Rondo Energy’s complicated path to building heat batteries

In 2022, John O’Donnell and Peter von Behrens figured out how to design a heat battery that would deliver heat at very high, constant temperatures. The breakthrough came on the heels of two years of research and development, some of which took place in Peter’s garage.

Now, John and Peter were ready to prove their technology at commercial scale. So they approached a long-time innovation partner, Calgren Renewable Fuels, about deploying a 2-megawatt-hour heat battery for industrial heat delivery in Calgren’s Pixley, California plant.

That’s when the next wave of problem solving started. It turns out, designing the technology was just half the battle. Manufacturing and installing a commercial scale demonstration proved harder than expected. In October 2023, they started on a six-month construction project that threw a myriad of challenges at the new start-up.

In this episode, Lara Pierpoint talks to John O’Donnell, co-founder and chief innovation officer at Rondo Energy, about the bumpy road of building a first-of-a-kind commercial demonstration. They cover things like the structural engineering challenges of scaling a new technology, and finding the right construction partner. Plus, John explains what new design specs mean for the company going forward.

The Green Blueprint is a co-production of Latitude Media and Trellis Climate. Subscribe on Apple, Spotify, or anywhere you get podcasts.

Transcript

Tag: Latitude Media, podcasts at the frontier of climate technology.

John O'Donnell: One of our earliest molds that we made, the release compound was PAM, the spray from your kitchen. That was a problem because when we turned that one on, the PAM released some smoke, and my co-founder, his wife, said, "You only get to smoke the house once."

Lara Pierpoint: I will say you've officially checked the box of real innovator, because I don't think unless you've set a fire in your own garage of some sort slash produced a large amount of smoke, you can really count that. So well done.

John O'Donnell: The second time the fire department came to Pete's house, it was not his fault. There was a neighbor who had run into a fire hydrant or something.

Lara Pierpoint: In 2020, John O'Donnell and Pete von Behrens were holed up in Pete's garage, experimenting with ways to stack bricks for heat conductivity. Their goal was to figure out a way to use bricks as heat batteries and industrial processes. In theory, it's a simple concept.

John O'Donnell: If you put a stone in your oven, you heat it up, you put it in your bed, it'll keep your feet warm all night. You stored electricity in an unrefined material that was literally dirt cheap.

Lara Pierpoint: But in practice, it's more complicated.

John O'Donnell: The challenge of heat batteries is moving the heat, and the other challenge in a heat battery is you must deliver heat at constant temperature.

Lara Pierpoint: In Pete's garage, John and Pete were testing different ways to mold and configure bricks that would create narrow passages for air to move between the bricks. The air would then trap that heat and deliver it as steam.

John O'Donnell: The very small air passages are instrumental to being able to deliver the same outlet temperature all the way through discharge. So how are you forming those things? And if you're making thin structures with some sort of mold, can you pull the mold out before you use the brick?

Lara Pierpoint: This is where the PAM came in. They sprayed the mold with it, hoping it would make it easier to remove from the brick, but if you've ever sprayed a skillet with the stuff and forgot to turn off the burner, you know what happened next. The PAM left behind a residue that created a lot of smoke when they heated the bricks. In fact, there was so much smoke that a neighbor decided to call the fire department.

John O'Donnell: We've gone down paths where we don't reuse release compound today. Let me just say it that way.

Lara Pierpoint: But John and Pete still had to figure out how to create these very small air passages. Two years later, they would have a solution, and they took it to their first industrial customer, where there was very little room for error.

I'm Laura Pierpoint and this is The Green Blueprint, a show about the architects of the clean energy economy. We've already invented most of the solutions needed to decarbonize the global economy, but many of those technologies are not yet commercial, and they need to get financed and built at scale. We don't have decades to get them commercialized. We have years. This week I'm speaking with John O'Donnell, a founder and the chief innovation officer of Rondo Energy, on building heat batteries to decarbonize industrial heating.

John O'Donnell: It's been true that the industrial sector is the hard-to-decarbonize sector. There were always solutions, but they were more expensive than burning fuel, and no one wants to be the high-cost producer of a low-margin commodity.

Lara Pierpoint: Today, producing heat for industrial processes is very dirty.

John O'Donnell: A quarter of all the fuel we burn in the world is for industrial heat.

Lara Pierpoint: Most often, heat is delivered as steam that comes from a boiler running on coal oil or natural gas.

John O'Donnell: A little more than half of all the world's industrial heat is delivered as steam because steam is wonderful, energy-dense, easy to move around.

Lara Pierpoint: But the rise of cheap and abundant renewable energy in the last few years has finally opened up an opportunity to serve that heat using clean electricity.

John O'Donnell:

We now have a new fuel that is cheaper than the fuels that industry used. It's intermittent solar energy, some places intermittent wind energy, and it is intermittent electricity that's becoming the cheapest source of energy humans have ever had available. We have now entered this world where on a primary energy basis, you can have electricity at a lower cost than burning fuel.

Lara Pierpoint: So John and Pete decided that they would build heat batteries that turn renewable electricity into steam, and in 2020 they founded Rondo Energy. Their technology is actually based on a process developed in the 19th century.

John O'Donnell: Back in the 1850s, a coal-saving invention was devised that stored heat in brick at a 1500 C, capturing heat from the furnace and delivering it back to the inlet here. We found a way to use that material in combination with the same electrical heating elements in your toaster. We had a physics insight in a structure that allowed us to combine those two very pedestrian things in a fundamentally new way that allowed us to build a storage unit out of stuff that we know will last forever, that charges very rapidly and then delivers heat at the same temperature constantly, all year long. And those are the premises that you need for a heat battery, and we build industrial-sized units. We build units that are approximately the same size as the coal and oil and gas-fire boilers and furnaces that we're replacing.

Lara Pierpoint: By 2022, John and Pete had figured out how to build the battery structure itself, and they had deployed a couple of batteries at a small scale. This included a 40-kilowatt-hour project and a 200-kilowatt-hour project. Finally, they were ready to go bigger. So they approached a longtime partner, Calgren Renewable Fuels, about building a two-megawatt-hour battery in their facility in Pixley, California. Calgren produces ethanol and gas using corn and cow manure from local farms, and both processes are very energy-intensive.

John O'Donnell: We started exploring with them quite some time ago how to use renewable energy as the source of energy to drive all that distillation, to drive the whole refinery, and in fact, to do cogeneration of heat and power to replace all of the purchased grid electricity and purchased fuel. And when we started Rondo, we had a good understanding of what the problem was. In fact, I had failed previously to come up with something that Calgren found acceptable using concentrating solar technologies.

Lara Pierpoint: Unlike previous attempts to use concentrated solar and thermal storage, the heat batteries could actually produce steam the same way that Calgren's existing gas powered boilers do. John and Pete thought the familiarity of the process would make it easier to sell. So they pitched Calgren on letting them try their two-megawatt-hour project.

John O'Donnell: They said, "Great, but show me. I don't want to be the first taking a technology risk. I want to know, is this going to operate in my location? Is it going to screw up stuff with my refinery team? So one way or another, we'd like you to show us.

Lara Pierpoint: And that's what they set out to do. I talked with John about how they designed bigger bricks, secured commercial-scale manufacturing, and what happened once they started stacking the bricks at Calgren's facility. It turns out designing and manufacturing bigger batteries was just half the battle. So you've built a 40-kilowatt-hour project, you've built a 200-kilowatt-hour project, you know that you need to build this two-megawatt-hour project. But as you said yourself, Calgren is saying, "Show me." They don't want to be the guinea pig. They want to be the one who's using something that is well-proven, and yet you're coming to them and saying, I need to build this as a proof point to get to something bigger, and I'm going to be trying some new things here. So how did that relationship work? It sounds like you'd known them for a while. How did you talk to them about the degree to which this was already something that you knew how to do versus this is something where you're really proving out some new pieces of the technology?

John O'Donnell: Step one, this was a pilot for the... The full-size installation will be about 600-megawatt-hours. And again, they recognize, "Look, that's too big a step." But the short answer and the thing that's true with every customer, I think, once they understand what a Rondo heat battery is, all the engineers, like, "Okay, this thing makes steam exactly the way we make steam today. We make steam in exactly the same heat recovery steam generators that we build in by circulating air. We blend it down to 650 C." So when an engineer looks at, okay, what is this thing versus what do we do today? Oh, it's fundamentally the same, and oh, and there's nothing in the box except brick and iron. There's nothing that is capable of causing any fire or gas release or whatever, because we are being installed in the middle of their refinery. Safety is the number one matter.

Okay, and now let's do a walk-down of all the design things, how we're going to connect. So there was technical diligence on their side, and then we chose with them a particular way of interconnecting to their process so that they put in the interconnecting flanges and the circuit breaker for us during a ordinary facility maintenance. Rondo has not caused five minutes of outage for Calgren since we started working there.

And then the last matter was let's create a commercial structure where we're putting our money where our mouth is. We're not asking you to take any technology risk. We're going to sell heat, we're going to buy electricity, and so we'll prove out the foundations of the commercial model.

Lara Pierpoint: Sounds like a couple of really key elements here. One is that you had an interface that they understood and that basically was the same as what they had dealt with before. And then also the piece that belonged on your side of that equation was not scary, it was not crazy. As you said, it was not something that posed any safety concerns for them. And then it sounds like the third piece really is this model for how you're... You called it storage as a service. So say a little bit about that. Do you think of it as storage as a service or heat as a service, or how exactly would you characterize the way you were selling a product to Calgren?

John O'Donnell: What we are doing is opening a completely new market for renewables where we're going to build utility-scale generation. The Calgren project will ultimately be well over like 150 megawatts of generation. But it may not be sensible to finance that generation at the same finance rates that a much newer technology, a juvenile technology needs. And so separating out storage as a service with its own financing package from the electricity can be beneficial. Now we have utilities that we're working with at scale in Europe who are combining these things to offer heat as a service to a customer. But of course when we offer heat as a service, that is what's the input electricity cost, what's the storage cost? And we're seeing enormous creativity among European utilities in bringing the power trading desk and the utility scale team and local construction teams together to come up with an electricity product that they're going to internally roll into a heat-as-a-service product.

But from our standpoint, there's really not a lot of value in Rondo structuring solar or wind projects. There are others who are better at that than we are, but our structuring and dealing with all of the matters of offering storage as a service, among other things, is maturing our team to be able to work in depth with others who want to offer storage as a service. It's proving the technology out. And maybe one of the most important things, which of course you're doing right here in this moment as well, is educating the market about this new opportunity to put capital to work at scale to build infrastructure.

Lara Pierpoint: Well, let's jump into this because the next step, of course you've got agreement. So how do you finance this thing? And let's actually start with Calgren's piece in that, because presumably they're offering to pay you for the heat that you're providing in this particular facility, or I should say for the storage as a service. What was the actual contract that you signed with them? How was that structured in addition to whatever else they provided in terms of support, and as you say, the kind of physical interface?

John O'Donnell: Because this one was the size that it was, we financed this on balance sheet. One of our board members says, "Look, it's easier to raise $3 billion than to raise 300 million. Get to work on the full-size project." And that's true from a projects to finance standpoint. If you want the A-team of the bankers, it's got to be worth their while. They're going to be putting money to work. Small projects, unless they're part of a big project portfolio, are very difficult to finance this infrastructure. And the other matter is, look, this is the first commercial one that we're doing. No one's going to want to do it. So we did that. First of all, we did that on balance sheet, and then second, because it was on balance sheet, one of the biggest issues in our particular business is, and you might call it fear or volatility.

When someone signs a power purchase agreement for solar electricity, or electricity as a service including a battery, they have a fairly good expectation that no matter what happens, electricity rates always go up. Whereas if I'm replacing combustion of fossil fuel, I mean during my career in solar and renewables, I've seen natural gas be $14 a million BTU, and $60 a million BTU in Europe, and I've seen it be $1.80. And structuring long-term off-take agreements for heat as a service, the buyer would love to be indexed to a fuel price. The finance party would not. It's like who's carrying what risk of volatility? And one often winds up pricing against fear of floor price of where fossil fuels might be versus pricing against an expected value. Here, just for simplicity, because it was on balance sheet, we said, fine, we will sell you heat at a discount to whatever the then gas price is.

Lara Pierpoint: So you've got an off-take agreement, you've got a plan to finance this on balance sheet through money that you're going to raise yourselves. And now you're starting to get into construction. One of the first decisions that you were facing, you mentioned this earlier, is about the size of the bricks that you're going to use. So can you say a little bit about what the challenge was there and the decision that you had to make?

John O'Donnell: We had a structure, but we were implementing that structure with individual cast bricks that each weighed like 50 kilograms, but then we realize we store a little bit more energy per kilogram than a typical lithium-ion battery pack in brick, but we're going to build a hundred megawatt hours. It's a lot of kilograms. And a 50 kilogram brick is almost perfectly the wrong size. It takes two people to handle it, and there are a lot of them. So to the extent that we could make larger single bricks, we could reduce crane operations or people hours building these things, and we have a trade-off of manufacturability and yield and construction labor building the thing. And we know that eventually we want to be using much larger things. We eventually want to pick up multiple tons with a single crane lift or lift as we're putting these things together. We also have tooling to choose. Are we going to use forklifts or telehandlers or hydraulic cranes, or how are we actually going to be constructing these?

And we made a decision that said, instead of building Calgren with 50-kilogram bricks and then deciding that we're going to build the next one with one ton bricks or half ton bricks, let's take that hit now. Let's make sure that we build Calgren of exactly the same brick and everything else as we build the large one, because otherwise we've got more kind of scale-up risk that we had not covered off. To the greatest extent possible, we wanted them to be the same except one of them has 10 tons, one of them has 500 tons of the same. And we wound up taking about probably six months longer than we thought it was going to take, learning how to make these large, much larger bricks that have fine structure that is not at all common. Almost everything is common, but the shape was unusual. And we were particularly blessed and helped that one of our series A investors, Siam Cement Group, SCG, out of Thailand, who are a diversified producer of paper, petrochemical, cement, and refractory brick doing business around the world.

We began working in depth with them about... We're doing pilot manufacturing. How are we going to scale up to high volume manufacturing? We want to explore with your team how we will make these shapes. So we did pilot manufacturing at the same time as we were manufacturing and building Calgren. We were also validating the handoff to large-scale manufacturing. So the journey that said we're going to try to build these the same way, we actually built an unbuilt Calgren several times, trying different tooling and different sequencing so that we had a reasonably firm base of experience of some alternates when we went to build the next one.

Lara Pierpoint: Okay. I have a really critical question at this stage. As you're manufacturing these 500-kilogram bricks, are you still using PAM in the molds?

John O'Donnell: No.

Lara Pierpoint: Okay.

John O'Donnell: We're not using release compounds. We found materials and systems that don't require release compounds, because it's not scalable and it's not needed.

Lara Pierpoint: Okay. But so in October of 2022, you're breaking ground, and you are now building Calgren. So talk about that process. Were there any major challenges you faced in the process of construction?

John O'Donnell: Not that were not self-inflicted.

Lara Pierpoint: So Rondo had sized the bricks and secured the manufacturing, but now, they had to actually build the battery, and they would learn a lot about construction. We'll hear about that after the break.

John O'Donnell: We learned lessons that had to do with, well, okay, let's look at tooling and handling that a different way because we're not getting things to assemble the right way. Oh, there's some manufacturing tolerances in the brick that we need to be looking at. And when I say self-inflicted, we were in this first unit not hitting tolerances in our own manufacturing of the brick mechanical dimension tolerances that we had established were the manufacturing tolerances, because we were in... like we're going to build it even though the brick is slightly out of spec, and we learned some of the reasons why the tolerances mattered.

Lara Pierpoint: And so, also, as you're undergoing construction here, you actually worked with an engineering procurement and construction firm. So tell me about that decision versus deciding to serve as your own EPC.

John O'Donnell: In these early projects, we are prime, we are the prime EPC. And in the set of projects that are in contract now, we have projects where we are not the prime EPC, where we are a component supplier. We have projects where we have a single sort of master subcontractor, single master contractor. We have other projects where we are doing more managing separate trades. And I know the answer, "It depends," isn't a very satisfactory one. But I will say I think a lot of what Pete and I lived through in our previous company where building a series of things that went from rooftop to something that fit in an office to something that was 200 kilowatts to seven megawatts, to all the 200 megawatts, and building things 500 miles from town in the desert in Oman, we learned an awful lot about contracting strategies and especially about learning.

Both on the manufacturing side and on the execution side, it is critical with novel technologies to be learning. Pete in several companies previously recognized it's important not to automate manufacturing processes too early, because you may lock yourself into a design that you didn't like, that, okay, and now you've made it hugely expensive to change that. Similarly thinking that you're going to give a completely cooked book of IKEA instructions and someone else is just going to do it without your supervision, you're going to get there, but the way you're going to get there and have that book be good is by having seen it and done it yourself and then work as rapidly as you can to move from time and material subcontracting to fixed work packages and work really hard on moving to that how do we make the whole thing a standard kit that we can give to anybody. But again, just if you want to do manufacturing, scale up manufacturing, you need to do some yourself.

Lara Pierpoint: So it's early 2023, you're nearing completion of construction, you turn the unit on, and you get first heat. Can you explain what that means and what that felt like?

John O'Donnell: The way things roll, it was a little bit anticlimactic, because of all the subsystem testing that led up to that moment. It's not like launching a rocket. This is a really very-

Lara Pierpoint: Probably that's a good thing,

John O'Donnell: ... simple thing. Yeah, well, I suppose that's right. Yes. And I will say that within about three days, there was a real moment of triumph, because one of the things that we did not mention earlier that was critical about this system was validating the computational models of what goes on inside and seeing the observed set of temperatures within the system as we charged and discharged it. One of the members of our design team in his last job was working on Mach 6 hypersonic missiles, computational fluid dynamics. Now he's working on computational fluid dynamics at six miles an hour, and it's just as hard. These things are really simple, but they could not be designed without modern multiphysics simulations of exactly how does all the gas and heat transfer work, because we'll either not charge properly, we could have brick that was deteriorating because it wasn't being heated evenly, or especially, we wouldn't have the same storage capacity. We would not achieve this thermal climb that I mentioned earlier, and that's about the gas dynamics.

Three or four days after we turned Calgren on, CTO was distributing this picture that, again, it doesn't look like much, but it's a picture of temperature sensors, thermocouple measurements, and what the computational model said. And the fact that those matched was just a wonderful moment.

Lara Pierpoint: I was going to say, these are the moments we live for in climate. They're the moments where the graphs show exactly what we're expecting or ideally something better. So I'm with you on this. That's exciting.

John O'Donnell: Yeah, we were actually slightly above our planned storage capacity because we had sort of haircut that a little bit.

Lara Pierpoint: That's very cool. Was that just because you were being conservative on your models or was there a reason that there was actual upside there?

John O'Donnell: Well, actually, way back at the beginning when we were doing modeling, we didn't have enough computer power to do all the multi-physics stuff, and we had taken some shortcuts in modeling and we built the early prototypes that were better than with the models, and it took a while to figure out, oh, we have to run this other stuff that makes their model run a hundred times slower. Okay, now we get match. We had sorted that out. So I think it had to do with assumptions about exactly what are the sizes of the air passages, what are the average yield of those that affects the temperature distribution as you discharge the core. We did slightly better in storage capacity. I can't even remember why. There was a mix of several reasons, but the biggest piece was that we had been a little bit conservative

Lara Pierpoint: And what's the state of the project now? How are things going?

John O'Donnell: The unit's been in a relatively continuous operation, but it's taught us a bunch of other lessons too. When we started the conversation, we heard, okay, we turned off the refinery once or twice a year for overhauls. Other than that, it's all dead-level temperatures, flow rates, nothing changes. So okay, here's how we're going to operate. None of those things were true. And there was one element of this particular unit that required manual intervention to start it up. There's an emergency safety thing and we required someone to go physically there and pull a rope to enable the unit to start up, and we thought, "Oh, it's going to do that twice a year. There have been dozens and dozens of times that we needed to have somebody there." So there were lots of lessons about, oh, look in a real world refinery, all kinds of things. And yes, there's a nominal operating condition, but we have now exercised the whole box of temperatures and emergency and non-emergency shutdowns. That's also been quite valuable.

The other matter is it's been a place for independent engineers, owners, engineers of customers, and an IE that we engaged to go look at a real operating unit, look at its operating history, and obviously it's a pretty boring thing to look at a box, but it's also a pretty riveting thing that it's not PowerPoint and let's look at the control system and let's look at the archive of all the measurements.

Lara Pierpoint: I feel like boring is good when you're talking about industrial processes, right, and all of the steadiness that you need to provide there. And as you say, it's also really exciting to have something real in the field. So you've built this two-megawatt-hour project, you're now building 100-plus-megawatt-hour projects. Are you maintaining pretty much all the same technical specifications around the bricks and some of the basic elements of the plant? Is there anything new that you're putting into place with respect to materials or construction processes or anything? How are you thinking about this next scale up step?

John O'Donnell: There are no changes in the materials. The materials that we used in making these first things are the materials that we're using in all the current things. One of our independent engineers said, look, you're using stuff that's 10 times stronger than it needs to be. That's on purpose. There will be a period where we're doing value engineering and bringing things down. But being conservative so that anything that we had not anticipated, there's a very large safety factor, is part of the current design philosophy.

There are things we learned, as I mentioned earlier, about tolerances, and how to design around and create greater flexibility for tolerances that there were some improvements. Those were particularly not the brick, but there's a system by which the electrical heaters are suspended in space with ceramic, they call it kiln furniture. They're ceramic supports. And there were several iterations to make it easier to install and make them more reliable, and make them more manufacturable. So there have been things, lots of small component things. And then the hundred-megawatt-hour units are a different system design. They're made of more of the same brick, but everything about air flow ducts and boilers and all those things are different. So there are system design things that the system around the thermal core, and there's a lot more of the thermal core, but deliberately the bricks, the heating elements are the same. There's just more.

Lara Pierpoint: Well, I am really excited to see what happens over your next couple of projects, and hopefully to have you back on the show to talk about how they go in maybe another year or two. But in the meantime, let's talk about this for the community. Are there any decisions related to building your first projects that you might've made differently knowing what you know now?

John O'Donnell: It's easy to say that we would have probably raised more money earlier and tried to build a larger engineering team earlier. Some of these challenges in just the system level engineering, the product level, were larger than we expected. We had expected to outsource more that we ultimately wound up insourcing, particularly in structural engineering and some of the product mechanical design engineering. What we are doing is very, very similar to what industry does today, but for many of the industrial partners that we work with, they're really good at doing the same thing again and again, but the amount of change that we wanted, actually we had to get much more deeply involved in outside things like boilers and electrical controls that we had expected they'll be able to handle all of that. There are some point things like that, but I think broadly, the raise more money than you think you're going to need and expect it to take longer than you think it's going to take, I've done six startups, that lesson has been the same in every single one of those.

Lara Pierpoint: Yeah, that sounds about right. Okay, last question. If I could wire a hundred million dollars into Rondo's bank account tomorrow, what would you do with it?

John O'Donnell: Right now, we are building out the execution organization to deliver these projects in Texas and Kentucky and Denmark and Germany and Siberia and Southeast Asia and California, and growing that team and also completing the contracting on the set of teams that the execution side and in fact growing the engineering team even further, those are the top priorities. And as it happens, we are raising money right now, specifically for those purposes.

Lara Pierpoint: John, thank you so much. This has been a great conversation. Really love that we've gone from cooking oils to ways that you're revolutionizing industrial heat and energy. This has been really, really great to hear, and I think really informative for our listeners, so thank you for joining us here.

John O'Donnell: Yeah, thank you so much. No one else has brought me back to that moment. To look forward, nothing can go to this speed and this scale, right? This is just this giant opportunity. The things line up and yeah, it wasn't always clear right at the beginning whether that would be true or not. Thank you so much.

Lara Pierpoint: John O'Donnell is a founder and the chief innovation officer of Rondo Energy. The Green Blueprint is produced by Latitude Media in partnership with Trellis Climate. The show is hosted by me, Lara Pierpoint. Our producers are Erin Hardick and Daniel Woldorff. Anne Bailey is our senior editor. Sean Marquand is our technical director. Stephen Lacey is our executive editor. If you'd like to suggest topics or guests for the show, send an email to editorial@latitudemedia.com. You can listen to The Green Blueprint at latitudemedia.com or subscribe wherever you get your podcasts. And if you have a fellow clean energy or climate tech traveler who would benefit from the insights in this show, send them a link. This is The Green Blueprint, a show about the architects of the clean energy economy.