[Disclosure: AgFunderNews’ parent company AgFunder is an investor in Faraday Earth]
Can a container-sized plasma reactor compete with the century-old Haber-Bosch process to fix nitrogen from the air and turn it into green ammonia at industrially relevant cost?
Faraday Earth thinks so. The US-incorporated startup, which has R&D operations in India, is using non-thermal plasma—a high-voltage electric field—to excite notoriously inert nitrogen molecules so they can react with hydrogen to form ammonia, without the high heat, high pressure, specialist catalysts, and mega-scale plants associated with conventional ammonia production.
The company claims its AI-optimized system could reach a levelized cost of around $500/ton, putting it within striking distance of fossil fuel-derived gray ammonia before factoring in the transport costs and supply chain vulnerabilities that come with shipping ammonia from centralized plants to farms and industrial users. Couple that with modular reactors designed to fit inside shipping containers, says the firm, and the economics of green ammonia start to look pretty compelling.
AgFunderNews (AFN) caught up with cofounder Debayan Saha (DS) at AgFunder’s AGM to learn more…
AFN: Give me the 30-second elevator pitch…
DS: We are here to completely flip the way ammonia has been historically produced, which is through equilibrium reactions such as Haber Bosch, which come with all the problems that you can think of including high costs, extreme vulnerability to supply chain disruptions due to growing geopolitical tensions, and large amounts of CO2 emissions.
AFN: How are you getting nitrogen and hydrogen to combine to make ammonia without using high heat and pressure?
DS: We are using the power of non-thermal plasma, which means you generate an electromagnetic field and you try to vibrate or excite [Nitrogen] electrons in such a way that they become reactive and are willing to form bonds with hydrogen to make ammonia. But the bulk volume of this gas is still near room temperature and pressure, it’s not at elevated temperature and pressure like that of Haber Bosch.
AFN: So you’re exciting nitrogen, if you like, to make it reactive?
DS: We’re exciting electrons of the nitrogen to make them reactive, rather than heating them at 700 kelvin or 250 atmospheric pressure to break the [Nitrogen] bonds.
AFN: How does your approach compare to that of others making green ammonia?
DS: There are several pathways to produce green ammonia. You can make it through greener Haber Bosch [using renewable power and green hydrogen], but you need renewable electricity, which is generally intermittent, so you’d need batteries and hydrogen storage alongside electrolyzers [to manage intermittency], which increases the capex, and that’s why this is very costly.
Other technologies have come up in recent years such as lithium-mediated approaches [which use lithium in the nitrogen reduction process]. But lithium is facing a huge challenge from the EV industry, so it’s a scarce resource, and there are also stability issues with lithium.
Electrochemical ways of making ammonia have also been tried, but you have to first dissolve the nitrogen into aqueous media [water-based solution] before the other reactions can happen, and that dissolution itself is very challenging because nitrogen is not very soluble.
By contrast, we are directly converting gaseous nitrogen into ammonia, so we don’t have that problem.
There are other companies who have tried to do the nitrogen to ammonia route through non-thermal plasma. But they were first trying to convert the nitrogen into nitrates [rather than directly reducing nitrogen to ammonia], basically oxidizing the nitrogen, and then reducing it back to ammonia, and ended up spending more energy and making the system more complex.
AFN: Where is your hydrogen coming from?
DS: Currently the hydrogen is coming from an electrochemical reaction [electrolysis] from water, but as natural hydrogen and other cheap sources of hydrogen start coming in, we can exploit those sources, especially natural hydrogen [naturally occurring geological hydrogen].
For other companies, it would be difficult for them to capitalize on natural hydrogen, because the sources are at distant locations in limited volumes. So you cannot place a Haber Bosch system there, because you need large systems for Haber Bosch economics. But we are very well placed for those kind of systems [owing to Faraday’s modular, distributed model using reactors in shipping containers].
Likewise, as we’re not primarily dependent on [highly specialized nitrogen-activation] catalysts and things like that, we can deal with impure forms of hydrogen, which are very low cost, but which will render the catalysts of our competitors ineffective. We don’t have that problem because of the way we are doing it. For us, the non-thermal plasma acts like a catalyst.
AFN: You’ve been using AI and a digital twin to optimize plasma. How does this work?
DS: We knew the Google DeepMind team and we knew they were working on fusion plasma and heavily using AI models. So we got inspiration from there and started using physics-inspired neural networks and ML models.
We are generating a plethora of data so we train models with our reactor-specific proprietary data, and they learn the physics, which help optimize the plasma. This is a huge time-saver for us, and is helping us reduce the price of ammonia substantially.
AFN: What levers can you pull to optimize the reaction?
DS: In plasma, in every nanosecond, a billion reaction dynamics are changing, so it’s very complex, but because of this AI layer that we have got in between, we can understand the ideal voltage, current, flow rate of the gas, and the ideal combination of these.
AFN: Who is your customer?
DS: Our modular decentralized units can be used at farmgate or at chemical plants that need ammonia as a feedstock. They can also be used for supplying ammonia to cold storage industries.
AFN: What progress have you made so far?
DS: We started the company a couple of years back and had a miniaturized proof of concept, and from there we have grown significantly. We now have a demo unit producing green ammonia and we are sharing that with customers and potential customers.
Based on that, we’ve got a good amount of industry traction in terms of pilot agreement offers and potential offtake, and the next stage is to scale it up towards a pilot demo. Once we do that, we will start generating revenues.
AFN: How easy is your tech to operate?
DS: It’s a completely turnkey system, which means you can switch it on and off whenever you want, unlike Haber Bosch, where there is a warm-up time. We don’t have that, which means we can directly match the dynamic profile of the renewable electricity from solar or wind. Our system is also very easy to handle [for customers].
AFN: Your target is $500 a ton levelized cost. How achievable is that?
DS: It’s quite achievable. We are already very close to that number, which is the manufacturing cost of ammonia today. But by the time [ammonia produced via Haber Bosch] reaches end customers, the prices are already higher because of all the transportation costs.
We are already matching the price of ammonia for end customers [due to Faraday’s localized production model], and hopefully with the pace at which we are progressing with our AI model, we should be able to match the manufacturing cost of ammonia pretty soon.
At the theoretical limit of fixating nitrogen using plasma, the energy required would be 2.5 times less than that of Haber Bosch, so the potential is huge.
Further reading:
US Senate fertilizer hearing: Farmers ‘can’t take much more’ as prices surge
Brazil Potash makes the case for local supply as Iran war roils fertilizer markets
Persian Gulf fertilizer crisis: Global food prices could rise 12-18% by the end of 2026, warns Helios AI
Guest article: Food’s fossil reckoning; energy crises are the new normal, and food is next
