The Exchanges

Q have seen us already starting to approach a ceiling in terms of performance of the existing chemistry. So, like, What was the chemistry at that time? And then I guess bringing us forward to today, like how much have we seen performance improvement relative to that time? Have we broken through the ceiling by shifting chemistries or whatever? Or in your view, has the improvement to date been like marginal?

A Yeah, it's, it's really been marginal when you look at the chemistry level. So back then the chemistry was lithium cobalt oxide, uh, and, and graphite. And that really was the, I mean, that was the chemistry from 1991, all, you know, all through the first, through the first Tesla Roadsters were built on, on lithium cobalt oxide chemistries. And, um, and, and cobalt's actually the highest performing of the oxide cathode. So you have your cathode, right? We talk about high nickel cathodes. We talk about LFP cathodes today, but cobalt cathodes are actually the highest performing. And the reason we're on nickel is because it's meaningfully cheaper. And the reason we're on iron is it's yet cheaper still. Uh, but both of those are, are actually lower performing, uh, than, than, than the cobalt-based cathodes.

AI assessment note: “it's really been marginal when you look at the chemistry level.”

Q question is to ask it, Relative to when you started Sela, like what was your, so this is, we're now in 2010 ish. What was your view at that time on like, what was the, what was going to happen in the EV market and then where there was going to be a gap that you thought needed to be filled? Like what was the founding thesis of Sela basically?

A So I think, I think there were a couple things. One was that, you know, so 2011 when we, when we incorporated There were about 2003 thousand electric cars sold worldwide, right? Not 100,200 thousand, but, you know, just 2003 thousand. So, um, you know, we're, we're, we're talking, uh, you know, almost no vehicles. It was a novelty, but we certainly, uh, all, all the early folks at Tesla had the belief that all cars were going to go electric. Um, and, and I think that looks sort of nutty in retrospect, but it's kind of obvious today. Um, and, and the thing that made that Possible was the improvements in batteries leading up to 2004 to 2008, right? Really starting in 1991 as lithium ion got much, much better in performance and the costs dropped a lot. The gap that I saw was that, uh, lithium ion was already hitting its theoretical performance limits. So the high kind of, um, you know, the high energy densities that, that drove down the cost, uh, you're starting to really reach those limits and the scale was already really large. You know, um, There were maybe 50 gigawatt hours of lithium ion produced that year. So my, my view was that if the chemistry didn't improve, uh, the performance would stop going up and the cost would stop going down at some point in the near future. And you would be kind of left with, with a gap where, you know, systems were cheap enough for many, many ty…

AI assessment note: “The gap that I saw was that, uh, lithium ion was already hitting its theoretical performance limits.”

Q Can I just pause for 1:02? Can you define performing? Like, what metric are you referring to?

A The most, so my view, the most important metric for EVs is energy density, defined as watt-hours per liter, volumetric energy density. You know, people sometimes mix up watt-hours per kilo, call that specific energy. Um, Watt hours per liter is the most important, and the reason watt hours per liter is the most important is it translates to how many cells you need to build a certain capacity pack. So back when we were using little tiny, 1806 fifties in the roadster, you know, it had 29, 6900 cells, uh, or so, but if, if the performance was two X, if your energy density was two X, then you'd only need 3400 cells, and it would be a lot cheaper to make those cells and a lot cheaper to package those. So I think about performance about energy density. And then, because a lot of your costs are fixed on a per-cell basis, right, you, you need to, you know, to make a single cell, no matter what energy it stores, you have a lot of fixed costs. The higher Amount of storage in a cell, the lower the cost goes in general. Um, and so, yeah, so performance and cost are very related.

AI assessment note: “the most important metric for EVs is energy density, defined as watt-hours per liter”

Q similar story to the LFP story in the sense that it is, it is less performance in exactly the metric that you're describing, but also potentially, and this is, I think, to be proven, but Theoretically it could be even cheaper than LFP and then has these side benefits of like, oh, you avoid like the lithium, uh, price volatility and stuff like that. What's your view on sodium ion?

A I think sodium ion makes a ton of sense for the grid. I don't think it makes any sense for EVs. Uh, you know, we're, it's hard enough to convince consumers to buy a 200 mile or 250 mile EV with LFP. It's going to be really hard to convince them to buy a 100 mile EV. I feel like that's, that's really kind of going backwards, even if it's, even if it's a bit cheaper. So, uh, you know, it's not, there may be markets in, in the developing world where, you know, a hundred miles or 50 miles is totally fine because you're using it for, you know, say a daily route of delivering, uh, you know, delivering food or, or something like that. Um, and, and you're just looking for the absolute lowest cost, but, but I think LFP serves that need pretty well. And we are so Over capacity in the world on, on battery production today of, of LFP, which is part of the reason why the costs are so low, um, that I think sodium, sodium would struggle to compete there on the grid makes a ton of sense. And I think we're going to see, we're going to see it on the grid for sure.

AI assessment note: “I think sodium ion makes a ton of sense for the grid. I don't think”

Q new manufacturing capacity of various kinds, uh, particularly in the U S. Do you think that that all, I mean, given what you're describing about the dynamic of the market where like adoption is a little slower adoption growth, I guess is a little slower than expected. And we have this huge overcapacity in places like China, like do, does that jeopardize this new wave of manufacturing? Do you think?

A Absolutely. Uh, you know, the IRA can subsidize, you know, the, the startup cost of building these plants, but they have to run for 30 years. They have to be competitive. And in the first five or seven years, while they're going to be able to get benefits from the IRA, there still has to be demand. Like you still have to have customers to buy. So if you build, you know, if you build five million EVs worth of production of battery production in the US, but there's only one million of demand. You're in trouble. Um, you know, there's nothing, nothing, you know, if there's no revenue, there's no, uh, there's no amount of subsidy that can, that can save you. So I do think some of these projects are still, are, are not going to go forward at certain stages. Um, you know, and that's kind of classic. A lot of times, you know, maybe half the announced projects end up getting built. The pipeline is way bigger than the demand today. Um, and so some of them will get, will get mothballed for, for some amount of time. Um, but I do think overall the IRA is helping Bootstrap domestic production of batteries, battery materials. I think, you know, it could be tweaked to, to do that even, even in a way stronger way, but it, but it's certainly having a good tailwind effect, um, for, for us and for others in the, in the field.

AI assessment note: “Absolutely. Uh, you know, the IRA can subsidize, you know, the, the startup cost”

Q build data center capacity, which is sucking up like all the locations that have available power capacity at the kind of scale that you would need actually going to make it more difficult to build new manufacturing in places where you have sufficient power and where it's cheap enough. So you guys set up a big factory in Moses Lake, Washington. How much power do you need? For that factory?

A So today we're using, we have about 20 megawatts plumbed to the, and 20 MVA plumbed to the site. We'll get to 60 in the next two years. You know, as we scale, we may need 200. Uh, so the numbers get big, uh, depending on how, how big a plant you build. Uh, but you're absolutely right. The, the, um, build out of that power is, was already slow. And, you know, we see, you know, Folks wanting to put Bitcoin mining or some other kind of, you know, there's, there are data centers, you know, down the street or in cities nearby because it's, it's hydropower. It's very low cost electricity. Um, you know, I think the pessimistic view is these things are going to come in huge conflict for sure. The optimistic view is it's now such a problem and there's going to be such a concert of voices saying, Hey, we got to fix that. And if, you know, if big techs, you know, machinery can help solve this, uh, Whether it's politically or through supply chain, you know, increases of, of key components, um, you know, that could be the silver lining in it is you're bringing to bear, you know, these huge economic players in the, in the, in the big tech, um, world to that, that may actually help if, if, uh, if we kind of work together.

AI assessment note: “we have about 20 megawatts plumbed to the, and 20 MVA plumbed to the site.”

Q Going to Washington state and especially for manufacturing, I immediately am thinking, oh, you wanted like a lot of cheap power, right? Like that's, was that a big factor for you?

A Huge factor. Not only that, we, we bought a site with 20 MVA, so we didn't have to wait in the, in the queue for getting power. And we know of companies, you know, that, uh, may miss their entire factory startups by a year because they can't get power in time and there's just nothing you can do. I mean, our ability to expand at that site is constrained first and foremost by power. And so, you know, we're working really hard with the local community to make sure that We're able to expand there, because if we're not, we're going to go to, we, we're going to build the factory. We just would build it in a different state, and that's not good for us, and that's not good for local community, and so working through all of the, like, state, federal, local elements of getting power to these manufacturing sites, I think, is, is really critical, and, and can frankly be a competitive advantage for, for states, uh, and a competitive advantage for local communities that are able to, sort of, uh, operate You know, faster. You don't have to, you know, you don't have to be really fast. You just have to be faster than everybody else.

AI assessment note: “Huge factor. Not only that, we, we bought a site with 20 MVA”

Q not signing these long-term take or pay contracts. So is it true that like when we see it, if it's, if you could go buy, you know, tier one or tier two LFP batteries for 50 bucks a kilowatt hour in China right now, will we start to see that flow through Through to consumer prices or to auto OEM margins? Like, either way, are they taking advantage of that?

A Um, maybe. I think some of those, some of those contracts, they, they may well be long-term contracts, but the reality is if the cars aren't selling, these car makers are going to find a way to renegotiate and get out. So it's, it's sort of, they, you know, if everything goes well, yes, it's a long-term contract. If everything doesn't, like, You know, it, it, there's no one, there's no one, there's no one to buy it anyways. Um, and, and, and they're not going to, you know, stockpile, you know, a million cars worth of batteries, right? Like the, the parties work out a, a negotiated agreement. Um, I think to your question of like, are car makers able to take advantage? There, there are oftentimes indices baked into these contracts for spot prices of materials and, and, you know, or some kind of trailing average prices of materials that allow them to get a lower price if, You know, the cost of inputs goes down. And the reason for that is because the, the cell makers want to protect themselves in a way that allows them to increase the price of the cells if the cost of the inputs go up. So, you know, the volatility of global raw commodities is something the cell makers certainly try to avoid. Um, and so there's, there's usually escalators and deescalators that, that, that sort of track with those indices. So it's not that they're necessarily buying spot batteries, but the, but But t…

AI assessment note: “there are oftentimes indices baked into these contracts for spot prices of materials”

Q for V to G or, or whatever. It's bi-directional. But I know one of the reasons historically has been concerns about the battery. Like if you're going to do all this bi-directional charging, actually, maybe you wear the battery out sooner. Do you have a view on that? Do you think there's a technical limitation to bi-directional charging with EVs, or is it really just like a market structure problem?

A I think, um, the technical limitation can be overcome fairly readily. If the, if there's market signal that there's the value to overcoming that technical limitation. So what I mean by that is, you know, yes, cycle life in, in, in EV batteries is limited, but it's sped as a set as a specification for a certain type of service life for the vehicle, you know, 200,000 miles, 15 years. You need to stay above, you know, 70, 80%, uh, kind of state of health. Uh, so, you know, pretty, pretty extreme. 200,000 miles, you know, is, is a lot. And, and most vehicles beat that, right? Because that's like the war, the warranty that they never want to hit. And most vehicles don't live in quite as hot of an environment as the limits. They don't drive as aggressively as the limits of the warranty. So, um, most vehicles do way better than that. Uh, you know, they're, I think Tesla Model S is out there that have gotten over a million miles on their, on their packs. So I don't think it's technical. There's a little bit there, but it's, it's much easier to overcome. I think if the market signal exists that that's worthwhile, I agree with you completely. I mean, you know, we, we have a fairly large household. We have like 300 kilowatt hours of batteries sitting in our driveway, right? We could, we could, we could be off grid. Um, but I, I think that the other constraint you have to really consider i…

AI assessment note: “I think, um, the technical limitation can be overcome fairly readily.”

Q to get the same range was worth doing in order to get to like a cost point that maybe brings more consumers into the market. So there's this implicit trade that the industry has been driving toward for a decade. Plus, do you think that that's a false trade, or do you think that Or do you think it was the right thing to do, I guess, for the industry?

A I think it's necessary, but not sufficient. Uh, you know, I, I think, you know, if you're really trying to replace all combustion fuels for transportation, you know, some consumers, for some consumers, a 200 mile EV is perfectly fine to get around the city. It's certainly much better than the Nissan Leaf back in the day that had a hundred miles. But, you know, for a lot of consumers, it's not good enough. And I think you see this by virtue of the fact that even high-end vehicles are Are still mostly combustion, right? Even people who can afford to pay a premium for their vehicle, uh, are still not buying EVs for a number of reasons. I think there's, I think there's really three reasons. Range being one of them, uh, folks, you know, we live in the Bay Area. A lot of folks want to just be able to get up to Tahoe without having to stop to recharge, right? Like you cannot do that in, in, without like hypermiling it in a, you know, in most Tesla vehicles today. Uh, and those are the best, right? Um, You know, two, recharge time. People don't want to wait around for a long time. If you could do 15, 10 minute recharge, that would, that would help things. And three, infrastructure. People don't trust the charging network, although I think with the recent Tesla opening up the supercharger network, I think that's a complete game changer for, for, for that one. So those are your three rea…

AI assessment note: “I think it's necessary, but not sufficient.”

Q and around the market. But of course there are trade barriers that have emerged and, you know, come and go. Uh, as you mentioned. So, what's your sense of, like, what the, what is, what's the geopolitical landscape in battery world as it stands today, and what prospects do you see for Western, let's call it, like, North American or, or European suppliers to maintain any kind of long-term advantage?

A Well, I think one, uh, Good thing about batteries, if you can call it that, is they're really heavy and they're a very, very big part of the value of a car. And so those two factors make it challenging to, uh, produce them on one continent for consumption, so to speak, for assembly into vehicles on another. It becomes a lot more effective if you can at least produce them on the same continent, um, so that your transportation times are much shorter. And because they're so heavy, you can't fly them. They're also hazardous to transport, so it's more expensive even, even when you, when you ship them. Um, so, so there is a lot of value to building battery factories on every continent, and I think that'll lead us to a natural distribution that's going to be a lot healthier than what we have with fossil fuels, which are sort of naturally concentrated by, by geology.

AI assessment note: “there is a lot of value to building battery factories on every continent”

Q in the battery manufacturing world generally? And the second one is I feel like in those, those Announcements. It's never clear to me who can take advantage of that. Like who's buying batteries on spot in China versus long-term contracts where you just like aren't exposed to those kinds of price volatility anyway. So like what's happening generally and what are those numbers, the spot prices for batteries really mean?

A So I think what's happening generally is that we're in sort of the second like trough of disillusionment of EVs, uh, which means that we are just past the second, uh, hype hype cycle. You know, I remember the first hype cycle when, uh, you know, Tesla was launching, you had sort of Fisker come, you know, Fisker, one point out, not, not two point out coming on, on the scene. And there were reports that said, you know, 20% all cars by, you know, 2015 are going to be electric. And it was, it was sort of laughable when you really drill down because the amount of just factories that would have to get retooled for that to be true, it was just nuts. Um, and so when that didn't kind of come to pass, everyone sort of Said, oh, this is going to take a lot longer, right? And it took until really twenty-twenty, um, you know, or 20, I guess, 18 or so, twenty-nineteen, when, when the Model Three took off, that everyone said, oh, wow, no, EVs are going to be for real. And so that started the second kind of, you know, wave of excitement. And in that wave, you know, again, the projections were all, half of all cars are going to be EVs, battery costs will continue to plummet at an insane rate. And it just, it just didn't happen. Consumers didn't adopt as quickly as folks wanted. Battery prices, Dropped some, but not so much. And again, even with LFP being 50 bucks a kilowatt hour, it's not makin…

AI assessment note: “So I think what's happening generally is that we're in sort of the second like trough”