Arbor Adds Natural Gas To Rocket Engine Power Plant

Arbor Energy, the startup founded by former SpaceX engineers, has scaled its “vegetarian rocket engine” concept to an omnivorous power plant and raised a $55 million Series A to scale it. The company says its oxy-combustion system, which was made to work with biomass as fuel, will also burn natural gas and yet still capture most of the CO2 produced for storage — part of a response brought on by surging power demand from AI-hungry data centers.

The Series A round was led by Lowercarbon Capital and Voyager Ventures. Arbor’s wager is that dispatchable low-carbon power with embedded carbon capture can win contracts where only wind, solar and batteries would be insufficient to provide around-the-clock power or where biomass feedstock risks proving scarce.

From biomass to omnivore: Arbor adapts its power plant

Arbor’s early concept was to torch agricultural and woody residues in a high-temperature reactor, capturing CO2 from emissions and providing carbon-negative power — BECCS for the grid, basically. The reason the pivot to embrace natural gas is pragmatic: the refuse is local and finite, whereas pipeline gas is everywhere. With AI data centers sprouting faster than interconnection queues can be cleared, firm power is the constraint that counts.

Crucially, Arbor is not giving up on the biomass route. It argues that, fed with sustainably sourced residues and coupled to storage, the system can generate net removals. Fueled by gas, it’s a CCS plant — not carbon-negative, but possibly low-carbon if methane emissions and capture rates can be strictly regulated.

Why oxy-combustion matters for capture and efficiency

Arbor’s core trick is oxy-combustion. The hydrocarbons are reformed to syngas, which is combusted with almost pure oxygen and not with air. With no nitrogen in the flame, the exhaust is primarily CO2 and water vapor, allowing for simple separation of a very pure stream of CO2 for compression and sequestration. It’s a chemistry-forward, capture-enabling means of simplifying the process that also reduces NOx.

The trade-off, though, is energy overhead: The production of oxygen and the compression of CO2 impose a known penalty to efficiency. Industry evaluations of oxy-firing systems generally report a single-digit percentage net efficiency penalty. Whether Arbor can counterbalance any of that with thermal integration or clever engineering is at the heart of its economics, particularly when it will be vying with combined-cycle gas plants and increasingly cheaper renewables-plus-storage.

Methane leakages are the swing factor in gas CCS

Natural gas appears climate-friendly only in an ideal scenario, with limited upstream methane leaks and high capture rates at the plant. Methane has about 84 times the warming effect of CO2 over 20 years, according to the IPCC, so even small leaks could erase gas’s edge over coal on near-term warming.

Official inventory estimates had placed the average U.S. leak rate near 1%, though measurements from satellites have tallied closer to a national rate of 1.6%, and higher within certain basins; previous peer-reviewed work by Environmental Defense Fund pegged it at 2.3% throughout the supply chain.

Some analyses suggest that if capture is incomplete, sub-1% leakage could cancel out gas’s advantage over a 20-year time horizon. Arbor says it will acquire certified low-leakage gas and is aiming for a lifecycle intensity of less than 100 g CO2e/kWh — more or less in line with the EU taxonomy bar for “sustainable” electricity. To hit that figure would likely need verified low-leak gas, capture in the 90–95% range and lean energy use for oxygen production and CO2 compression.

The catalyst: demand for data centers drives firm power

AI and cloud campuses are leading to a hunger for power 24/7 that intermittent resources can’t meet by themselves. The International Energy Agency has warned that data centers’ electricity consumption globally could top 1,000 TWh by 2026, increasing steeply from recent years. Hyperscalers are pursuing 24/7 carbon-free power targets as well, which favor firm, dispatchable resources alongside wind, solar and storage. A fleetable, capture-ready plant that can site near load and be fed by a variety of fuels is conceptualized for this timeframe.

Louisiana project and carbon removal deal

Arbor says it’s also moving forward with biomass at its first commercial plant in Louisiana, including plans for a model that depends on wood not used for lumber and maintains other beneficial environmental effects.

That project is backed by a $41 million advance purchase agreement from Frontier, the Stripe- and Google-backed carbon removal buyer consortium. Under the commitment, Arbor is required to deliver 116,000 tons of long-lived CO2 removal by 2030 with strict measurement, reporting and verification in place.

The geology of the Gulf Coast is appealing — U.S. Department of Energy maps indicate vast saline formations available for such storage, and regulators are now signing off on Class VI wells in the region.

The larger questions center on biomass sourcing practices, lifecycle accounting and long-term storage durability — matters that Frontier has pushed suppliers to transparently document, and independent groups, including CarbonPlan, are pushing them to be as transparent as possible about.

What to watch next as Arbor scales its omnivore plan

Arbor’s omnivore turn steps up the proof factor. Among key milestones are third-party validation of suppliers’ gas leakage rates, demonstrated capture rates and net emissions (less than 100 grams CO2e/kWh), as well as permits and offtakes for CO2 transport and storage. Reliable procurement of biomass should be conducted to ensure no land use-related emissions.

• Third-party validation of suppliers’ gas leakage rates
• Demonstrated capture rates and net emissions below 100 g CO2e/kWh
• Permits and offtakes for CO2 transport and storage
• Reliable biomass procurement to avoid land use-related emissions

If the company can hit those boxes and deliver firmed power to data centers, then it will have built something that’s truly rare in today’s grid mix: dispatchable capacity with a credible carbon story.