> ## Content Index > Fetch the complete content index at: https://nexi.fund/llms.txt > Use this file to discover other available public pages before exploring further. # Reverion raised $175M to build a 250 MW fuel cell factory - URL: https://nexi.fund/reverion-reversible-fuel-cell-series-b/ - Published: 2026-10-12T07:00:29.000Z - Updated: 2026-10-12T07:00:29.000Z - Description: Reverion closed a $175M Series B led by Kembara, funding a first Megafactory that takes annual manufacturing capacity from tens of megawatts to 250 MW by 2029. The 74.2% efficiency figure was measured in live customer operation at a biogas plant in Baden-Wuerttemberg. - Author: Nexi.fund Labs - Tags: Energy & Climate, #mode-5, #hook-number, #track-E, #layout-asymmetric-long A biogas plant outside Mühlacker, in Baden-Württemberg, was burning methane to make electricity. In September 2025 its operator swapped the engine for a stack of solid-oxide fuel cells the size of a shipping container. The unit came back at 74.2% electrical efficiency. The previous world record for any thermal power plant was 64.18%, set by the Keadby Unit 2 combined-cycle gas turbine in the UK on 21 May 2024\. A good modern gas plant lands closer to 60%. Hartmut Spliethoff, who runs the chair of energy systems at the Technical University of Munich, puts the practical ceiling for thermal generation at roughly 65%. Reverion's figure came from pre-series hardware in live customer operation, with carbon capture already counted in the losses. Not a laboratory result. On 29 September 2026, the company behind those stacks raised $175M. ## What the round is buying Reverion was founded in 2022 as a spin-out from TUM and now employs more than 200 people in Eresing, near Munich. Kembara, the European scale-up fund of Mundi Ventures, led the Series B. Allianz, KfW Capital, Aurum Impact and Carbon Equity joined as new investors; Extantia, Energy Impact Partners, UVC Partners, the European Innovation Council Fund, alfa8 and Possible Ventures followed on. Mercom Capital, writing on 6 October, named the destination: a new production facility with 250 MW of annual manufacturing capacity, up from tens of megawatts today, with 2029 as the target. The efficiency figure is the hook. The factory is the bet. 74.2% electrical efficiency ↑ 15.6% vs 2024 record #### Thermal plant efficiency Measured on pre-series units running at a customer biogas site in Baden-Württemberg. The prior thermal record was the 64.18% combined-cycle gas turbine at Keadby Unit 2 · *Reverion / TUM, 2025* > This level of efficiency is impressive and significantly exceeds what was previously thought possible.— Prof. Hartmut Spliethoff, Head of the Chair of Energy Systems, Technical University of Munich ## The megafactory is the actual story Hardware scale-ups fail at a predictable point: after the prototype works and before the factory works. Money raised for a pilot line is not money raised for a production line. Welding, sealing, gas routing, thermal cycling and QA at ten times the volume are different engineering problems from building unit number forty, and they decide whether a deep-tech company survives its Series B or spends it. Kembara's framing is that the technology risk now sits behind the company and the manufacturing risk sits in front of it. Units are in the field, the pipeline is pre-sold, and most of the money goes into the new site. One account puts the job creation at up to 800 roles at the German plant. 🏭 **Capacity path** Tens of megawatts of annual manufacturing today → 250 MW at the first Megafactory in Germany, targeted for 2029. A tenfold jump funded by one $175M round, against roughly $79M of total disclosed funding raised before it. ## Reversibility, and what it is actually worth A solid-oxide fuel cell runs backwards. Feed the same stack biogas or hydrogen and it makes electricity. Feed it cheap electricity and it makes hydrogen, or uses the carbon in the gas stream to make e-methane. Reverion says the switch takes under a minute, so the container covering a biogas site's demand at dusk can run as an electrolyser at noon while solar is being curtailed. Kembara sizes the prize with an installed-base argument. Germany alone has more than 6 GW of biogas-to-power combustion capacity that could be repowered. Replacing each engine with an electrochemical unit roughly doubles the electricity the site draws from the same gas, and opens a route to shift summer solar into winter through existing grid gas storage at close to zero marginal storage cost. The Reverion team call it the death of the Dunkelflaute, the German term for the dead calm when a winter week arrives with neither sun nor wind. That supply constraint is what makes the timing legible. Wood Mackenzie puts average gas turbine lead times at around five years in 2025, with prices expected to triple to $600/kW by 2027\. A fuel cell factory is a furnace-and-container supply chain, and $175M buys rate on it. 🔁 **Three revenue lines on one container** Gas-to-power at dispatchable efficiency, measured in customer operation. Power-to-gas in electrolysis mode, flipping in under a minute. Carbon removal on the captured stream, sold under multi-year offtakes rather than claimed on a slide. ## The carbon line is doing real work Frontier, the carbon-removal marketplace, has signed a $41.1M multi-year offtake for 96,000 tonnes of removal from Reverion installations, with delivery scheduled between 2027 and 2030\. The technical point behind it matters more than the headline figure. Conventional biogas carbon removal captures only the CO₂ fraction of the gas. A fuel cell splits the methane as well, so the same feedstock yields roughly double the captured carbon. Frontier puts the global ceiling at more than 2 gigatonnes a year by 2040 if that approach spread across the world's 120,000-plus biogas sites, and reports 60 signed pre-orders and 120 letters of intent against Reverion's pipeline. ## Why data centres keep appearing in the pitch Robert Trezona, climate lead at Kembara, is explicit about who the growth market is. "The growth market lies in powering data centers, where Reverion's negative carbon footprint counts as much as efficiency," he said in the round announcement. Hyperscalers need firm power they can switch on inside a procurement timetable, and a unit that can also book a removal is easier to place inside a climate review than a turbine is. This is the second dispatchable-power round in a month to land on the same data-centre thesis. As we wrote in September, TAR raised $120M at a $1B post-money valuation to build off-grid capacity for exactly that buyer. The arithmetic behind both is the same. RBC Capital Markets, compiling Department of Energy data in May 2026, put ERCOT's data-centre interconnection queue at roughly 356 GW by March, and behind-the-meter gas generation announced across the US at about 101 GW. Reverion's angle is that firm power no longer has to mean a combustion engine. ## What could break it Start with the ratio. Reverion's Series A in September 2024 was $62M, and roughly $79M has come in across every round before this one. The Series B is about 2.2 times that. This is what industrial scale-up costs, and it is also what turns a funding announcement into a balance-sheet event if the ramp slips. Industrial factory programmes of this size have a long history of slipping, and this one has not started construction yet. Then the order book. The company disclosed more than $100M of customer pre-orders at the time of the Series A, and Frontier now counts 60 signed pre-orders and 120 letters of intent. Pre-orders in emerging hardware are evidence of demand rather than revenue, and their cancellation terms rarely get published. Sixty units is also not a 250 MW line's worth of volume. Any early shortfall has to be closed by customers who have not yet signed anything. Feedstock geography is the quieter constraint. The 6 GW German repowering base is real and it is local. The company's own materials also list natural gas and hydrogen as fuels, and on natural gas the carbon-removal argument disappears entirely — the unit becomes an expensive, high-efficiency generator competing against turbines it outspends and gas engines it outlasts. And the carbon price is doing more work than a conventional power project. Roughly $428 per tonne is what $41.1M implies for 96,000 tonnes. If removal pricing or measurement rules move against the project, the line that makes the units financeable at data-centre procurement shrinks with it. 🎯 The efficiency record is real and measured in the field. The 250 MW factory is the actual bet, and it is larger than everything Reverion has built so far. The carbon-removal offtake is what converts a good generator into a financeable one at hyperscale procurement. [ Reverion closes a $175M Series B led by Kembara The company's own announcement of the round: investor syndicate, the tenfold capacity target, the 74.2% efficiency claim and the data-centre growth thesis. Reverion ](https://reverion.com/en/reverion-raises-175m-series-b-to-scale-production-of-its-dispatchable-carbon-negative-power-plants-tenfold/?ref=nexi.fund) Primary document for the round. Company figures come with the usual caveat: an efficiency record measured by the company on its own pre-series hardware. [ Why Kembara led: the Megafactory plan and the 6 GW repowering base The lead investor's own note on the round, including the capacity path to 250 MW in 2029, the next-generation 500 kW unit targeting over 80%, and the German biogas-to-power installed base. Kembara ](https://kembara.vc/reverion-series-b?ref=nexi.fund) The only place the 6 GW repowering number and the Dunkelflaute framing are written down. An investor document, so treat the forward figures as intent rather than as guidance. [ Frontier's $41M offtake for 96,000 tonnes of removal The buyer side of the carbon line: contracted tonnage, $41.1M contract value, delivery window, and the argument that capturing carbon from methane roughly doubles the removal available per unit of biogas. Frontier Climate ](https://frontierclimate.com/portfolio/reverion?ref=nexi.fund) This is the document that prices the carbon leg at roughly $428 a tonne. Everything downstream of that number depends on the removal market holding its price.