not building something new if the existing solar PV space is already modular. Yeah?
Hannan Parvizian:
Yes, with the difference that the existing solar PV has two disadvantages, solar PV plus battery electrical systems. Number one is that while it's modular, the supply chain is very dependent on rare earth materials and mining rare earth materials and is mostly coming from China. And that's a big component of this. China controls and manufactures over 80% of the supply chain and manufacturing for solar and wind today. The same argument is true for battery electrical systems where again, most Chinese companies dominate that space. So that's from a supply chain manufacturing perspective.
The other aspect is really about the cost, which cost of electrochemical batteries is high relative to thermal batteries by almost an order or two orders of magnitude. But more importantly, part of the reason it's high is that over their lifetime, electrochemical batteries degrade and you have to replace them and augment them. So the best example that you probably know is you have any electronic device, whether it's your phone or your laptop. Over time the battery just becomes weaker and weaker. And the same is true with these battery storage systems, especially when they are getting charged and discharged so rapidly and aggressively throughout the day to provide power to the grid or to the end customer.
And with heat batteries, you don't have this degradation issue essentially because there's no chemical reaction, there is no degradation. The material, especially if it's a sensible heat battery where it just stays solid, just has the same level of performance for 30, 40, 50 years of service life and that brings a cost down. The materials that you're using, typically materials that are abundantly available don't require any rare minerals and materials. There's no lithium, there's no cobalt, there's none of that in there, and it makes it easier to source them from domestic supply chains and makes it easier to also manufacture them. So those are all the reasons why solar PV plus battery electric systems, electrochemical batteries is a good solution for sure, for applications that are just looking for intermittent available power or just for short duration storage for two to four hours. But anything beyond that, they become economically not feasible.
Cody Simms:
I'm hearing you say two main things that you believe were insights that led you to build and design the Exowatt current system, which is one, that the existing photovoltaic solar and electrochemical storage mechanisms that we're trying to deploy across the US are heavily reliant on a Chinese-dominated supply chain, and two, that electrochemical batteries in particular have a shelf life, have a degradation to them.
I actually thought you were going to say something totally different, so I'm interested to hear your feedback. When I think about Exowatt, it seems like one of the main advantages in theory is time to market, which is these systems are modular. You can take them, I think, off of a semi-truck and put them on the field and use them pretty quickly to get power up and running. And in this world where AI energy demand is something like doubling in size every few months or something crazy at the moment, how do we get systems up and running and in place very quickly? I don't know if you factor your time to market ultimately as having an advantage over a traditional solar farm, for example.
Hannan Parvizian:
Yeah, no, that's a great point. I think that number one constraint or the number one ask that we hear from customers, especially data center customer, is the time to power. So to your point, how fast can you deploy this so that I can take my data centers live? The biggest challenge data centers have today is that there's not enough capacity available on the grid. And so one of the advantages of this technology is that you can deploy this behind the meter off-grid and essentially take an application or data center live sooner rather than later.
But as it relates to the time to market or time to power as compared to solar PV plus electrochemical batteries, I don't think we necessarily have an advantage there. The solar PV is widely available. Battery electrochemical batteries are widely available. The advantage we have there is really, as compared to other sources of firm clean power or baseload power, renewable energy sources of firm clean power or baseload power, for example, like nuclear or geothermal or hydro or large solar thermal installations.
And those are also technologies that theoretically could provide baseload power to a data center, but have the downside that I was mentioning that because they are non-modular, they take forever to set up, they have billions of dollars in CapEx. Sometimes they take 10 to 20 years to set up, and that's what really kills those technologies as it relates to being able to power data centers today. As a result of the fact that there isn't really a great firm clean power solution available to data center customers today, the tragedy is that most of these data centers, for the most part are being powered by traditional fossil fuel sources and diesel generators as backup generators and natural gas.
And that's something that we want to hopefully prevent happening at Exowatt. That's part of our goal and mission is to get these units out, these modules out to customers as fast as possible and really provide them firm clean power and give them an option as opposed to going back to traditional fossil fuels to power AI. Because AI is an amazing application and has amazing benefits for humanity and society, and we definitely believe in the vision that intelligence and energy have to ultimately free this new world that we're embarking upon. And we want to make sure that we don't power AI with essentially fossil fuels, rather we power it from the sun, the cleanest, freest source of energy that we have available to us today.
Cody Simms:
And Hannan, this isn't your first rodeo. You built another company over the last seven or eight years that raised multiple rounds of financing, I think was a YC company. Lightspeed backed you a few times. Ultimately it sounds like the company didn't wholly work out, but maybe share a little bit about what you learned from that experience, what the company was and what you learned.
Hannan Parvizian:
So the inspiration for that company was when I worked at Tesla, we had this supply chain logistics issue that we wanted to address using long range drones. So I partnered with two co-founders from Stanford and launched Volansi, which was the name of the company from YC in winter of '17. And what we did is we developed a vertical takeoff landing drone with a thousand-mile range, 50 lb. payload capacity designed for enterprise delivery applications and also military applications.
And as you said, we had an amazing group of investors including YC, Lightspeed, our customers and many more who back the company and we service many enterprise customers and also most of the US armed services and eventually sold the technology and the company to one of our defense partners middle of 2022. And tons of lessons learned for that. That was definitely my first startup, if you will, and it was probably the most difficult startup to start because not only we were building hardware, but hardware in a regulated market and selling hardware to a very difficult customer, both on the enterprise and military side.
So all of those are lessons that I've learned and I'm now taking to Exowatt, which in many ways is similar but also different, but there's so many things that I can see one-to-one happening there. Not only in the product development from a hardware perspective and software perspective, but also in how to bring this to market, how fast we can bring this to market, what are the challenges and hurdles that we have to overcome and it's like I'm playing a game where I've seen a lot of levels before and then now it's just going through those levels as fast as possible.
Cody Simms:
One of those levels is certainly what you announced a few months ago with your Seed round, which is much like your first company Volansi, which had some tier one investors like YC and Lightspeed. You announced a Seed round with participation from Andreessen Horowitz and Sam Altman himself personally. How did that all come together?
Hannan Parvizian:
Yeah, we've been incredibly lucky to have not only Atomic who has backed us since the inception of the company, but also amazing investors who led our last round, including Sam Altman, who of course is the most important person in AI in many ways today, who really bought into the vision and the mission of the company to be able to deliver renewable energy data centers as soon as possible.
Cody Simms:
Sam, just for listeners, in case you haven't been following, Sam Altman is the CEO of OpenAI.
Hannan Parvizian:
Yes. Hopefully everyone knows that by now. Yeah, also Andreessen Horowitz, Felicis and 89 and many other great investors that we have on the cap table that helped us get to where we're at today and are helping us scale the business and bring this product to market as fast as possible. It's been great really to have the support of these types of investors.
Cody Simms:
How did it happen? What was the process of raising that Seed round?
Hannan Parvizian:
It was really where we were at a juncture where we figured the concept for the product, basically the journey for the product and the company was one that was really focused on unit economics from day one. So we had set a very ambitious and aggressive goal for ourselves to be able to ultimately generate electricity at 1 cent per kilowatt-hour. And from that goal, we worked our way backwards to a product essentially by iterating on different configurations and different materials and overall we went through 50 plus different designs and prototypes and systems to arrive at the product that we believed could get us to that goal ultimately as we scale manufacturing.
And so once we had that figured out, we went to our network of investors and Atomic's network of investors and Sam and really got them excited about this product and this approach to the technology and they bought in and we quickly got a round together and have been executing against our plan so far, including launching the product to the public just over a month ago now, actually exactly a month ago now.
Today at the RE+ conference in Anaheim where we launched the Exowatt P3 to the public, be able to place pre-orders, and we announced a very significant backlog of reservations already for the product over this point, I think 80 gigawatt hours from customers in data center space, the renewable energy developers of the world and data center developers of the world. And we're super excited that there's so much interest and demand in the product today and we're really working as fast as we can to get it to market and get it to customers.
Cody Simms:
That's a huge number. We're going to come back to that, the traction and the customer demand you're hearing, I'm glad you mentioned RE+ because I was going to ask you about that. You guys had this big unveiling plus I think famously Leonardo DiCaprio was in attendance. He's one of your investors I believe. And then I think the next day or two afterward, there were a few articles that came out in the energy tech press that had some questions about your technology and the basic crux of them were a few things.
One was in Latitude Media, one was in Canary Media. The basic of them was, "Hey, we've seen most of this technology before, right? We've seen these concentrated solar for heat. We've seen sterling engines that take heat and turn it into power. We've seen a lot of this technology. Why is now the time to take technology that didn't work before in a scalable way from a renewable energy perspective and bring it to market, especially given that cost curves on photovoltaic solar and electrochemical storage have come down so far since the last time those things didn't work?"
Hannan Parvizian:
Great question, and I think the answer is exactly what you said. The last part, which is if you look at the cost curves of solar PV and battery electric solutions, they have come down this learning curve of manufacturing so rapidly because of the scale of manufacturing. The other component that we say at the company and we say also to our stakeholders, investors, is we are standing on the shoulders of giants because of exactly those previous attempts that people have made and capturing solar energy in form of heat and storing it and being able to convert to electricity. There is so much that we can look back at and see what mistakes people made and what they didn't make, and that's ultimately allowed us to invent this product as it is today, the Exowatt P3 in this configuration, which has never been done before.
And really the reason we have so much confidence that this is a better approach is really because it's a modular system. It's a modular design that basically drops the cost of learning down to tens of thousands of dollars as opposed to tens of millions of dollars. So that's what the key insight behind the company is. And I mentioned we've developed different prototypes and systems over the past couple of years and wrapped the iterator on the product, and we were only able to do that because of this modular design. If we had to go test a 25 megawatt system or a hundred megawatt system and just a prototyping phase would be like five years of setup, that would just be the same exact recipe as let's go build something huge that's optimized around economies of size.
And by the way, it's very tempting to do that because on paper, the technical efficiencies of large systems, especially large thermal systems, are always better than smaller ones. So the best example is if you grab a small rock in your hand, a hot rock, that one is going to cool down faster than if you have a very giant rock that's hot. That cools down slower just because of the mass and the amount of heat that it's losing.
There is this temptation to go build bigger and optimize for technical performance and efficiency. Our approach has been very much different that we have optimized for economics, cost, time to market, time to power, and technical efficiency and performance have been secondary factors for us. Not that we are building things that are not as efficient or as performing, but that we haven't prioritized or optimized design around. That's why we think we have this winning recipe today, and we also appreciate all the skeptics that are out there. We have surrounded ourselves with them. We have an advisory board that's full of these folks who've been through these previous attempts and challenges of the industry and are helping us make the right decisions. That's I think, a great position to be in today.
Yin Lu:
Hey, everyone. I'm Yin, a partner at MCJ, here to take a quick minute to tell you about the MCJ Collective membership. Globally, startups are rewriting industries to be cleaner, more profitable, and more secure. And at MCJ, we recognize that a rapidly changing business landscape requires a workforce that can adapt. MCJ Collective is a vetted member network for tech and industry leaders who are building, working for, or advising on solutions that can address the transition of energy and industry.
MCJ Collective connects members with one another with MCJ's portfolio and our broader network. We do this through a powerful member hub, timely introductions, curated events, and a unique talent matchmaking system and opportunities to learn from peers and podcast guests. We started in 2019 and have grown to thousands of members globally. If you want to learn more, head over to MCJ.VC and click the membership tab at the top. Thanks and enjoy the rest of the show.
Cody Simms:
You described the product at the very top of our conversation, but now that people have had a little bit of time to absorb everything you said, maybe describe the product again. Physically, what does it look like? How does it work?
Hannan Parvizian:
Yeah, absolutely. So again, it's a three in one solution and modular. So what we mean by modular is each module is roughly the size of a 40-foot shipping container, so about 40 feet long, eight feet wide, eight feet tall, and it's made of three elements. The first element being the solar capture system, energy capture system, which is an array of optical lenses. These are very large Fresnel lenses that capture the energy from the sun throughout the day over the eight or 10 hours of solar resource that's available on a given day.
Cody Simms:
Just to clarify, capture or focus? I envision them as they basically are magnifying the sun into even tighter heat. Would that be the right way to think about it?
Hannan Parvizian:
That's a conversion, yeah, so essentially they capture the energy from the sun. That means they're tracking the sun and then they are converting it to high temperature heat, similar to the function principle of that magnifying glass that you mentioned. And then basically we are impinging this heat directly onto our sensible heat battery cells.
Cody Simms:
This is the thing that looks like a box? I mean it looks like a shipping container basically.
Hannan Parvizian:
Yeah, exactly. That's the enclosure is the shipping container. Inside, there are these cells, very large cells that are made of proprietary silicon composite based material that essentially stay solid. They just get hot with the energy that they receive from the sun or actually from other sources as well. We've built the capability to charge them from other sources of electricity and heat as well.
Cody Simms:
This would be in the event that it's a cloudy day, for example, you still need to be able to generate heat into this system. Yes?
Hannan Parvizian:
Well, not only a cloudy day, but more importantly if this is connected to a grid or some sort of other source of energy, imagine it's paired with a solar PV field. You can take excess energy from the solar PV field and charge it in these batteries, so you can really use them as a purely functioning battery as well as a three in one full stack solution.
And anyway, so you store the energy in the form of heat, the cells are inside an insulated enclosure, and five of these cells make what we call a battery pack. We take two packs and essentially connect them to the heat engine. The way essentially the heat engine works to convert the heat to electricity is we circulate air through the battery packs. This is our working fluid, if you will, and the air picks up the energy from the battery cells and takes it to the heat exchanger of the engine, and essentially the engine on the hot side starts expanding the working fluid there, starts expanding, pushes a piston inside a generator coil on that basically makes electricity.
So that's really the simple functioning principle of it. The beauty of this design is that you can size every component independent of each other. So for example, when a customer comes to us and says, "I have this a hundred megawatt site and I want to run it for say, 16 hours a day," that gives you a total amount of energy that's needed per day, and we can size the battery packs exactly, and the number of engines exactly to meet that energy profile.
And so it's very linearly scalable with any project size, any load curve. It's what some people call shaped power, so you can really shape the way the power is dispatched, and this is the key ingredient of enabling us to not only serve small data center projects and small commercial industrial loads, but also huge ones. Data centers today on average are 60 to a hundred megawatt and powers, and they're becoming bigger. There's gigawatts and hundreds of gigawatts of data centers that are being planned today. So that's the beauty of the system. We don't need to change the modular design, we just scale the number of modules to meet exactly any given project size.
Cody Simms:
In hearing you describe it's steel and glass and silicon bricks, and you said air circulation. It sounds like the materials are not going to be subject to significant challenges in terms of sourcing.
Hannan Parvizian:
Yeah, exactly. That was one of the key ingredients of the thesis for us was to really focus on materials that we can source from the US that don't have any need for rare earth mining, that don't come from China. And we experimented with a lot of materials to arrive at this particular selection, but this also means that we can ultimately get the cost of these modules down to the raw materials costs as we scale. And raw materials costs is rather insignificant. It's just really a matter of the total weight of the system that determines the cost of the system and the raw materials are fairly available and cheap.
Cody Simms:
Geographically, are you constrained to places that have significant sun? Is this mostly a Sunbelt deployed solution in the US for example?
Hannan Parvizian:
Yeah, yeah, definitely. If you are looking to deploy this solution, the Exowatt P3 solutions as a behind-the-meter off-grid solution, which is many of the applications that we're serving are like that. You want it to have enough solar resource available, and that's typically available to it, and that's typically along the Sunbelt, anywhere from eastern Oregon to California to Nevada, Arizona, Texas, New Mexico, Georgia, all the way to Florida. Those are the places where it works best and has the highest performance and capacity factors.
Of course, you can deploy it in areas that have less solar resource available to them. It still will work. It's just a matter of the capacity factor, as in how many days of the year does it actually work to its full performance level. So we don't necessarily recommend it for those geographies today. The beauty of data center application is that the majority of the 2,400 plus data centers in the US that have been built today and are being built are along the Sunbelt. So over 60% of them are being built in the Sunbelt areas. And primarily that's because that's where the land is the cheapest and most available and the energy is the cheapest because of the solar resource.
Cody Simms:
Let's sure hope they continue to have access to ample water for their cooling operations in the Sunbelt.
Hannan Parvizian:
Yeah. And there are many new cooling technologies that are being developed as well. I think one key statistic here that many folks may not know is 60% of the operating expenses of a data center are basically related to the energy bill, and that's for the compute and for the cooling. And I was about to say the only data center that are not in the Sunbelt are right now around mostly the Virginia area, and those are the ones that you can't necessarily power directly with an Exowatt P3 solution, but you can definitely offset the power that's being generated there with a Exowatt P3 implementation somewhere else in the Sunbelt area.
Cody Simms:
And then what are we talking about from a footprint perspective? One of those articles I mentioned did some back of the envelope math and said a 50 megawatt peak for 24 hours in a sunny state like Arizona would require over a thousand acres to deploy Exowatt onto it, which is bigger than Central Park and substantially larger than an equivalent photovoltaic solar farm and storage would require for roughly the same amount of power. How do you square that or how do you think about that from a sort of a land use perspective?
Hannan Parvizian:
Couple of things there. So number one, it's really important to think about how energy hungry these data centers are. So a hundred megawatt data centers consuming the equivalent of 80 to a hundred thousand US households of energy, which is an insane number to think about. And these data centers are becoming even more energy hungry and energy dense in that sense.
In terms of the footprint that it requires to power these data centers, it is significant if you think about it as opposed to the data center, but it's not as significant if you think about the fact that you're powering a hundred thousand US households with it. And from a power density, actually the Exowatt P3 is more power dense than the equivalent solar PV plus electrochemical battery or best solution. But from an energy density perspective, you're essentially dealing with the same amount of solar resource available. So we still get the same amount of watts of solar energy available to us in Arizona as any other solution would.
And actually overall, the footprint density of an Exowatt P3 solution is less than a comparable PV plus BESS implementation. Where I think the feedback and criticism from some folks is, "Well, this is still a lot of land, why don't we go build nuclear or geothermal or something else, something that's more power energy dense?" And our response to that is those are of course great technologies to have and they definitely need to be part of the energy mix. But coming back to that time to power, they are not available solutions to any customer today, and even the most optimistic implementations of those are at least five to 10 years out.
The other factor to consider is while it looks like a lot of land in the context of Manhattan in Central Park, it's not a lot of land in the context of where these data centers are being built. We actually did a study if you wanted to power the entire US with Exowatt P3 24/7, it would take less than 10,000 square miles to do that, which is like a teeny tiny spot in the state of Nevada could power the entire U.S. So there's tons and tons of land available where these data centers are being built and it's cheap land, and if you factor in the cost of land into the total cost of the generation of electricity or the LCOE gets a fraction of a cent.
So we're not really worried about this takes a lot of land. That's a given. That's understood. That's because that's how much sun will receive per day per square meter. But the fact is that land isn't really number one constraint and maybe just a fun fact is if you want to make it take less land, the best thing you could do is to essentially try to capture more energy from the sun and space and basically focus it down on earth so then you can get more solar resource per square meter.
Cody Simms:
Just don't walk underneath that beam of light.
Hannan Parvizian:
Yes, exactly. That wouldn't be a practical implementation, but that would be the way you could get the physics to least work.
Cody Simms:
Okay, helpful to hear your perspective on that. I'd love to hear how you think about Exowatt as compared to other thermal heat batteries that we've had on this show before. We've had folks like, and you don't have to compare about anyone individually, but we've had Rondo and Antora and companies like that that are also taking renewable power in the form of electricity and turning it into thermal heat from a storage perspective to then I think mostly be used as heat, though I think they have applications where you can then convert the heat back to electricity if you want to.
You're, as I understand it, obviating the step where the power goes to electricity and then to heat. You're just keeping it as heat the entire time until you need it as electricity. But maybe explain to me how the use cases for those types of applications that would be different than the use cases that you solve for.
Hannan Parvizian:
The first thing is that the Exowatt P3 solution, as I mentioned, is a three in one solution. So it's not just a battery, it's an energy capture system, it's a storage or battery and it's a generation system. So that's number one difference. And what that means is that we don't have to be grid connected. We don't need a source of electricity or source of energy to connect to our batteries. They basically have the lenses that capture energy from the sun directly.
And so the idea of capturing excess electricity from the grid, storing it in a heat battery and then dispatching it in form of heat and electricity, there's definitely grid applications for that, whether it's to support and decarbonize industrial heat, which is what a lot of these companies are doing, or whether it's to take advantage of those particular energy arbitrages that might be available where essentially you have so much solar PV or wind available on the grid, the cost of electricity essentially is negative and you just have to take it if that makes sense to take it, store it in form heat and convert it to electricity.
But if the input cost of electricity is not zero or negative, that conversion mechanism or efficiency round trip efficiency doesn't really work out well from an economics perspective. And so that's why when we did the analysis, we really realized that if you want to be able to deploy this without having to wait for interconnection agreements and the grid and deploy this, where you don't have to worry about just these energy arbitrage opportunities that exists on the grid today, you have to have your own collection system built in. And that's why we have the optical table with the collection system.
Circumstances in areas and geographies where the customer really actually just wants to run it as a battery, we can also run it as a heat storage battery solution and generation solution. And from an application perspective, I think as I said, we're focused on the power market, on data centers in particular. A lot of the other companies that are out there are focused on industrial heat, which is also a gigantic market, but one that we're not focused on.
Cody Simms:
How much of the interest that you're receiving from your data center customers or prospective customers is for this behind the meter solution where the fact that you can do generation and storage and they can plug it in and run their operations and they don't have to get interconnect and they don't have to get the system plugged into the grid is the selling point for why they want to work with you?
Hannan Parvizian:
So 50 50 is the selling point on behind the meter versus in front of meter grid connected or not grid connected. I think the bigger selling point for customers is the time to power. This is something that they can actually deploy, whether it's grid connected or not, it's available as a technology. Scalable, that's another component that's really important here. And then yeah, whether you do an implementation where this is connected to the grid or behind the meter off-grid, that's really a function of where the site is, what kind of reliability of power the customer is looking for, how they're looking to use this power, whether it's for AI training models or whether it's for inference models or whether it's just your regular data center operations. Those are all things that affect how it's ultimately deployed and implemented.
But I think what customers are most excited about is that this is a technology. This is clean renewable energy technology that is available today that is going to be deployed with customers commercially in the US at least in the coming months, and then is going to be able to really scale to meet their demands. Because as I said, these data centers are not small. They're not looking for one or two modules. They're looking for hundreds of thousands and millions of modules that need to be deployed to power their data center operations. And that's what we are focused on as a company is really scaling our manufacturing to that point.
Cody Simms:
And what does the manufacturing look like for you? Are you building all of this in-house? Are you working with contract manufacturers? You mentioned 80 gigawatt hours of interest currently. That is a lot of system to scale, so how do you get there?
Hannan Parvizian:
Yeah, good question. I mean, we made a lot of business decisions that are focused on scale. So first of all, from a business model perspective, we sell the modules, the hardware and software to energy developers or data center developers or owner-operators. So we don't sell energy, we're not doing PPAs, we're not selling heat or electricity as a service. Rather, selling the hardware and software, which is the fastest, most scalable business model in our opinion.
Second, we wanted to de-risk manufacturing as much as possible, and therefore we partnered with the leading contract manufacturers here in the US that have capacity, that cap proven to scale products and to the gigawatt hours before similar products. And we're working with them to do our first implementations of the product, the first batches of the manufacturing and scale with them as well to multiple sites over time to the tens of gigawatt hours that is our goal and that's also driven by scale.
The fact that we want to bring this to market so fast, at scale so fast, we didn't want to spend any cycles on trying to figure out manufacturing and building a factory and having to raise a lot of money to do that and making all the typical mistakes that one would make in setting up a factory for the first time. So for us, it's always been driven by the speed of deployment and time to power and scale, and every decision we make is first and foremost around the economics and second around the scale and deployment timelines.
Cody Simms:
Hearing you say that you're doing just direct product sales is certainly different than I think just about any other solution that most big energy consumers, specifically data center developers, are probably finding on the market where they're not going out and selecting from different manufacturers of the power they want to put on site for the most part.
Hannan Parvizian:
Yeah, absolutely. And we're lucky to be partnering with the leading energy developers in the US that can help us implement this technology for the data center customers. And again, that's where we don't feel like we have a unique advantage there to take on all the risks of project development, all the project financing risks, all the balance sheet risks, debt risks as a startup. And we prefer to partner with folks who've been doing this for decades and have the means to do it and the expertise to do it. And we really want to focus on what we need to get right is scaling manufacturing and doing that as fast as possible to ultimately get the cost of product to the point where we can generate electricity at 1 cent per kilowatt-hour, which is our ultimate goal.
Cody Simms:
How much are you selling the system for today?
Hannan Parvizian:
So we announced standard 25 kilowatt-hour electric module at our launch event in Anaheim for $7,500. That's the selling price, retail selling price, and that means that the equivalent cost of electricity to a customer who buys that, the generation cost is about 4 cents a kilowatt-hour. That's before any subsidies are applied to it. Of course, if the Inflation Reduction Act persists, this product qualifies for a number of subsidies including the production tax credits, the investment tax credits for both generation of renewable energy and storage of renewable energy and domestic content tax credits because of the fact that it's made and sourced in the US and many more, of course, depending on the location where it's deployed at. So the effective cost of electricity for customers could be as low as one to 2 cents a kilowatt-hour.
Cody Simms:
That would be highly variable though on time of day and solar availability, etc. Yes?
Hannan Parvizian:
That's not really a function of the solar availability. This is the function of total energy made by a module over its service life and the cost of the module. With subsidies, this is going to be extremely competitive. Even without subsidies, it's an extremely competitive offering and our goal is ultimately over time to reduce the price and reduce the cost even further to make energy as our mission says, almost always available and almost free.
Cody Simms:
And you've talked about time to power a few times. Where are you in terms of your ability to start deploying?
Hannan Parvizian:
Yeah, we're actively working on a couple of deployments that we're going to be announcing also hopefully very soon with leading customers here in the US, sites in Florida and in Texas, and some more that we have in the pipeline. And these are commercial deployments of various configurations both behind the meter and front of the meter to support various applications including the data centers and excited to share more with you once these projects have gone live.
Cody Simms:
I'm sure there will be some who listen to this show and still come away just basically skeptical that, "Hey, why would you step in front of the freight train that is photovoltaic solar and electrochemical battery storage?" It is the largest driver of growth in the US energy grid system today. You've tried to explain why that is a great solution, why yours might solve different problems.
From where I sit, being an entrepreneur and trying to see where the world is going and seeing a problem that's ahead of us with the data center boom and all of that and how do we solve that quickly is a hard thing to do. I know you're probably getting people who come at you with all sorts of different criticisms, questions, whatnot. From where I sit, I think it's awesome that you're trying to come up with a solution to this problem that's in front of us and time will tell if it's the right one or not. I'm curious if you have any last thoughts or remarks or anything you want to share about that or anything else that's in front of you as you build out Exowatt.
Hannan Parvizian:
Yeah, of course. I think it's like what you said, you want to be where the puck is going, not where the puck is at today. Really for us solar PV plus BESS, as I mentioned, is I think a perfect solution for applications where you're looking for intermittent cheap solar and dispatchability beyond that for two to four hours. If you're looking for baseload power or dispatchability and availability of power beyond 12 hours, I'm not the one to say this. I think anyone in the solar PV plus BESS industry would categorically tell you it's not economically feasible because of the cost of storing electricity.
This is the only solution that I can think of, which is a combination of solar and batteries that can become economically feasible and is actually economically feasible even today as we are starting and let alone scaling. And I think what really drove the cost of solar PV plus BESS to come down over time so significantly is the scale of manufacturing. And we're taking that lesson one to one and essentially implementing the same thing here with a modular design that can scale to gigawatt hours of production very quickly. And I think we're not saying anything different there. It's like solar PV plus BESS is a great short duration storage solution, not a long duration source solution.
Cody Simms:
Hannan, thanks for joining and I'm sure you're probably hiring, looking for folks to join the team. If there are any key skill sets that you need or we have a lot of people listening to this who work in big tech, do you have any asks in that regard as well? Please share away.
Hannan Parvizian:
Yeah, absolutely. I mean, we are actively hiring at Exowatt. We're always looking for amazing engineers, operators, salespeople, anyone who is interested and passionate about our mission to join the team. And same to your folks who are listening who might be customers or partners or investors. We're always open to having conversations about how we can solve different energy challenges for commercial and industrial customers and even utilities. And hopefully we can solve those with a solution that's made in the US with raw materials that are sourced in the US that really helps stimulate US manufacturing and supply chain and creates jobs in many communities and ultimately give the US the leads that it once had in renewable energy and take that back from China.
Cody Simms:
Hannan, thanks so much for joining.
Hannan Parvizian:
Thank you, Cody. Take care.
Cody Simms:
Inevitable is an MCJ podcast. At MCJ, we back founders driving the transition of energy and industry and solving the inevitable impacts of climate change. If you'd like to learn more about MCJ, visit us at mcj.vc and subscribe to our weekly newsletter at newsletter.mcj.vc. Thanks and see you next episode.