🌍 Heat pumps go steampunk

Decarbonizing industrial heat with AtmosZero

Decarbonizing heavy industry is going to mainly depend on finding emissions-free mechanisms to generate heat. AtmosZero wants to start with solving for steam.


  • About 20% of global CO2 emissions could be avoided by decarbonizing heat, which makes up ~75% of all industrial sector emissions. Of that segment, ~50% of process heat is delivered by steam.
  • Most industrial facilities use fossil-fuel burning boilers to generate that heat, and while there are options to electrify or fuel-switch in order to decarbonize the process, most market-ready solutions are still more expensive than current GHG-emitting equipment.
  • AtmosZero is focused on replacing the gas boilers that create the steam to power industry. Unlike high-temperature thermal storage tech (from companies like Rondo and Antora), the Fort Collins, CO-based startup is developing heat pumps that can deliver the lower temperatures (~100 to 225 degrees Celsius) that make up much of the demand for industrial process heat with higher efficiency—and therefore lower costs—than other decarbonization solutions.
  • This solution still pulls heat from the ambient air, but compared with residential or commercial heat pumps available today, AtmosZero is developing and aiming to mass-produce much larger heat pumps that deliver higher temperatures and steam, rather than air.
  • While investments and demand for heat pumps have been growing rapidly over the last few years (VC funding for the technology increased from less than $6M in H1 2022 to more than $200M in the first six months of this year), process heating made up less than 2% of the $53 billion to $68 billion global heat pump market in 2021.
Source: GP Thiel and AK Stark. “To Decarbonize Industry, We Must Decarbonize Heat.” Joule 5(3), 2021

Last week, AtmosZero announced a $7.5M seed round, a pilot project with New Belgium, and a manufacturing partnership with HVAC giant Danfoss. Co-founder and CEO Addison Stark believes the company’s technology could displace some 2.25 gigatonnes of emissions per year. We spoke with Stark about what it will take to retire gas boilers, the necessity of mass-manufacturing for industrial heat solutions, and why no one has tried the supersized heat pump approach before now.

What’s your background? How did it lead you to start AtmosZero?

I did my PhD in mechanical engineering at MIT, where I was working on technologies to decarbonize the chemicals industry, in particular, advanced biofuels and developing advanced utilization of biomass to supply hydrogen and other reactive gasses to industry. My first job out of grad school was a fellowship at ARPA-E, where I was looking at improving efficiency of the chemicals industry—really doing industrial decarbonization before we called it that—and applying innovation to the industrial sector. That's where I met my co-founders, COO Ashwin Salvi, who actually hired me, and CTO Todd Bandhauer, who is a professor at Colorado State University and received ARPA-E funding during that time for a project in his lab.

After ARPA-E, I was the director of energy innovation policy at the Bipartisan Policy Center, where I worked on developing policies that led to the bipartisan infrastructure deal, the Energy Act of 2020, the creation of a lot of the different applied programs like [The Department of Energy’s] Office of Clean Energy Demonstrations (OCED) and the Industrial Efficiency and Decarbonization Office. But I really started to miss working on technology itself. In the pandemic, everyone took up a weird hobby, right? Mine happened to be pushing back against the concept that industry is hard to decarbonize and that we wouldn't see a lot of traction in that space until the 2040s. I ended up publishing an article in Joule in early 2021 on this with Greg Thiel, director of technology at Energy Impact Partners, essentially saying: to decarbonize industry, we need to decarbonize heat first. The article was a broader R&D roadmap, but I identified a nearer-term opportunity as I was writing it, which is going after steam boilers. So I pulled the band together—reached out to Ashwin and Todd. We wanted to build a company that can displace gigatonnes of emissions on a realistic timeframe. Steam drove the first industrial revolution, and we really do believe that by decarbonizing it, we're going to catalyze the second.

What are the current options when it comes to decarbonizing industrial heat?

Source: GP Thiel and AK Stark. “To Decarbonize Industry, We Must Decarbonize Heat.” Joule 5(3), 2021

If you're a plant manager today—let’s say a brewer—looking to decarbonize your brewing facility by eliminating emissions from your on-site boiler, the first thing that everyone wants to do is fuel-switch. You'll go out and look for [renewable natural gas (RNG)] first. The challenge is that it's a really small resource and also there's a lot of demand for it. To switch to RNG, if you can get hold of it, you're talking about a 3x to 4x increase in your cost of process heat. So that's your first dead-end. Second dead-end is hydrogen. We think hydrogen is super important for high-temperature applications or for applications where you need a reactant: steel, cement, other aspects of the chemicals industry. But it’s not really great for just burning in a boiler. Similarly, you're looking at a 4x increase in cost, at least, and availability challenges or a really big, complex upgrade of your facility to put a hydrogen generator onsite.

Facilities today are buying electric resistive or electrode boilers, which are both versions of taking electricity and resistively heating water to boil it directly. The challenge with that is long-term economics—you're signing up to pay electricity prices for process heat. Usually, that's also going to lead to a 2x to 3x increase in the per-unit of steam generation on site.

How do heat pumps fit into the equation?

We need a better solution that's higher efficiency, and therefore yielding better long-term economics for a facility, which led us to heat pumps. The challenge when you look at the market for heat pumps is they've been deployed in Europe before, but generally in one-off projects. Usually they're rather bespoke, designed around the specific temperature of waste heat available and the specific temperature needs of a facility. Everything has been engineered and designed as a project, meaning very expensive upfront capital and integration complexity.

At AtmosZero, what we recognized as a real opportunity is to deploy an air-source steam heat pump—a fully-packaged heat pump boiler that goes directly to generating steam, just like if it were a combustion boiler. Instead of tying into waste heat or ground heat or anything else that increases the complexity of a project, we're directly sourcing heat from the ambient air outside the facility. That allows us to go towards mass manufacturing scale, which gets us to the economics that have allowed building heat pumps to be so successful.

When you look at manufacturing and industrial facilities across the US and the world, 65% of facilities have a thermal demand below 10 megawatts (MW). With below 10 MW, you don't want to have a project-based approach because you're going to overwhelm the overall cost of solving that problem. You need a mass-manufactured product to solve it. What we're focused on is productizing a heat pump boiler that can replace a fossil fuel boiler at a cost competitive with gas boilers today.

What makes your technology more efficient than what’s being used in industrial settings today and how is an industrial heat pump different from heat pumps people might be familiar with in residential settings?

An electrode boiler is really what people are buying today in industry. You put one megawatt of electricity in to get one megawatt of steam out. By using a heat pump, what we're capable of doing is delivering one megawatt of steam using one-half megawatt of electricity. So for every one unit of electricity in, we get two units of heat out, and that's what drives the economics to work. That's just the reality of using a heat pump, you get this efficiency gain. That's why people are turning to them for building HVAC replacements, because of the cost benefits.

The heat pump that runs in my condo is designed to deliver heat at about 50 Celsius, transfer that heat to the air inside, and maintain around 25 degrees. But that's the upper limit of the temperature it delivers. For an industrial application, you need to go to higher temperatures. For our pilot at New Belgium, we're going to be delivering 165 degrees Celsius steam. One of the core differences there is a much higher temperature lift to be able to achieve, and the other thing that you have to do is you're delivering steam—boiled water instead of air. The last thing is they're much larger. In buildings, you're probably talking about a kilowatt-scale system. We're developing a product at roughly the 1 MW scale. Our first prototype at our pilot with New Belgium is going to be 650 kW, for example.

Tell us more about the partnership with New Belgium. What are you hoping to learn through the pilot and who are your other target customers?

New Belgium is an incredible first pilot customer. They are a nationally known craft brewery that has been very climate-conscious and climate-forward for a long time. They've gotten to the point that it's time to figure out a solution to replace gas boilers to get rid of their onsite fossil usage. They went through that similar journey that I mentioned before—looking at solutions and finding that you need something more cost-effective to do this. We are working with them at their flagship location, their headquarters in Fort Collins, where they have a boiler room of four boilers and are looking to replace one with our system. The goal is to be able to put our system in, show that we can shut down one of the gas boilers, and prove that this can be a true drop-in replacement. Then figure out how we can go scale it from there.

Brewing is a great example of the food and beverage market and how steam is used in that vertical. Steam is used to clean vessels. It's used to sterilize bottles, cans, kegs. When you process food or beer, you have to boil water for a while, so you also use steam to drive that. Steam is a big fraction of food and beverage companies’ scope one emissions, generally around half. Right now we're also actively exploring chemicals, particularly biochemicals. These are similar fermentation processes—thinking about cosmetics, separations through distillation—and then pulp and paper and other applications. Steam is used in every vertical for something. If you need process heat from 100 to 225 [degrees Celsius], you're using steam to do that. We do see food and beverage as being a very important first market, because steam is super important and the majority of onsite fossil fuel usage is steam boilers for that sector.

What does your traction look like so far? What do potential customers need to see in order to use your tech in their industrial facilities?

Since our announcement last week, we're having a lot of conversations with people who have reached out across food and beverage and in chemicals, biochemicals, and other verticals. What people need is an available drop-in solution, so the metric that we're trying to achieve is a true productized drop-in ready manufacturable solution. The pilot in New Belgium is a step on that path that allows us to prove to the world about the value of this technology and how it works. But ultimately, it's about getting to the optimized and fully-productized solution that we're continuing to do in our engineering work—keep on driving up efficiency, keep on driving down costs, keep on optimizing for longevity. Those are the things that we focus on, because at the end of the day, this is an industrial machine that is expected to live for 20 years.

What do you see as the biggest challenges for AtmosZero as you scale?

We're bringing a new capability to the boiler room that hasn't existed before, so there's a lot of interest and demand for what we're doing. For us, the challenge is engineering it correctly, developing the right partnerships and manufacturing partnerships to allow ourselves to be able to scale up to being able to deploy a gigawatt-worth of these per year. One of the other things we announced earlier this week was our manufacturing partnership with Danfoss. They supply compressors for us, and they're a world leader in compressors for refrigeration and for industrial HVAC applications, including heat pumps. We continue to look for similar supply chain partners in order to be ready to deploy. We’re focused on maximizing the utilization of the existing HVAC supply chain so we can have an impact in a relevant timeframe. At the end of the day, a lot of the components we use are to some degree off-the-shelf things—heat exchangers, compressors, valves, pipes. There's no “unobtainium” in what we do. It's all about smart engineering design and integration and packaging.

What will this funding be used for and how are you thinking about expansion?

We have raised $7.5M in seed funding from Energy Impact Partners, Starlight Ventures, and AENU, and we also have a grant from ARPA-E. We're using that money currently to focus on building a prototype at our lab. We're now in the final stages of fabrication, and we're going to be doing a lot of testing on it in the next couple of months. It’s 650 kW, so the same scale that we're going to pilot over at New Belgium. We’re doing a prototype for proof of concept, but then turning that into our testbed research facility that we’ll continue to operate to keep on improving efficiency of systems, components, things like that. We've also been growing our team to enable our technical product design and product development team to move as quickly as possible towards a productizeable and manufacturable solution.

Why hasn’t this been done before?

It's a confluence of a few different things. One of the most important things is there just hasn't been demand up to this point. When you look at trends in corporate climate goals, we've seen tech companies who don't emit a lot looking to buy a lot of carbon dioxide removal (CDR) offsets. Now we’re seeing companies that are large, and actually make physical assets, trying to decarbonize as well and meet their own climate goals. We're running into the point where the boiler needs to be reckoned with, essentially. Companies have continued goals that can’t just be met with efficiency improvements anymore. You have to start replacing your onsite fossil fuel usage. We are also starting to see the true impact of geopolitics in the natural gas supply chain like we never have. Since the invasion of Ukraine, the pandemic, ongoing tensions in Europe, we see that there's not just a climate need, but an overall energy security need to move past natural gas in Western Europe manufacturing.

The other thing is we are innovating on top of a couple of decades of continued innovation in heat pumps, compressors, refrigerants—all these other component innovations and improvements that have occurred for applications in residential and commercial HVAC heat pumps and chillers. We’re taking advantage of continuous improvement and applying it to this.

AtmosZero is still looking for one or two additional pilot projects, especially with companies working on products other than beer. If you’re interested in meeting your steam needs with heat from the air rather than fossil fuels, reach out here for a conversation or send a note to [email protected] to learn more.

Related posts