Talking behind-the-meter power generation with Bloom Energy (Ep. 234)

All right. Welcome back to Energy Sense, an S&P Global Energy Podcast covering all topics on the intersection of energy and finances that your host, Hill Vaden, you're with the other host, Sam Humphrey. Sam, how's it going? It's good. Thank you, Hill. How are you? I'm doing well. Do you have a good weekend? Did, did, six nations is on. So for anyone who has any interest in European rugby, it is on at the moment. I'm peeing with the Olympics. Well, the six nations is on every year. So they have moved a few of the matches just to accommodate the Olympics coverage. The GB is not great in the Winter Olympics. I think we've got like two medals and I might be massively understating it now, but we're not. We don't have snowy mountains and stuff, so we're not. So with some Olympics is better for us. I'm a much bigger fan of the Summer Olympics, which seemed to be a much more inclusive Olympics if I don't have money to buy a bobsled, I'm not we're going to be a very good bobsled. But I could learn how to run and be a really good runner. Exactly. I skiing if you don't live next door to a mountain is prohibitively expensive. Yes, even when you do an extra mountain experience. Yeah, absolutely. All right, well, we just finished a fascinating interview with Amanjashi. The chief commercial officer of Bloom Energy and Bloom Energy is a fuel cell company, a publicly listed fuel cell company that really caught people's attention a few years ago with the hydrogen boom and has absolutely taken off with the behind the meter data center. I'll call it infatuation where natural gas is being used, that natural gas, that the need for dispatchable generation for data centers has really opened up this space for fuel cells. And the US in particular, but I think Aman does, you know, hints at some other areas. Can you give people a few things to listen to over the subsequent 30, 35 minutes? Yeah, sure. So what spots were part of the things that sparked this conversation is the phenomenal growth that Bloom had last year. I think their share price went up to 300% year-on-year or something like that. I think huge growth, their backlog and product has grown like two and a half times and they stand in a major partnership as well for around $5 billion. So for a single year, it was some big numbers that were thrown around and what's interesting is this technology has sort of been rumbling away for the last 15 years or more and it's this expansion with data centers and that need for off-grid power that is really driving this forward and you know, Aman does a really good job of sort of explaining the mathematics behind why this is so beneficial to people and why they might want it. I'm not going to try and break that down now because that would take away from him, but it is fascinating and like you said, it's not just in the US where this technology is applicable and being used, you know, it is a global company. But what about you, any key things you want people to be listening at for? You've captured most of it there and Aman does a great job describing it, but it's just another instance of energy innovation that a listener should keep an eye on and we will hand it off to the episode now. Adoli. All right, well, Aman, thank you so very much for joining us. We know that you and probably all of these executive team at Blue Energy is busy this year coming off of really a wild 2025 where I think your stock was up something like 500% start of year to end of year and there's still a lot of enthusiasm around what Blue Energy is doing with fuel cells and with the data center space. So there's a whole lot to unpack so we're very grateful for your time, but maybe if we could kind of get a brief introduction of Blue Energy for those who are listening and are less familiar with the company and you know, you're somewhat, I guess kind of the hydrogen focus and the continued hydrogen ready projects as you really boom with the data center investment. Blue Energy is the largest fuel cell energy company in the world. We have a platform for solar locksite that is best suited for stationery power generation needs. The company was started in 2008. That's when the Blue Energy was formulated before that it was Iad Americas when our chairman and founder Dr. Kair Shreeder has started his lab and started working on this technology and they got publicly listed at 2080 and in the last decade the technology has rapidly grown in terms of customer accessibility, but also in terms of technology progression. We're currently in our sixth generation of our fuel cells. There's still a long run way to go in terms of this technology. Cell cells as a technology is getting accepted by customers for on-site power generation and the way for your listeners to think about this is we are the most efficient and the clearest way of converting natural gas molecule into an electron. So that's it. This up quite nicely and I think one of the first questions that we want to get out of the way is fuel cells themselves, what is the advantage for customers to go ahead and invest in fuel cell technology as opposed to having gas turbines on site that are behind the grid? The way to think about this is here's seven facts that translate directly into value for customers. We can provide firm dedicated primarily on-site power. You run on pipeline quality natural gas, UPS quality power in terms of very high quality power that we are able to provide zero air pollution because we don't come back combust. We have no particular pattern, talks or socks and roughly 30% less carbon emissions. Extremely scalable once you understand our volume architecture. We are able in gigawart scale for rapid deployments again with sustainability benefits around low emissions and zero water usage except for start up. So once we are in our operational mode, the self-degraded water so we don't even need water. That makes our permitting very straightforward and easy and in most cases we qualify for a minor source permit. And then with all these advantages we are extremely cost competitive and can enable true time to power for our customers. So in our confess I was talking with a friend of mine a couple weeks ago and he was talking about all the work that he's doing financing data centers and he was quick to group reciprocating engines and fuel cells in the same conversation. I didn't realize how fast fuel cells were moving and relative to reciprocating engines are loud and have some of the emissions concerns that you just described that the fuel cells don't. On the speed to power question how fast can I get a fuel cell today if I were to put in an order and compare that so the turbines are waiting five years reciprocating engines are waiting 18 months. How should we think about fuel cells on the speed to power question? Yeah so I'll give an example that is available in public domain. We did a large on-site power project for a big hyper-scaler in this case being Oracle for their AI workloads. From the time they signed the contract the fuel cell getting delivered and commissioned the engine really electrons was already done in 90 days. That's a rapid turnaround I wasn't expecting that. So your fourth quarter results were announced on the 5th of February and just for our listening we're recording this on the 17th and one of the stats that you had in there that I found really, really interesting was that your backlog growth has really, really grown in the last year and I guess I wanted to ask why because this may answer the speed to customers actually getting the technology but as a remarkable thing it was like two and a half times you may please correct me if I'm wrong but what was the driving factor behind that rapid growth there? Yeah, you're correct, a backlog, backlog through the half x versus last year but his, his but I would tell you if you'd be surprised you should expect significant growth like this to cut you in years to come. But here's how fundamentally I want you to think about it, right? Many years ago maybe you was 25 years ago with power generation industry adopted wind technology at scale. And a couple of years later, it was solar. Then energy storage came out, right? And now you're starting to see fuel cells come out, come off age and starting to get a doctor. Right. I tell utilities and I tell all I could ease each one of them are going to have a bloom fuel cell in their portfolio. Give it. Take it aside. When they want to deploy us for on-site generation, behind the meter or in front of the meter and the timing of when they do the first project, it is exactly like other technologies. This technology has come off age, blue has a technology board. It's 15, 20 years ahead of anyone else. What a sixth generation and there's so much more runway on this technology. It's just fascinating. The vision of the company that our founder and chairman has is we'll be powering the planet and powering the planet goes back to the comment that I made earlier that we are the most efficient and environmentally friendly way of converting natural gas molecule into electrode. And we're already at charity with other technology, but very soon we'll also be the cheapest as we scale up. So let's talk about that if we can offer some details about it and please redirect me or direct me into some numbers. The way that I understand the call stack today, if I'm thinking about behind the meter and data center application, if I'm looking at traditional gas bar generation, either a turbine or a reset, it's somewhere on the order of $2,025 to KW. I've also got to add a battery to that to manage the volatility in the load as AI ramps up and ramps down and that's particularly challenging for a nurse with the turbine where you've got to manage the problem there. The fuel cell doesn't have that a nurse problem. The numbers I've seen on the fuel cell are more in the 5,000 plus per KW. Do I get to parity and of course fuel cell is newer technology and should appreciate call stick lines as things move forward? Am I able to strip out battery installations from my cost because of the load balancing, because of the way that the fuel cells are constructed and does that help to get me to parity or fuel cells cost falling more than I realize or both? It's both but let me do the math for you. So the 5,000 KW is a gross cost then we also qualify for tax credits and once you take half 40% of that or 30% in the new regime, so already your gross cost here at 40% is down to $3,000 a KW net. So you're starting with a cathex on turbines of about 2500 to a net cathex on a fuel cell of 3000. Then you look at it from fuel efficiency advantage, we will guarantee somewhere around mid 50s, 54, 55% life cycle efficiency over 20 year period. I repeat it's 55% over a 20 year period that captures all degradation because fuel cell starts with high 60 set efficiency. So now you look at the delta fuel advantage if a crop up recipient engines, you know, maybe they're at 40 to begin with but then over time, the recipe or gas turbine goes through significant degradation. So the delta advantage on fuel cost is pretty significant and then you brought up another very trusting aspect around our ability of clume fuel cell, we able to handle AI workloads without a battery because we have super caps that are inbuilt and integrated with our fuel cell systems. So actually we don't need a battery. So fuel cells are able to handle workload variations between minutes and hours and then super caps are there to be able to handle three seconds to seconds. So now when you do a full math, not the gross cost but the net cost, you stack up all the efficiency advantage that you have on your fuel cost. You eliminate batteries that are not here. You add ease of permitting because we're going to get the most sustainable solution that is there with an actual gas. You add time to power. Not sure, we're expensive. Post customers once they realize the math, they start to realize that we're actually much cheaper versus alternative technology. One more quick question, Sam, before I let you jump in. So if I pull batteries out of my stack, how long does it take for me to fire up a fuel cell? If I compare, so if a turbine of a reset is what a couple hours or less, is a fuel cell faster? If I've lost the battery to bridge me to that generation perspective, does this thing, once I press go, does it work or do I need time to warm up? Great question. So, fuel cells are solid oxide fuel cells internally within the hard parts. The electrochemical reaction is happening at 850 degrees C. So when we first start the fuel cell, it'll take us 12 hours to really get to that max temperature. That max temperature is our most optimal operating point. However, once we're up and running, you can turn down a fuel cell. You can bring the output all the way down to as low as 20 or 10% with no loss of efficiency or low impact on emissions. And then when you want to bring it back up, it's a matter of seconds. Think of like a gas burner in your house. All you're doing is regulating the flow of gas. It's almost instantaneous. So that's how you should think about it, right? From a compute core start, we'll tell the customer first time when we start, it takes us 12 hours to get to our operating point. But once we are there, a bit like the press the engines, how they kept warm. You can turn the fuel cell down as low as 10% and then be back up in seconds. So I want to switch gears a little bit and talk about the partnership that you had with Brookfield, which was announced in October 2025. This was a huge partnership deal. And can you give us a little background into what that means for 2026 in the future of the company, sort of securing that partnership? We're very excited about our partnership with Brookfield. We already be aware that they have a fully vertically integrated AI infrastructure fund. The first ones amongst the ordinary asset managers who've got a fully integrated fund that is investing in the entire AI value chain. And as part of that partnership, there's multiple facets to our partnership. They are financing projects, power projects, agribins that blow contracts for our end customers. That's first aspect to it. Brookfield also owns a lot of data center companies and are building their own AI factories. Brookfield is also a customer to us. They're also using our fuel cells for their own affiliates and group companies and as they create their own key factories and sovereign cloud. So that's how I would want you to think about it. They are a partner in terms of financing customer projects. They're a partner in terms of not just data centers, but other industrial affiliates that Brookfield owns who are deploying fuel cell for their own power generation needs. And then the third facet is for their own AI factories and sovereign cloud business that they are developing. They are going to be using fuel cells, bloom fuel cells for on-site power generation needs. So it's very clear the growth opportunity is there for you. Can you talk a little bit about the portfolio today and how much your installed user bases, how much sits in the US versus global and where you see things going as a percentage of behind the meter versus front-on-the-meter investment? Hilling back to your earlier question on backlog as well, look, every technology gets to that inflection point wherein technology in techno economics and market acceptability is there and blooms gotten to that point. It took company 15 years, 2008, 2024 to deploy first one and a half. Kigawatts of fuel cell, which you know coming from harbor generation and being a gastropic guy for or a half decades before that that's small lot but you have to understand that number in the context that when fuel cells were first lodged in 2008 you were talking watts, watts became kilowatts became hundreds of kilowatt became megawatt, tens of megawatt to now hundreds of megawatt so coming back 15 years for one and a half gigawatt last year our factory was that of fact sitting a gigawatt that we are commissioning this year we're doing two gigawatts and we have plans to wrap up our factory do five get which is why I gave you the example of wind and solar panels at battery storage and say hey work the scale and where we had headed I would describe it by how our chairman says we're just getting started in them so where the built the now coming back to your geographic picks outside of US one big established market that we have you've got roughly about 400 megawatts commission maybe this is a couple of hundred megawatts getting commission anytime in the first half this year or if they go up to 500 is South Korea largely because there was what country that recognized the importance of fuel cell as part of the energy transition let's slated it so there's like an end-belowction that happens and you would take you know it's like about a hundred-ish megawatt a year in the market that we participate in and then obviously US with electricity prices progressively going up with AI infrastructure investment is a pick market that is rapidly growing for us in Europe we see projects in Italy, Germany, UK still at Spogscale but again as the AI wave and the demands apply mismatch shows up which is which is inevitable as as the European countries start investing heavily on AI infrastructure the same phenomena as going to play there as well and then in Asia outside of Korea the other market where we are starting to see rapid traction is Taiwan because of the semiconductor value chain the need for power there is also urgent and their space could straight but the way I always tell our customers is we're we're we're pretty like smartphone or an iPhone if I may want to use that as an example will be everywhere I don't have to go prove myself out and have we country at the war or just go to chew it with the customers adopted in a few places or few big countries and economies and rest of the world will fall no different than we don't have to go prove battery storage in TV solution in every country right just then you scale up so that's when that's when going forward as we see our backlog growth you can come back in reference but I said every utility every IPP in US is going to be a dual customer just matter of time so one area we haven't really talked about much or yet really is is hydrogen obviously that there is a lot of appetite for hydrogen a couple of years ago and that has sort of suddenly dwindled in terms of fresh coverage and and sort of outlooks we see what is looms stance on hydrogen where opportunities do you see there within this company so as it relates to power generation a solid oxide platform fuel cells platform is a glass day if you bring hydrogen today we can use hydrogen in the fuel cell today every fuel cell that we're shipping is how is hydrogen ready fundamentally if you look at what is happening in the fuel cell electrochemical reaction what is that gas chemical formula is CH4 in fuel cell natural gases reacting with air doesn't allow nitrogen to pass which is why we don't have any oxides the compound CH4 gets broken your carton gets went dead it's really the hydrogen so there's a steam methane reformation that is happening on the surface of fuel cell it's really the hydrogen from the natural gas that we are using today create electricity in future if you inject direct hydrogen no problem even today our fuel cells can handle any mix of hydrogen 2% 5% 20% we're agnostic we get fuel cells blue fuel cells can run on natural gas can run on hydrogen can run on bio gas believe it for the customers to decide whatever fuel gives them the best certainty about availability and economics because of our high efficiency it doesn't really matter right no matter what fuel you bring we are the most efficient in converting that molecule into electron and so you spent you mentioned this a second ago but you spent I don't know 20 plus years or something at GE right and the gas turbine industry is really dominated by GE Siemens and Mitsubishi as they're as they're looking at what you're doing do expect copycats do you expect to see more people getting into the business and does does what you all do do you have a secret sauce that even if Sam or I wanted to start a you know fuel cell business today it's going to take us some time to replicate it how should we think about that yeah I think I think it's natural as bloom succeeds that there will be other companies who'll try to get it to the fuel cell business both myself and our chairman loves competition because it only it only inspires us to become even better is how I would characterize it now look fuel cells have been around right the technology is going around for 50 years but a lot of companies you mentioned electric and early 2000 had a few subdivision Siemens had a similar division and so did other OEMs they've all tried invested hundreds of millions of dollars but what able to make it work and go beyond lab scale experiments to commercial create and that's where blooms God roughly at 20th technology board so you wish everyone all the best I really like that so just looking at time we always like to sort of round these conversations off with a look forward to the next six to 12 months now obviously bloom had a remarkable 2025 so what do you think people can expect from the company or something it doesn't have to be you know any one particularly thing that you want people to just pay attention to in the next six to 12 months yeah another so I'll get to answer your question but what I want all your listeners to take off bloom as a platform a bit like smartphones or iPhones right just like when iPhones game I remember in 2009 when when I would hear talks about this gonna be accident or platform I couldn't stand as to what it means now I tell the same story to customers think of bloom as a platform wherein there are these different use cases that customers can dial in over time right and once such big use case is carbon catcher and carbon sequestration because we don't combust we have a pure stream of carbon just all part what happens is look if that's your guess is where to be the dominant fuel that powers the the energy needs of the world for next two decades if we have to all stay true to our our decarbonization and sustainability goals it has to be on the backs of carbon capture and sequestration all right that I think you will see play out in a big way in next six to twelve months and the bloom advantage is if you look at an exhaust of a gas turbine there's only about five percent carbon content the meaning ninety five percent is oxides extend because of combustion worst is that in bloom once you once you knock out water, our carbon composition is high 80% with almost 16 next lower mass flow. So you go and you mentioned there's a lot of oil and gas, people who listen to your podcast. So if you have high carbon content, less oxides, you know, like really high carbon content of high 80% 16 next lower mass flow. What that means is your ability to capture carbon form of fuel cell is extremely economically advantages. So just like we all use different apps, people, my house or uses a lot of Amazon. And sometimes I tell my wife, you know, what made Amazon success who was not just Amazon was also a smart phone. So it's a bit like that for us as the as the carbon sequestration permits start to come through, you will see fuel cells get adopted at scale for powering AI workloads and also capturing carbon and sequestring, creating a net zero power with natural gas. That is what you will see get announced this year. And then you go and do your own assessment in terms of how big that market is. So that's number one. So that's another one of those apps, right? But what we tell the customer is, hey, we're still renting your carbon today. All you're doing is because sequestration gets an April, but now just giving you that pure carbon in a team, tailpipe, capture it, treat it, and then sequester it. Similarly, the other app that sits is CHP. I told you inside a fuel cell, there's 850 degrees C reaction happening. You know, one of the other byproduct that we have is high quality steam at 350 degrees C that we meant today. We mix in air and then just went it out. So we tell the customer, hey, as your data centers have cooling needs, we can give you 350 degree steam, which is a fee byproduct that you can use to absorb and chiller for your cooling needs. Again, no change in my platform. Whenever you're ready, we'll enable it and give it the-- I already talked about AI workloads and how we can handle it without batteries. Then let's talk about DC power. Fuel cells natively produce DC power. Today, what is happening is that DC power gets converted back into AC because that's how grid boss set up. And then the power goes in. Then it gets converted back into DC because then eventually, the data center internal infrastructure needs that. As the power density of racks goes up, you've already seen in India and other hyperscalers have announced their intent to have DC racks where the only technology in the world that we go DC directly enabling a DC power drag. But that does is tremendous savings for the center customers. In terms of transformers, power distribution units, switch gear panels, UPS systems, that's just Catholics. You were asking the math that gets eliminated. So that's where we go and tell the customers that, hey, Bloom is a platform and there are these apps that sit on it. You as a customer get to dial in as to when you want to enable what aspect of this. So once you're descending math and the customers and all large hyperscalers have started to understand including big utilities, and I think fuel cells are expensive. I think they're cheaper. Once you are in set, all these attributes and what it means overall for customers own value that they can create. I'm going to piggyback on Sam's last question before I let you go. So one of the big value propositions here is also the modularity of the technology right. And if I think about some of the commercial success of batteries, the batteries have conformed to rights law where the more batteries manufactured, there's a relationship where calls are declining in a predictable way. Can we think about fuel cells? You're talking about scale and that same rights law paradigm where the more because they're modular, the more you produce, the more calls come down. And as these apps to use your word get cemented, that the cost of clients are just going to have more and more and more on a predictable basis. Yeah. Yeah, absolutely. Absolutely. You're exactly right. And while the good costs are going up because of transmission distribution, infrastructure upgrades, prices of other OEMs have gone up. So overall cost of electricity is going up. OEM as its scales, the costs are coming down. So it's the opposite of what is happening in our world. I'll explain it to you by giving an example. When you look at our most Vadella value prop and how and what Loom was doing for first 15 years of its existence, we were selling OZEC power to industrial customers and telling them that hey, I can be at parity or cheaper than the grid. Why is that? Because I eliminate the entire transition distribution value chain, right? All I need is gas. I convert it on site. It doesn't matter. You don't have to be grid connected. You can choose to be. I've been running microgrid, completely islander applications for decades. Or with the upper spore to five megawatt range, because that's the kind of power that industrial is needed. So now look at what is happening to grid. Every state, the grid cost are going up. So every year, we can't new-- we're can't see an acceleration on this concept where we can't need to become at parity or cheaper than the grid, which is also why we also announced that our conventional consumer and industrial segment is also growing rapidly, because I go to an industrial customer and say, hey, I know you don't care about load following and you don't care about all these other attributes, things that should parity use at cheaper than the grid, giving you very reliable high quality power. And by the way, I have this byproduct of the 150 degree C high-quality steam that maybe of use for you. Use that. And it's just pure economics. If you're cheaper than the grid by a couple of cents, that's just funny that they can pocket and lower their operating cost, right? So coming back to your question hill, you're exactly right. The operating leverage in this business is nothing like I've seen before. If technology as it scales up and as a factory cart use to double its production, cost, cart view to come down, bloom has an unbelievable culture of being able to take cost out. And we've proven it by taking 10% cost out year over year. And I think I foresee that trend to cart view. And rest that are tailwinds behind us, because the grid cart use become expensive. Power is out available. And all these differentiated attributes that I talk about, which is why I'm very confident we're going to be able to wear. Well, this is great. And I could keep this conversation going for another 30 to 60 to 90 minutes. But I know you're busy. We're very grateful for your time. And we'll adjourn here and hopefully welcome you back to get an update here in a few months or next year. [MUSIC PLAYING] (gentle music)