Shoals Technologies Group, Inc. ($SHLS)

Good morning, everyone. Thank you for joining us today for this webinar on Battery Energy Storage Solutions with Shoals Technologies. My name is Christine Cho, and I'm the clean tech analyst here at Barclays. We have a great discussion prepared for today. We're going to start with a little bit of why storage has increasingly become important in today's backdrop, the type of storage needed for data centers, the difference in architectures, the value proposition of Shoals and the size of the opportunity. I'm going to turn it over to Matt Tractenberg, VP of Investor Relations and Finance, who will introduce the Shoals team and give some introductory remarks. We will have some slides during the discussion and also have a Q&A session at the end. So if you have a question, please e-mail [email protected], and we will try to get through it all. Matt?

Thanks, Christine. Good morning, everyone. Again, my name is Matt Trachtenberg. I'm with the Shoals' IR team. And before we give it back to Christine to sort of drive through some of the questions, I wanted to just introduce our 2 speakers, add a little bit of context for today's event. So as you see here on the slide, Jeff Tolnar is our President, and he's responsible for our commercial efforts at Shoals. And Kishan Ponnadurai is a product engineer specializing in our BESS solution. The subject of BESS has been top of mind for most investors since last fall when it really started to pick up steam here. We get a lot of questions about what we sell, who do we sell it to, why Shoals and the specific application that it fits into. So this call is designed to be educational in nature. It's not intended to provide financial updates or forecasts. Those would be provided on any of our quarterly earnings calls. So Christine has been doing a lot of work in this space, and you're kind enough to host us today. We have reached out to a number of shareholders in advance to ensure that we're addressing the most relevant questions. But again, as she pointed out, if there's something that the audience wants to hit, you can e-mail your questions to her e-mail address that you see under her picture as we move through the hour. I think we have a great discussion today for everybody. So Christine, I will give it back to you.

Great. Thanks, Matt. So good morning, gentlemen. I wanted to start off with -- I think it's really important to understand the events of the past to bring us to where we are today. So can you provide a brief history of the evolution of batteries in recent years?

We'll do. And thanks, Christine. Matt, if you can pull up the slide for that. I won't hit all these bullet points, but I'll hit the main areas. So starting left to right, one of the -- in the early stage, batteries were relatively small. They were AC coupled, typically a 1:1 ratio of the battery stack and an inverter. That was about 10 years ago. And then going to the bottom, you look at about 8 years ago, began scaling and then largely driven by EV battery innovation, the batteries started to become more bankable and they began to grow to multiple megawatts. And then a few more years go by were about 5 years ago, once the spike in lithium came back down to normal range, the battery blocks continue to grow in size. I think one of the most telling points is that the energy duration increased. So you started seeing more and more of multiple hours of capability, megawatt hours of capability as opposed to just minutes of ride through. And then the continued improvement in economics driven largely by scale. And then if we look at now, there's quite a few things that have happened and a lot of today will be diving in more deeply. The first is that there's alternate technologies that have been introduced other than lithium ion, some of the long-duration energy storage systems that have been announced recently are very compelling. They can provide tens, if not 100 hours of long-duration storage and the new innovations around BESS as a firming mechanism and a UPS infrastructure for data centers. So that's a really exciting part that we're going to dive more deeply into today, I think.

Great. On our webinar last year or maybe it was at our conference, you mentioned that the product, your specific product is needed for projects that are DC coupled. Can you talk through why?

Yes, Jeff, I'll take this one. I think it's important to categorize the 3 major components of why our product is useful in DC-coupled installations. It really boils down to augmentation planning, protection and cost. I mean those are the 3 biggest factors. I think a lot of DC-coupled applications or infrastructure, what they kind of go through is they're planned from the very beginning, and it's hard to retrofit. It's hard to build on top without adding more and more product. And what we offer is a method to plan for augmentation from the very beginning. And so that's one big value proposition that our recombiner has in the market. We also have an extremely high short circuit protection. And I think that's more -- it's more important as we continue to involve the battery technology. Short circuits are getting higher, batteries are getting bigger. So it's important to have that over current protection. And of course, cost is a very big one. I think our value proposition of being able to aggregate a bunch of DC blocks to lower your -- potentially lower your -- the amount of inverters that you're putting on your site is very attractive for a lot of our customers. And that's simply due to the fact that we can simplify your storage charge solar -- sorry, solar charge storage and help with that augmentation planning.

And just for those who aren't as familiar with these terms, can you just give a little one-on-one on what you mean by augmentation?

Yes. Augmentation is basically the idea of expansion. So as you're building a site, your project -- your demand changes, you need to kind of move with that demand. And augmentation is simply just growing your site to meet that demand.

Okay. Do you have any statistics as to how many storage projects are AC versus DC.

Yes, it's a great question. I think you have to look at with where batteries are going right now, I think it's hard to look at all storage projects as AC coupled, DC coupled. What I will say is that when you look at any sort of interconnection site, so any solar plus storage site that's supporting the grid or any stand-alone storage site that's supporting the grid, you can -- it's pretty easy to look at what is AC-coupled versus DC coupled. When you get into some of the other types of applications for batteries, it becomes a little bit more blurred. So when you're talking about those interconnected sites, the solar plus storage and stand-alone storage supporting the grid, typically, what we're seeing in the market is about 80% AC coupled and 20% DC coupled.

Okay. And then what are the puts and takes for each as to why a project developer would want one over the other with respect to AC versus DC?

Yes. So another great question. And it's a question that we're asked a lot of. So we'll start with DC coupled. I think the biggest puts for DC coupled is truly a cheaper upfront capital expense. If you're able to lower your -- the number of inverters you have, potentially use a recombiner to help you do that, that's a pretty big put. The other thing is a compact footprint. So instead of having a 1:1 battery-to-inverter ratio or -- and a solar-to-inverter ratio, I think one of the big things is if you're able to reduce your footprint and increase your power density in a smaller scale, I think that it's a very attractive option for developers to go with. And the last one -- the last big one that I think captures the biggest puts is the solar energy capture. And what I mean by this is typically, in a DC-coupled infrastructure, you're using a common bus that is fed DC energy. So your solar is DC coming out of the solar panels, your battery is DC coming out of the batteries. And what you're doing is you're aggregating these DC inputs on a single bus and allowing solar DC energy to charge a DC battery as opposed to on the AC-coupled side, converting your solar energy into AC and then converting your battery into AC and then going into an inverter. And so I think what I'm trying to say here to simplify things is that you have lower conversion loss when you're looking at a DC-coupled infrastructure.

So the earlier statistic where you said 80% was AC. Can you just sort of maybe help us walk through if there's all these benefits with DC, why is the majority AC?

Yes. What it really comes down to is operational flexibility and revenue. And so what I mean by operational flexibility is I've mentioned augmentation earlier. It is much easier to augment an AC-coupled infrastructure. So it's easier to grow an AC coupled infrastructure to meet your demand simply due to the fact that there is a 1:1 inverter ratio. This is important because what you'll realize in an AC-coupled infrastructure is that the solar and battery operate independently of one another, and they're not connected to a DC bus. Why is this important? It really comes down to how you're using your energy, how you're gaining revenue from your energy, things of that nature. So because in a DC-coupled application, you're tied to a common bus, you can't operate solar and energy storage at the same time. With AC-coupled, you can. So I'll give you an example. Let's say your price for energy is high and you want to sell some of your stored energy from your battery storage while your solar is still producing energy. In an AC-coupled application, that's possible. In a DC-coupled application, you run into the risks of curtailment. You'll have to stop one of your operations in order to have -- in order to operate the other. So that's probably one of the biggest reasons why AC-coupled is seen as a more prominent infrastructure in today's society. It is driven by revenue and operational flexibility.

And specifically for data centers, are you seeing it skew one way or another with respect to AC versus DC?

Yes. It's a great question. Data centers operate a little differently, and we'll get into why that is in later down the presentation. But I'll simply say that in today's world with new infrastructures evolving, it's hard to just simply say that one installation is DC-coupled or AC coupled. And we'll get into exactly why that is.

Okay. Great. When you guys kind of initially teased out this product, you had mentioned that it was initially for front of the meter, but it's just that you got a reaction from the behind-the-meter market when the product was released. So curious on how the market is evolving. Is it -- are your customers primarily BTM? Or are you seeing FTM interest? And if you could provide a split?

Yes, I'll take that one, Kish. I guess when I look at it, it's front of the meter is really grid-facing. So I look at the 2 use cases that Kishan just had up on the projection. If I have solar plus storage and providing energy to the grid, that's going to be grid-facing. If I have storage as a resource for the grid, that will be also front of the meter. I think what's emerging more and more is with the large load growth driven by data centers and AI training data centers is there's a mix. And Kishan alluded to that a little bit. And without diving too deep technically, what I would say is that what we should focus on is what's the predominance of where the energy is going. So we're seeing with the data center use case, that the energy is predominantly being provided to feed the data center, but that energy can also be used if the data center were to drop in power, it can be used as a buffering resource back to the grid. So it really could look like a combination of the 2. So I think a hardline view of is it front or back of the meter, I think it's blurring right now. And I do think -- and you didn't really -- you didn't ask this question specifically, Christine, but I think it's important to note that I really do believe that I get questions about is the grid going to be bypassed and is it going to be purely microgrids. I guess in my point of view, the grid will be connected in the vast majority of the cases. There will be fringe cases where a microgrid could feed a data center. But grid energy is sustainable. It's been there a long time. It's highly protected. It's redundant, and that's something that's difficult to replicate in a pure microgrid. So I do think that, that blur of front of and behind the meter will continue. And it is, in my mind, all about which direction the energy is going and what it's feeding.

At the risk of getting too technical, just because you gave these -- and I know we're going to talk about it a little more later, but you gave 2 examples of like the batteries are used for buffering, but they -- the data centers can also be pulling from the batteries. In those 2 separate use cases, is the battery placement separate, like different?

It's not. The batteries are an energy source, and it really comes down to is in the initial engineering design, are the batteries contemplated as just outage ride through. So again, it gets back to the use case of the data center. Do they just want to ride through an outage in which case the batteries are there to provide minutes or hours of support if the grid were to fail? Or has the design been put into place where the batteries can actually be used as a buffer back to the grid because we're hearing more and more from electric utilities and IOUs that -- or the independent system operators, the ISOs that they're concerned about data center impact to the grid itself. Things like GPU ripple and voltage stability are potentially causing issues in neighborhoods and with other adjacent C&I facilities. So in that case, if the design is done upfront, then the batteries can be used as a firming resource back to the grid. So it goes back to initial design and intent of what use case the data centers are trying to achieve.

Okay. And I know we're going to go into the use cases a little more, so I'll just leave it there. But -- and you kind of touched upon this, but let's just dig a little deeper. What has changed about data centers in the last year or so that is warranting so much interest in battery energy storage?

I think we've got -- a slide will help in this one. It goes back to the prior question, which is where -- what's the primary purpose of the energy that's being provided. So in the upper right-hand portion of this slide, you see the substation. That's grid connected. It could also be natural gas turbines that are behind the meter that are connected. They then go to a transformer that steps down the voltage to the point where an inverter can take that power in. It then goes through a -- we're hoping in every case, a Shoals recombiner, which then goes to a battery stack battery container charges the batteries. The batteries then are used to discharge through an inverter that steps it down to a voltage that the data center can use and then it feeds into the data center as AC. So you think of that power flow coming from the grid or natural gas turbines. This use case is very new. The diagram that you're seeing is publicly available. It's indicative of the way that many are looking at it. And I see it as quite interesting because if the design is considered upfront, you can see that the power could be used to firm the grid, just like the use case that Kishan showed of batteries facing the grid or the batteries can be used to feed and make the voltage and continuous power more available to the data center. So that a design like this enables the best of both worlds for grid firming and then also for data center operation.

So we have -- at the risk of oversimplifying, we generally have heard that the use case for batteries comes down to 3 things: one, power quality; two, speed to interconnection; and three, backup needs. Is your product necessary for all 3 scenarios? And how does the demand differ for each of the scenarios and the relative sizing for each?

Yes, Kish, do you want to take this one?

Yes, I'll take this one, Jeff. Yes, our product is -- our product plays across all 3 scenarios. And in many cases, it's hard to say that each of these scenarios are independent of one another. I think there is a lot of meaningful overlap between them. However, it's up to our customers to prioritize which one of these 3 that they want to go with. So when we're working with our customers that really want to prioritize power quality, it turns into, okay, what exactly are your product requirements? What are the needs that your site actually needs for this specific location, this specific area and these specific power electronics. And that becomes a very custom tailored product that we build for those customers. And typically, we do see that as more of a CC&I opportunity, although we do work with utilities in that regard as well. But when it comes to power quality, cost efficiency and flexibility in how we design the product is truly the name of the game. So it becomes a very, very tailored product. When you're talking about prioritizing speed to interconnection, once again, power quality is obviously very important at all stages. But if speed to interconnection is a priority here, and typically, what we're doing is building a larger system and a standardized system that they can use for repeatability. We want to be that bankable partner for that customer that's trying to scale their business and quickly get developments up and running. And so when you look at a standard product, scalable DC architectures are what's needed. And those are something that we can support as well as those flexible power quality-focused products. When you go into backup, it's funny because it's almost like we want everything when it comes to backup because the fact of the matter is power quality is very important when it comes to these backups, especially with data centers. But also with all the demand that we're seeing for data centers, speed to interconnection is also extremely, extremely important, mainly because these data centers, what we're seeing, they can't exist without a sort of interconnection. Right now, there's a big push to make sure data centers have power packages alongside with their overall development. And so what you're seeing is that power quality and speed to connection, speed to interconnection are both extremely important in that regard. So you'll often see a combination of the product we build for a customer that's specified -- that is prioritizing power quality and a customer that's prioritizing speed to interconnection. So we are building extremely custom solutions for these backup these UPS solutions that are able to be scaled at very, very high volumes. And so that's one of the big focuses that Shoals has taken in recent times is how can we standardize on maybe a skeleton, but allow for flexibility on the internals of our product, so we can meet those customer needs and scale with the increasing demand. I think it's important to kind of talk about our -- where we're going with our business. So one of the big things is that when we're looking at these custom products, our biggest goal is to make sure that we can repeat even the custom product. And so that's why we focus on building that skeleton framework and then focus on designing the internals very specifically.

So you went through power quality and speed to interconnection in more detail. Would it be fair to say that backup is kind of on the back burner for now?

I would say backup as a -- if you're talking about backup as a solution for data centers, it's at the forefront. If we're talking about backup as an energy plant, an interconnected solar plus storage, stand-alone storage site, I would say speed to interconnection and power quality do take priority over backup.

And then for the bookings that you do have so far, do you know which case they are using it for?

Yes. For our current bookings, we're seeing the strongest pull from backup data center-specific backup and high-power applications. And that's primarily driven by the AI infrastructure that's being built up today. And so we're able to handle that high current scale DC distribution and support very, very tight reliability requirements at a large volume. Power quality alone and speed to interconnection, those use cases are still very much apparent, but they're typically a smaller portion of our demand right now. And they do tend to be more cost driven.

Okay. What is the decision cycle of a data center and the difference in architectures that are being evaluated?

Let's go back to the 3 use cases. The decision cycle for BESS added to a storage site is pretty much the same as solar as the solar -- utility scale solar sites. So we're seeing 9 to 18 months as the typical range for solar plus storage. And I would also say that most solar sites these days do have some amount of BESS associated with them and attached with them. The grid firming use case, those are typically through electric utilities, and we're seeing a longer sales cycle for those. And I think a part of -- the reason is twofold. One is that electric utilities are a little more conservative and they want to see a trial and then before they go to moderate scale and then to large scale. The second reason would be long-duration energy storage is at the forefront for the grid firming use case. And those are relatively immature in the market and are just emerging now. So the sales cycle has been longer on the grid firming side. I'd say, upwards of 12 to 24 months. And we've been in on some of those for quite a while. And then the third use case that we've talked about is the data center where BESS is a firming resource for the data center, a UPS and also a firming resource for the grid. That one ranges in the 9 to 18 months. So it really varies by use case. But I don't see a quick turn 30-, 60-day sales cycle in any of them. They're typically going to be 9 months to up to 2 years for some of the fringe cases.

Okay. And the stuff in your backlog, I think you guys have said that like half of that would turn this year and then the rest is sort of next year. Is that what we should think? -- extend beyond that. Okay.

No, that's exactly right. That sales cycle started about 9 months ago. And by the time it turns into revenue, it's going to be about a year.

Okay. What is happening right now that is creating this specific opportunity for you guys?

Yes. I think we talked about some of it already. And I want to tie back to some of those comments. BESS as a ride-through technology. So if the grid were to fail, and we had some issues in Texas a few years ago, we had issues in other parts of the country where the grid would fail and you needed energy to be provided to, in this case, a data center. So the batteries were in place for ride-through capability and outage -- basically outage ride through. What's changing now is with data centers and AI and when a data center is in training mode or whether it is in inference mode where it's in standard normal operation, when an AI data center is in training, is a training data center, you have something called GPU ripple. And basically, those are high voltage fluctuations in the signal that cause problems on the grid itself when that occurs because the loads are so massive out of a data center that it causes cascading issues within the grid. So GPU ripple is a new challenge that the industry has had to face. The second is tied to it, which is voltage stability. Because the power fluctuations are severe and they happen so fast, you can see a swing in voltage on the grid itself. And then providing continuous power is a third. If the data center were to drop its operating load substantially, the grid would see that and might not be able to react as quickly. So the grid might have a continuous power issue in that area. Conversely, if the grid isn't able to provide power, the data center that the data center needs, it may have an issue. So those 3, I would say, use cases within the data center segment are relatively new last couple of years and the solutions that we're seeing on the market are addressing those. So it's using BESS as a UPS resource as opposed to the more standard low-voltage UPS resources that were deployed over many, many decades.

So let's talk about your specific product in your process. What do you exactly sell? What does it do? Why do people buy it from you? Who buys it from you? And then if we can get maybe an idea of product road map, what's next for this product?

Yes, sounds good. You want to go on Slide 5. You can jump in.

Yes, you do the what I'll do the who.

Sounds good. So what do we sell? So it's a good question. I want to bring your attention to the left 2 boxes, the green boxes. What we sell right now is a product that allows you to aggregate multiple DC circuits. We have combiners that can be wall mounted, and we have recombiners that are freestanding. And the main thing is we obviously play in the DC-coupled market because we are consolidating and aggregating DC inputs. But we offer a method for developers to consolidate and reduce the number of inverters on their site and in doing so, reduce the footprint of their site as well, maintaining their power density. We go into a range of amperages. The highest one and the most common one we typically see for battery storage products is that 4,000 amp recombiner, where we can take up to about 16 fused inputs into a cabinet. Now that does require a larger cabinet profile. But the fact of the matter is we can take up to 16 in that 4,000 amp recombiner. And I think the biggest thing that we've started to see more and more of is as batteries continue to evolve, as batteries continue to get bigger, as duration starts to get longer, as footprint starts to get smaller, the fault current becomes more and more of a risk. And so what we're seeing is that the fault current no longer is just 100 to 150 kIC. Now we're going up to the 200, 250, 250-plus kIC. And right now, we're designing our product to take on that high of a fault current. So that's kind of what we do with the combiners and recombiners. But we have an attachment that I think offers a lot of control to our developers, which is the multi-load break disconnect. And basically, what this is, is all those inputs that you put into your combiner or recombiner, if you want individual control of each one of those inputs, then you can attach a multi-load break disconnect to that unit. Why is this important? Well, when it comes to doing service or any sort of maintenance, if the developer wants to maybe not shut off their entire operation, they can have that individual control on their inputs and really isolate what needs to be shut off and what doesn't need to be shut off. And so we offer that as an attachment if the developer does want that. So what's next? I think the -- where we see the market going is definitely in reducing footprint while maintaining power density. And the best way to do this is to really consolidate your components onto a skid. And so where we're looking at right now to take Shoals to the next level is looking at skidded power electronics for our next launch.

Yes. Christine, as far as the -- why do people buy from us, I think I'd look 2 different directions. If I look at this portfolio, all of the variables that are in there make it a high mix solution. So every -- we found that every data center, every solar field is a little bit different. We have differentiated ourselves over the years in solar by providing high mix at a high volume. That same approach is resonating with and in the best market. The cabinets themselves range in size. What goes into the cabinet is designed for that site. The configuration is designed for the site. The fault current rating is designed for a specific site. If I look to the large multinationals that I'm sure all of you know, we compete against them because we are very good at and very willing to do custom solutions on a site-by-site basis. That's what we've cut our teeth on over the years. And we've got a scaled production line set up to do just that. So our ability to compete against the multinationals is tied to this high mix, high-volume capability that Shoals has had. On the opposite facing realm are the smaller, I'll call them the small panel shops that can create and do quite well at creating 5, 10 of these similar cabinets per month, but a data center is looking at hundreds of cabinets in the scope of their project. Plus you look at the bankability of those companies, the smaller panel shops that may or may not be profitable and have their own cash flow issues, whereas with Shoals, we're a $0.5 billion-ish company. We're cash flow positive. We are bankable public company. So our customers view us as steady and bankable and able to take on these large multiple hundreds of units of project -- products within a project. So I think we're at the sweet spot of the market of willingness to customize, but also able to -- and also able to scale and able to differentiate ourselves through the volume and size that we are as a company. Now the who buys the question is it's going to tie back to the use case. If I look at the solar plus storage use case, those are typically procured by the EPCs. So the engineering procurement construction companies. The grid firming use case, those would be an EPC, but could also be an electric utility or an IPP, a public power operator. And then the data center operator, it's varying right now. It could be an EPC, it could be an OEM. It could be an energy and engineering firm that is coupling solutions together. So we're seeing a broader variance. And I think that's a testament to that the ecosystem hasn't quite settled yet on the data center space. So our customers are a little bit more varied there.

Okay. You touched upon all other things that I would like to go a little deeper into. But before we do, when I just think about your solar EBOS product, which was your flagship entry into the market, the value proposition here has always been labor savings. The higher CapEx more than offsets the labor. And so it's like very easy for investors to understand why someone would take this product. What about for this BESS product? What exactly -- can you just talk through what is your product "replacing" or maybe just talk through the value proposition of why you versus someone else or something else?

Yes. So I think it's back to the use case. Solar plus storage we enable DC coupled. Kishian talked about some of the advantages earlier of a DC-coupled solution. We have -- we are DC experts. It's what we've been in for the duration of our company's history. So a DC-based product at high voltage, so up to 1,500 volts, 4,000 amps, so incredibly high-power DC systems. are our specialty, and we get quite a few opportunities to come in just by our -- being in the DC side of the equation for grid firming and solar plus storage. They're used to doing business with Shoals on the solar EBOS side. They will continue to work with us on the best recombiner side. And then it gets down to the difference of would they rather buy from a large multinational that may not be willing to customize, I'll buy a SKU off the shelf. What we've heard more than once is our customers said, well, they've asked in the market and the response is, well, I've got this box, this box and this box. and they kind of fit your need. And with Shoals, we're able to say, no, we'll make a system that fits your need for that site. So it's that flexibility that helps us compete in regardless of use case, but primarily in solar plus storage and grid firming. And then the other alternative would be the panel shops. And what we've seen of most of the panel shops is that they are AC experts. So they're making large AC panels. They're making AC switchboards, but they haven't really gone into the DC area as yet, they might. But we've got quite a lot of, I'd say, a decade plus of experience on them in that regard. So I think that sets us apart for quite a while. And then also, we're continuing to evolve our thinking of it's not just a cabinet. Now it's a cabinet plus a power system on a unified skidded solution that will reduce labor from the field. So eventually, we're going to get back to our cut and dry. We're going to take X amount of labor out of the field, put it into the factory. That's when we start to introduce skidded solutions.

In the past, you've talked about 3 sales channels, leveraging existing customers, selling directly to data centers and then co-engineering with integrators. It's been about 7 months since you started talking about this product in more detail. Can you update us on what you have seen since then? -- which sales channel is getting the most traction and which one has the biggest tail?

Yes. I think the quick answer to that is the one that we're talking the most about, and that's the BESS as a resource for data centers. If we look at the other 2 use cases, I think what we released at Investor Day is we're thinking that the solar plus storage is about $360 million TAM. Okay, interesting, but not massive. And then grid firming, we're seeing that emerge more and more with long-duration energy storage. So we think that TAM could be about the same, maybe larger. But then when we look at the data center opportunity, and there's a graphic that I would like to show that ties some numbers into what we talked about earlier. So I'm going to give a couple of examples and some ranges. What I can't do today is provide you a TAM, and I'll explain why in a minute. But I'll give you some tools that can help you get to that point. So one -- and again, this is a energy flow diagram that is publicly available. And if we equate it to a 1 gigawatt data center, depending upon the size of the battery stacks, you'll need some number of Shoals recombiners, 400 to 600 would be a range that we would be comfortable with based on the size of the batteries. If I have much larger batteries, I would need fewer recombiners. If I had a smaller battery, I would need more recombiners. The ASP for the recombiner varies pretty broadly, whether I need 1 cabinet or 3 cabinets, the number of interconnections that Kishan talked about earlier, do I have 8 feeds coming in, 16 feeds coming in? How many fuses do I need per recombiner? What does the disconnect structure look like? So it's a fairly broad ASP. And then if I do some math, and the math won't be obvious, I'm going to take you through it. If I look at the lower end ASP, that would be a smaller recombiner. But if I have a smaller recombiner, I'm probably going to have more of them. So if I look at the lower end of a range, it would probably be the 70,000 ASP times 600, that gets us to about $42 million per data center. If I look at the higher end of a range, that would be a multi-cabinet recombiner. And if I'm putting a high end together, I would use the high-end ASP times the high-end number, and that gives us a range of 42 to 72, again, just pure math. We've been saying for a while now that our range is $50 million to $60 million per gigawatt. That's about right, and that's where those numbers get derived from are these ranges. And I would say that we haven't seen enough volume to know what the predominance is within that range. So I'd say let us sink our teeth into it a bit more and see whether it skews to the high or low end, but that's a good range for you to use. And then if we look at the data center market, and we've looked at it a lot of different ways. We use WoodMac as a company. A lot of companies use BNEF. We've done a math calculation to roll up the number of NVIDIA chips and the power consumption of NVIDIA chips bottoms up. They all round out to about a 10 to 15 gigawatt per year U.S. data center market. Now here's the difficulty in getting to a SAM or a Shoals addressable market is -- we don't -- this is so new, this BESS as a UPS or as a firming resource is so new. We don't know of that 10 to 15 gigawatts, how many will have a BESS system. We also don't know the power consumption that will be required from that BESS system. So it's too early for us to say that some percentage of that 10 to 15 is going to have BESS versus not. And then within that number that will have BESS, how many are going to be this configuration versus the use case, which is just outage ride through. So I think there's too many variables right now, and I'd have to give you a wild range that would make no sense. So -- but I think we would -- as a company, we would stand firm behind these 2 numbers. And what we're trying to do is figure out where they land from an overall addressable market perspective. Is that a fair way to hedge it?

You're asking me?

Yes.

I guess my -- maybe my follow-up would be because -- and I totally understand what you said about why it's difficult to sort of put a tighter range. And -- but also just sort of since you first started talking about this topic, there have been new products, new ways to sort of maybe solve for the same thing. And so I guess I'm curious -- and maybe this is an unfair question, but if you just kind of had to compare what you thought or what the range of outcomes of what TAM could be was 7 months ago when it was very early stages versus today because you do have more information today, would you say it's up, down or about the same?

I would say it is significantly up. Because if you would have asked me, and I think you did ask a few months back, and it was really the first 2 use cases. It was solar plus storage and grid firming. And I would say this opportunity is substantially larger than those 2, and it's moving faster. What we don't know is where it's moving to and at the pace and what the adoption curve is going to look like. So it definitely is larger. It is a larger overall market opportunity for us. And I think we play well in all 3 use cases.

Okay. What about -- earlier, you talked about shrinking footprints and increasing energy density. So what about the move to 800-volt DC? Does this move impact the need for this product at all?

Yes, love the question. And we see the white space of the data center. So within the data center itself as a different product category. A DC recombiner could be advantaged by a data center going to DC. But I don't know what that could look like. But I also know that we've got a product that we're -- that we unveiled a little bit at Investor Day, and we've continued to work on within that white space, a cabling system concept that is moving along quite well within the data center. Now think of this, Christine, if you've got DC that is feeding the racks and the rows or the rows and the racks there's still going to be AC power needed within a data center. You're still going to have to feed lights. You'll still have to feed standard HVAC systems that are AC powered. So there's going to be a combination of AC and DC that goes into a data center. But what I would say is our DC focus is advantaged by the move in the industry to go to DC. And I see that it makes total and logical sense for the data center operators to do that. They're looking at ways to reduce power consumption and heat dissipation is one of the primaries. So if I go to a higher voltage, I go to 800 volts DC or even eventually to 1,500 volts DC, I can get the same power with lower current, which means less heat dissipation. So I do think the DC architecture advantages us for the best product that we've been talking about. That's a little bit separated because you're still going to need some type of inverter in the middle and power conversion in the middle. But I would say our product that's within the white space of the data center definitely is a great fit for 800 volts DC or if they continue to work down the AC path.

Okay. So just to make sure to clarify, that's a separate product.

Separate product.

Yes, one that I'll be bugging you about later on.

Please do.

I'm going to -- I have one last question. Just as a reminder for everyone on the call. If you do have a question that you would like asked, please e-mail [email protected]. The e-mail is right below my video. So the Shoals BESS recombiner box is usually discussed synonymously with data centers. Is there a need for this product with other customers?

Yes. Jeff, I'll take this one, if you don't mind. It's important to consider the birth of this recombiner. It actually was initiated by your typical solar plus storage, stand-alone storage sites. And I think what we've seen in those sites in our ability to consolidate inputs and simplify those inputs to a singular output. I think that value proposition has been realized in this battery storage UPS system because what you're seeing is kind of a reverse recombiner. You are consolidating 2 inverters and going to a singular battery. And I think the theme that you're seeing here is that there is a need for consolidation and there's a need for simplification. And I think customers are always looking for simplification. And when you look at our recombiner, we offer a custom way to offer that simplification based on the customer's specific needs. And so I think it's important to make that distinction that this product did start with your interconnection sites, solar storage, stand-alone storage and that value proposition of simplification and has been realized in this data center world that we're in right now. And so I think what we're looking to do is continuously improve that value proposition, continuously improve the product to add to that value proposition and grow.

Okay. Well, thank you, gentlemen. We have, let's call it, roughly 5 minutes. So I did get a number of questions that I'll try to move through. It seems like a lot of them has to actually do with competition from what I can tell. How does your DC to DC recombiner box differ from those offered by Schneider Siemens? Also, what is the typical solar to storage ratio? Is it 1:1, 1:0.5?

Yes. Kish, I'll take first half, maybe you take second half.

Sounds good.

I think it goes back to, Christine, the question we talked about earlier, we compete against Eaton, Schneider, Siemens. And our ability to customize is what differentiates us. I would say we see them less than we see some panel shops when the customers are looking at because I think customization is top of mind right now. And what we're hearing about the multinationals is that they have standard SKUs on the shelf and they're trying to sell the standard SKUs and whereas we're willing to customize, but can also scale. So that's where we're winning today. Where a customer could buy a standard SKU and they're very comfortable and maybe they've got a bulk order in for substation gear from one of the big multinationals, they may just couple in the best cabinets there. But I would say that our customization and scaled line capacity is what sets us apart, and that also sets us apart from some of the -- from the primary panel shops. Kish, can you take the ratio question?

Yes. So I just wanted to add on to that. One other thing we did touch on this earlier on is that when you look at an off-the-shelf cabinet, you're typically looking at a lower short circuit current rating for those cabinets. When we design our cabinets, we understand the increasing need to handle a much higher short circuit. And so that goes into our design as well. So -- it's not just that we're able to scale and we're able to be custom, it's that we're able to work with our customers and become a sort of partner with them as we're designing the product that works for their site that needs to work for their site. So that's just one piece that I wanted to add. As far as ratios go, it's an interesting question. Christine, was that about ratios of solar to recombiners? Is that what we were going with? Or what was the question, that second one?

I thought it was solar to storage.

Solar storage, yes. So typically, what we're seeing on our side is we would see about, I would say, 6 inputs going into the recombiner to 6 solar inputs going into the recombiner to 2 to 3 battery inputs going into the recombiner. So to answer your question, I mean, it's about a 2:1 ratio, solar to battery.

Yes. If we would say megawatts, a 500-megawatt solar site is going to see about 250 megawatts of battery.

Yes.

Okay. Great. And then I got a question asking if you can comment on the margin profile of the BESS system by use case. So I understand if you don't maybe want to give explicit numbers, but maybe if you could talk relatively if there is a difference between the different use cases that you've given.

Yes. I would say that the margin profile is consistent across the 3 use cases and that it is accretive to the solar margins.

Okay. And then I think this is going to be the last one, but I figured it would maybe be a great way to sort of provide a segue for your next product. But I did get a question of can they talk a little bit about the inside data center product that you guys alluded to just a little while ago?

Sure. Yes. And Kishan, I'll defer to you at some point. I'll stay high level. Data center rows and racks are fed today primarily using 2 technologies. One is busway systems that are large copper bus bars that run along the row and then you have a tap-off unit that feeds power down to the rack. The other technology that's used are called WIPs. So basically, home run wires from the rack back to a power unit that provides the power to the WIPs. We're -- and we announced this at Investor Day, so I'm not disclosing any material nonpublic information, but I'm refreshing it. We're introducing a product that we feel fits very nicely in between the 2. So we're using our BLA, our big lead assembly that's been in solar fields for over a decade. We have -- are testing it to the AC environments that you see within a data center and basically envision now where the busway would go above the row of racks. Now envision BLAs to have a cable management configuration that runs along the row of racks and then the drops that would normally go to a solar string. Those drops then feed into a tap-off box, which then goes down into the rack. And we believe that the system we're developing will substantially save on labor, but then also can be turned around very quickly for a specific site. So no more long lead times for busway.

Yes. I'll just add one more thing to that. The other thing that we kind of mitigate is short circuit risk with using these BLA cables. If you have cables in a bus, potential for positive to ground, negative to ground, I'm not going to say it's super high, but it's higher than using the BLA cable. So that's just another point. Jeff did a great job bringing that to life.

I'm sure I'll be asking you for a webinar when you make more progress there. So I look forward to that. Well, gentlemen, thank you so much for the time. And for everyone who tuned in for the last hour, thank you for joining, and I hope everyone has a great rest of the day.

Thanks all.

Thanks, everyone.