The Exchanges

Q about what proven means, but not de-risked in the eyes of infrastructure investors. So what does that look like? I mean, give me, you know, representative imaginary technology, like how much do you have to do to be at the point where you are quote unquote proven? And then where is the, like, how does an infrastructure look at That same thing and see it as unfinanceable at that point.

A Yeah, um, let's start with the end part, which is how do infrastructure funds work? How does a project finance bank work? And the reality is that infrastructure funds just do not want to lose money, right? There is, there is 70, 60, seventy billion dollar infrastructure funds that have a 15 year track record that have never lost money on any project they've invested in. Right? It's a huge contrast, obviously, with the VC world, where you can lose on nine and make it up in the one, the 10th one. Um, so the biggest thing here is the fact that there is such a rigid criteria for infrastructure investors to invest in, right? You have to have a long-term track record for the technology. You have to have hours of operation. You have to have a really solid EPC, um, Um, you have to have it long-term contracted, um, and you have to be able to not only invest in a small one-off project, but you have to be able to invest hundreds of millions of dollars from that fund, um, into that technology. And I think that's where, you know, a lot of these new technologies don't make the cut, right? Um, so what we try to do, uh, in our work is how do we shepherd These technologies from the point where, again, they came out of a lab, and they've done a pilot. And out of that pilot comes a lot of data, and comes the fact that the original scientific thesis has been proven. XYZ reaction happened, and you …

AI assessment note: “out of that pilot comes a lot of data... scientific thesis has been proven”

Q risk and so on. And so in their mind, ideally, they want to send the shortest term contract that they can to allow you to build the project. You want the longest term contract you can get from them for certainty. And so there's a, there's some middle ground there. I don't know where it sits. So how do you think about how, how long the offtake needs to last?

A So two things, it does depend on the product, right? So we've seen there are some products that we were involved in structuring the first 10 year ESAF offtake, uh, you know, fully bankable offtake. Um, and we were able to get it 10 years because, you know, The airline, uh, American Airlines was very motivated to have access to future product, right? They see that demand supply, uh, imbalance, and they want access to future products, so that's the reason why many people sign those long-term agreements. Um, in a space like cement, where cement is such a spot-based, uh, commodity, um, by the way, so is jet fuel, but again, jet fuel, the airlines can take a, a longer term view, but in cement, Where it's so short term, it'd be really hard to get anybody to sign more than a five-year agreement, right? However, again, you, with the right pricing, um, and the right motivated customer, you are able to show that, hey, without this, I'm not going to be able to raise the capital that I'm going to need for the plant, right? And I think that, you know, you don't necessarily have to go 20 years, um, and, and yes, I would argue that You know, investors will give some credit to the merchant piece, but you also can't do one or two year off takes. Uh, it's just not helpful.

AI assessment note: “it does depend on the product, right? So we've seen there are some products”

Q demo is end-to-end representative, right? So, like, you're running the full system, you're producing the product you're ultimately going to produce, whereas the pilot, maybe you are demonstrating the reactor performance or something like that, but you're not running the full end-to-end system. Is that right? And, and also on the demo, I mean, you mentioned another thing that people want to see, which is runtime, like, how much runtime?

A Yeah, no, it's, it's, it's exactly, that's exactly right, which is, You have to be able to create, have your whole, your whole team understand how you're going to design a demo. How do you build it? Most of the time, the team has to build it themselves. No EPC contractor really wants to get into this smallish projects, high risk projects, right? You have to be able to understand how to operate it. You know, what is the product that's going to come out of it? And is that product to spec? So a lot of the demos, what they Go to do is to prove out the specs inside the, you know, factory or the production processes in the end customers, right? Um, and then when you look at it, where a lot of companies, you know, don't necessarily, uh, follow these steps, what happens is they don't have an end-to-end system, and I think that's super important, right? What we know is At the end of the day, the investors that come in at the Series C, Series D, you know, most of the time funding or partially funding, um, first of a kind projects, they really want to see six months or more, or performance, uh, because things degrade if you continuously operate something in a way that you didn't do it on the pilot, things degrade in a way that you really have to learn from that, And figure out how you're going to engineer your first of a kind to make sure the performance is there. And that's why that six …

AI assessment note: “they really want to see six months or more, or performance”

Q or sells plant, I should say, whatever the product is, that's a balance sheet purchased by a customer. Rather than offtake. That's attractive, obviously, in the context of like, you don't need a 20 year offtake. You have a customer who's just going to take it off your hands. How do you think about the trade-offs there? And how much of that do you see in first of a kind?

A We've seen, we've seen some of that. Uh, the issue with that is you have to finance the construction, right? And if the customer never buy, one is you have to make sure the customer buys the plant when it starts operating at a certain performance levels. Um, so that you have to be making sure you're, you're very, you know, tight on the agreement on that. Um, two, How are you going to get that construction financing, which is the biggest thing, right? Um, there's different ways of doing it. Again, you're, you're, you know, you can base that on the fact that you have a hundred percent certainty that if you perform, the plan will, you can put the plant onto the customer. Um, however, where there is great lines or where that is in question, that's when you, what's the backup, right? So a lot of, we've seen a couple of cases is people actually signed a long-term off-take agreement, And that off-take agreement has the, the buyout option, right? So you still cover as a company building the plant, you're still covered from the perspective that if the customer doesn't buy it, um, you know, you could still have some sort of off-take, but at the same time, they have a buyout right, which at some point, if it's, it's going to be an economical decision for the customer to say, look, I can just buy the plant and rip, uh, up the off-take agreement.

AI assessment note: “We've seen, we've seen some of that. Uh, the issue with that is”

Q if the unit thing Is the same size as the ultimate unit thing, what you're building is a modular version of the thing, it's viewed as less risk. And I guess my question for you is, do you see that as being, is that an optical thing, or is it a real thing? Is there actually lower risk, if it's modular, or Is it just what will convince infrastructure investors?

A It is real. Um, and for us, it's about the execution risks, right? It's about the engineering, design, the procurement, and logistics of it, um, right? Large-scale construction has so much more risk associated with it, and then construction, operations, and as you said, the capital race, right? Um, the way, the way we think about it is numbering up Um, will require you to, and you approve your unit, and now you're adding numbers to it. Integration is not an easy thing. So it's not a panacea for all your problems, but it does require, it does alleviate a lot of this execution risk around just large scale projects. Um, I've been involved in my lifetime on putting together code generation Projects where, you know, you're using a G turbine that has been deployed thousands of times, and still a lot of things went wrong in building this mega project. So, again, building a billion dollar project versus a hundred or two hundred million dollar project does have real risks associated with it, um, that, you know, we need to think about. Now, many people in your audience are going to say, well, modularity doesn't work for everything, and it's true. It's there, you know, things such as reactors where bigger is better and bigger does get you to down the cross curve a lot faster. So where you cannot design for modularity on everything, you should at least try to find the design for modularity…

AI assessment note: “It is real. Um, and for us, it's about the execution risks”

Q secured, all the things you need to have. You've got, you've got all the boxes checked. What should you be thinking about in terms of the capital stack on the project? What is realistic? Is, is there a world where you can get debt on a folk project? Does it basically never happen unless it's like government debt? I mean, you know, what, what are you, what are you targeting?

A Project level investment or financing, as you mentioned, is, you know, even with a positive IRR, as I said, it's challenging because of that risk return. Mismatch there. So what you really have to do is you have to bring capital at different levels of, of your stack in order to be able to fund the project. A lot of times companies do raise TopCo equity, um, and most of that TopCo equity goes to fund the project, right? Um, I think that is, that is one thing that, that we've seen done. Um, some of the company putting TopCo equity do want to take a bet on the project level, uh, and therefore they put Money into that. Um, as far as debt goes, we haven't seen it. We've seen people be able to raise debt at the corporate level, um, via, via venture debt or via corporate debt. Now, those are very low leverage numbers, right? They're not the 80, 95% that, you know, you will get in solar. You will get very low levels. But again, our contention is that if you structure your project And the construct, uh, you know, the contract, contract or structure of your project correctly, you should be able to put some sort of debt, very low level of leverages on it, because as I said, you have hopefully protect the downside, and if you do have an infrastructure investor coming in to your project, um, it's, it's not a far leap to say, can a bank put 20, 30% leverage on it? Uh, again, that is the theo…

AI assessment note: “As far as debt goes, we haven't seen it.”

Q You talked a little bit about CapEx estimation. I'm curious where you've seen people get that wrong. When people underestimate how much a thing is going to cost, is there any consistency to what they have underestimated?

A No, there isn't. Um, again, the reality is the EPC CapEx world changed dramatically after, after COVID, right? Um, things that used to be very secured, large scale EPC companies that used to be able to give you a, you know, lump sum turnkey contract are just not doing that anymore in, in conventional technologies, right? Let alone this kind of technologies. So the CapEx blowouts that we see a lot of times is You know, again, as you go through the process, fail one is a plus and -50% sort of, ah, number, and as you go through it, there's a lot of things that the company didn't realize they needed in order to build the plant. Um, so one is, again, The, quote unquote, lack of knowledge around what it takes to build a plant. Hence, the demo is a, a mitigant, a large mitigant for that, but you still have to go through the process. Um, one of the things that we see on, on one of the projects that, uh, we committed to that, that was canceled was, you know, the balance of plant. The integration of four systems where the price was very well known in those four different systems, but the integration and the actual, you know, The, the, the steel and the cement that went into integrating those four components just was not sized correctly because nobody had ever done before. So balance of plan integration, as I said, around the modularity is still, it's still an issue. It's still, you have …

AI assessment note: “No, there isn't. Um, again, the reality is the EPC CapEx world changed”

Q All right. So I guess final thing, we've talked through a couple of these specific examples, but I think folks will be interested to hear more. So can you give a couple more actual real world examples of first of a kind projects that you guys have gotten involved in and just a little bit about how they're structured and what they enable?

A Sure. Absolutely. Uh, and a lot of it has been collaboration with, with you guys. So I do think that, you know, it goes to show that this world is not going to be conquered, uh, that value of death will not be conquered with specific, you know, VCs or infra funds or catalysts by itself, but rather the partnership. That, that's the main thing that we have learned from this whole thing. Put aside the 12 keys. If you can't really partner with the investors, with the companies, with the future investors, um, it's really hard to get this done. Um, a couple of examples of that are, you know, Rondo is a great example where, you know, a year and a half ago, um, you know, we talked, we spoke to Rondo and they had, you know, more than a dozen projects around the world that they wanted to execute on. Uh, but there wasn't, uh, you know, there wasn't a specific clarity around, uh, What is the commercial product that they really wanted to put out there, right? There were customers that wanted to buy the machine, there were customers that wanted them to build the machine, and then, you know, there were buyers from them, there were customers who just wanted the product, the steam in this case, and what we really focus on with Rondo is, where is it that you can deliver, um, you know, this, uh, your first of a kind projects, uh, in this case, uh, three projects that we funded, And what is the be…

AI assessment note: “Rondo is a great example where... we created a new commercial model where we called it Steam as a Service”

Q FEL one, FEL two, maybe my feed. So I feel like I have pretty good visibility into what this thing is going to cost, and I have offtake. What should I be targeting in terms of pick your metric, IRR, payback, et cetera? What, what is, what is going to be good enough to attract some kind of external project level capital to the, to a first of a kind?

A I think the answer is not a single number, but rather it's a combination of factors that get to that de-risk point of view from the infrastructure investor, as opposed to a hard, solid number. Because The reality is the risk associated with first of a kind, uh, plans most of the time will never compensate, you know, on a number, right? So you have to start thinking about the commercial construct around, um, around the project that will give the infrastructure investor a couple of things. One is very high certainty that the plant is going to finish construction and, um, And two, they really have to take a point of view around the downside. As I said, infrastructure investors do not like to lose money. So how do, how is that downside protected? So what you have to do is start putting together all of the risks around the contracts, right? Is your feet stock the same length as your off-tick, right? Did you procure the right PPA? Are you, are you exposed to PPA risk? In a lot of the long, Uh, long duration energy storage plays. That arbitrage, uh, you know, on, on negative price power might only be for a couple of years. And, and the utility of the area might only be willing to offer it three years because they know there's a new transmission light coming in, right? So how are you exposed there on the PPA side? Um, on the feed, on the offtake, again, you know, what are the obligatio…

AI assessment note: “the answer is not a single number, but rather it's a combination of factors”

Q space, and it's, it remains, you know, the sort of like quintessential challenge, I think, in the development of new technologies in, in the climate world, and there's like so much to unpack around it, but I want to start with This, which is, um, how would you describe why it's hard? Of course it's hard. Why is it hard, though, to figure out how to finance a first-of-a-kind thing?

A Yeah. Um, no, look, really good question, and, and I do think this is the main challenge of, of climate tech. How are we gonna get this technologies where people have spent, you know, hundreds of millions of dollars, years, decades developing, how do we get them scaled up, right? How do we get them to the place where solar And Wynn and offshore Wynn are. Um, and it's exactly why, you know, Catalyst was created, and, and it's not just me, by the way, it's a whole team of people that are trying to, um, to tackle this. Um, the challenge is the fact that these companies reach a point where the technology has been proven, but it has not been the risk in the eyes of large scale infrastructure capital, right? Whether that is Infra funds or project finance banks, right? So you need to find a way to get these technologies over that, over the, from the proven stage to the de-risk stage. Um, and that's where the expertise that Catalyst has, the infrastructure, energy infrastructure, investing, development, uh, and construction expertise, um, coupled with the flexible capital. That's Where, you know, we can play a role in this.

AI assessment note: “proven, but it has not been the risk in the eyes of large scale infrastructure capital”

Q between, okay, you have a customer who's going to buy your product versus going to buy your system and own and operate it versus the other ways that you can engage with customers. And are there You already described one thing, which is don't license before you build. Build first if you're going to license, license later. But beyond that, what are you looking for in, in the commercial construct?

A Sure. Um, let's spend a little bit of time talking about the demo because I feel you, you have it exactly right. Let's just get on with the demo and prove it. Uh, many times though, people think they're going to make money off the demo, right? When, in that, in our experience, we found that's not the case. The case is That most, in most cases, the demo is a very large R&D exercise. So money losing, unprofitable R&D exercise on every aspect, including the commercial aspect of how are you really going to market this? So most of the time, demos don't really get off the long-term off-take contracts because no customer really wants to take that risk. Um, and the company doesn't really feel comfortable signing up to something, to delivering something that they don't really know, they've never really built.

AI assessment note: “demos don't really get off the long-term off-take contracts”