Airthium: industrial heat pumps and energy storage
Airthium is a greentech – deeptech startup developing three solutions revolving around a Stirling engine, a kind of reversible heat pump:
- The production of process heat up to 550°C from electricity, at a price that will eventually be competitive with natural gas.
- A seasonal electricity storage system to bridge off-peak periods in wind and solar production, combining
- short-term storage of electricity in molten salts (in the form of heat) which, thanks to the Stirling engine, would enable storage using common materials with an efficiency of 70%.
- the production of ammonia, which can then be converted into heat using a burner, and then into electricity using the Stirling engine.
The Stirling engine makes a wind- and solar-powered electricity system more reliable.
The Airthium company
Airthium was founded in March 2016 by Andreï Klochko and Franck Lahaye. The birth of Airthium is recounted in the interview given by Andreï Klochko for École polytechnique:
” The idea for Airthium germinated in 2008 when I became aware of a technology at the crossroads of fluid mechanics and electromagnetism, but it wasn’t until I had completed my thesis at the Plasma Physics Laboratory at X that I was able to apply it to energy storage by working on the design of a new type of gas compressor. In 2014, when we won the Gérondeau – Safran prize and the Concours mondial de l’innovation de Bpifrance, I realized that our project had convinced people and I realized its full potential. I then devoted myself to it full-time, and Airthium was officially created in March 2016.”
https://www.polytechnique.edu/fondation/actualites/toutes-les-actualites/airthium-la-start-qui-revolutionne-le-stockage-de-lenergie
The company was incubated by the prestigious Y Combinator in 2017. It raised €500k from several investors, including Y Combinator.
Housed atÉcole Polytechnique until 2019, the startup was until recently based at Air Liquide’s Innovation Campus, before moving this October to dedicated premises in Villebon-sur-Yvette (91).
It raised ~1.3 million euros in crowdfunding in 2021 and is currently raising a second round, again in crowdfunding (source: interview).
On the industrial front, they are planning a 1kW prototype in 2023, a heat pump demonstrator the following year and a 1MW industrial model in 2025. (https://airthium.com/about_us)
Then, larger-scale heat pumps (20MW) would be available in 2028. Seasonal energy storage is planned for 2030 on a small scale (50MW) and 2035 on a large scale (1GW).
Questions and answers
– On the principle of the Airthium solution
- Discover The Greentech (DTG): If I’ve understood correctly, you’re proposing an industrial heat pump capable of capturing heat between 100 and 500°C or atmospheric heat from -50°C and delivering it at the desired temperature, is that correct? What’s the difference between heat you can only capture at 100°C and atmospheric heat?
- Airthium: Yes, our heat pump can capture waste heat (heat produced by a process but not recovered) or atmospheric heat and deliver it at the desired temperature (up to 550°C). We are able to collect heat from -50°C up to 550°C, so there’s no fundamental difference between atmospheric and 550°C heat.
Depending on the temperature at which the heat is collected and released, the heat pump operates differently: the working fluid may change (water between 20°C and 80°C, oil up to 250°C and molten salts to reach 550°C), but it’s above all the COP (coefficient of performance) that is affected.
A resistance has a COP of 1 (1 joule of electricity = 1 joule of heat), but the advantage of the heat pump is that it has a COP greater than 1. In our case, going from 20°C to 200°C we have a COP of 1.9, but if we go from 100°C to 200°C our COP is 2.7.
- DTG: Can you explain in a few words how a heat pump works? How is it more efficient than a resistance heater? A layman would think that 1 joule of electrical energy would give 1 joule of heat. Is it more complicated than that?
- Airthium: The greater-than-1 efficiency of the heat pump may seem surprising, but it’s based on thermodynamic principles. The heat pump’s job is to move calories from the cold source to the hot source. Electricity is therefore not used directly to create heat, but to move calories. In this way, it can be more efficient than resistance heating. In fact, they are already widely used in homes and industry.
- DTG: How is heat transmitted? Air, liquid, metal?
- Airthium: Heat is captured in the cold process by a heat exchanger, which transmits it to a fluid (water, oil, molten salts) which is then sent to our heat pump. Inside, it interacts with helium gas, which transfers calories from the cold source to the hot source. We then recover another fluid (water, oil, molten salts) that has been heated by the heat pump, which is sent to the heat exchanger on the hot side to reheat the hot process.
– On uses :
- DTG: I see you’re talking about using your heat for cement. However, it seemed to me that, to remove the carbon from limestone, it had to be heated to extremely high temperatures (>1450°C). Do you go that high? The same question applies to glass and metal. For what processes and how can your process be used?
- Airthium: Our heat pump can be used in the cement, metal and glass industries, but that doesn’t mean it’s used directly in the main process (metal or glass melting, for example). These industries require moderately high temperatures for other applications, such as drying, heat treatment, temperature maintenance… This is the kind of process we can help with.
– On the energy storage system:
[Here we’ll be discussing the schematic of their seasonal storage system in particular]
- DTG: I’m having a bit of trouble understanding your diagram: why produce ammonia? Take heat from water? Is the “Airthium Stirling Engine” dependent on these two specific processes? Or are they examples of use?
- Airthium: The Airthium Stirling Engine is the machine we build, which can be used both as a heat pump and as a Stirling engine. The Stirling engine is a thermal machine that can produce heat from electricity (heat pump mode, COP up to 3.5) or electricity from heat (engine mode, thermal>electric efficiency up to 50%). We plan to start by producing heat pumps to develop our technology and make the system more reliable, before entering the energy storage market (which requires a substantial capital investment).
Our storage system works in a hybrid way: for daily storage we use molten salts or sand as purely thermal storage, but our main advantage is seasonal storage (very long duration), for which we need ammonia (NH3).
With an energy mix based on renewables, there is a risk (a few times a year) of a major drop in production (more or less sun and wind at the same time). We therefore need to compensate for this shortfall with storage capacity.
Ammonia is a synthetic fuel whose combustion generates no CO2, and is fifteen times cheaper to store than hydrogen. When there is a surplus of renewable energy production, we produce ammonia from electricity, then store it in large-capacity tanks in liquid form (it is easy to store, and there is already a whole industry that has mastered this technology). The parallel can be drawn with the natural gas reserves that are in use today. When demand exceeds production, the ammonia is burned (in a burner developed to limit NOx emissions) to heat the molten salts, which are then sent to the Stirling engine to generate electricity.
- DTG: Can you give us an outline of this system? How efficient would it be? What would be the storage medium (liquid, gas, etc.)? Does the system’s stability require any special maintenance (maintaining temperature/pressure/other conditions)?
- Airthium: The round-trip efficiency of thermal daily storage will be around 70% (compared with 80-90% for lithium-ion batteries, which discharge more quickly, use rare earths, are expensive and present a risk of thermal runaway). The fluid (molten salt or sand) is stored in thermally insulated tanks to minimize heat loss. Molten salt needs to be kept at a minimum temperature to prevent it from solidifying. This will be guaranteed by the minimum temperature of the tanks, which will be much higher than the solidification temperature, and when solar and wind power are lacking, by the combustion of ammonia.
As mentioned above, the seasonal storage medium is liquid ammonia (-33°C under 1atm), with a return electrical efficiency of close to 30%. Its storage conditions are very similar to those of propane. Ammonia storage is already an established and mature industry, which we can build on.
Discussion with CEO Andrey Klochko
I then chatted with Airthium’s CEO, Andrey Klochko, to discuss a few points in greater depth, mainly around the energy storage solution. I’ve noted in [ ] the passages I’ve taken up or summarized.
The principle of the long-term energy storage system
- DTG: So the daily storage solution would use highly available materials with an efficiency of 70%? What’s stopping this solution from being more sustainable?
- Andrei Klochko: That is, to store for more than 40 hours?
- DTG: Yes.
- Andrei Klochko: It’s just too expensive. […] Seasonal energy storage compensates for a lack of solar and wind power, which […], 2 to 4 weeks a year, […] can drop significantly over very large territories at the same time. Whether it’s the whole of France or the whole of Europe. For those times, we’ll have to keep power plants [, currently mainly gas-fired, which will also] have to recoup their fixed costs with just 1 month’s energy sales. So either they sell their energy at a very high price, or they are subsidized. [For storage], it’s all a question of price per kWh. [Lithium-ion batteries cost $200 per kWh in capital […], with molten salt it’s $60 per kWh and for ammonia, it’s something like $2 per kWh. The problem is that $2 is an optimum. If you only use ammonia, then the ammonia production equipment is too expensive and, because of the yield (30%), you’ll need a solar [or wind] field that’s too big; if you only use molten salt, […] it’s too expensive. The optimum, to reach $2, is to use a mix of the two.
In other words, molten salt storage would make it possible to smooth out electricity production in the short term to limit the power requirements of ammonia production. The great advantage of ammonia is that it’s cheap to store. When production drops, the ammonia is burned and transformed into electricity using their Stirling engine, in a process that limits nitrogen oxide (NOx) emissions.
Then there’s the question of the power of the Stirling engine. Indeed, it will be used to smooth out the production of electricity for ammonia production throughout the year, and then, one month in the year, will have to burn all or a large part of the stored ammonia.
“Ammonia is produced in dribs and drabs all year round and burned one month of the year. So there’s a power asymmetry. If we ever have 1GW of power output, then the solar field will make 3 to 5 GW peak […] and the burner will make something like 2GW thermal, but the Haber-Bosch power plants the electrolysers will make something like 600 MW. [By way of comparison, if we depended solely on ammonia, we’d need at least 3 GW of electrolysers, plus a solar field more than twice as big. All these numbers depend, of course, very much on the case under consideration].”
Limiting NOx
Using a combination of burner and Stirling engine at the end of the chain is also an advantage in terms of limiting NOx:
- DTG: In your interview for X, you talked about converting renewable fuels into electricity. Is your solution more efficient in this respect than a conventional (turbine) system?
- Andrei Klochko: The advantage is that it’s mutualized [=the Stirling engine is used both for short-term storage by heat pumping, and for seasonal storage by ammonia]. [A strength of our solution] is that there is a combination of small advantages [which add up to one big advantage]. We could have burned the ammonia in a turbine, but that [isn’t interesting], because it produces lots of NOx. We compensate for this by using an external burner [equipment that can be optimized to emit less NOx natively (before filtration), which facilitates filtration].
The trade-off between fertilizer and electricity
- DTG: What is the trade-off between using the ammonia produced to regenerate electricity and using it to make fertilizer?
- Andrei Klochko: Both work. The first few times the system is installed, it will still be expensive, so it will be installed on islands or in Alaska [for example], in [remote] places. The more we install, the lower the unit costs will be, and eventually we’ll be pretty cheap to produce fertilizer too. […] That’s why we have the heat pump value proposition in the first place, to get the cost of the Stirling engine down to a minimum and industrialize it [following the example of the automotive industry].
Water supply
Andrei also explained a little about the bottom of the Airthium storage system diagram. The energy storage system uses water in a closed cycle and is exposed to freezing if it uses air/water that is too cold (which can be problematic in very cold countries). To counter this, water is stored for cooling purposes. There would also be cooling towers, as the system would still radiate heat, especially in hot weather. As regards the need for water for ammonia synthesis, a large proportion is recovered during ammonia combustion, by condensation of the water contained in the burner exhaust stream.