That experience changed everything. “I decided I actually really like planet Earth,” Hartwig told me, “and I wanted to focus my career instead on preserving it, rather than trying to leave it.” So he rallied a bunch of his former rocket engineer colleagues to repurpose technology they pioneered at SpaceX to build a biomass-fueled, carbon negative power source that’s supposedly about ten times smaller, twice as efficient, and eventually, one-third the cost of the industry standard for this type of plant.
Take that, all you founders humble-bragging about starting in a dingy garage.
“It’s not new science, per se,” Hartwig told me. The goal of this type of tech, called bioenergy with carbon capture and storage, is to combine biomass-based energy generation with carbon dioxide removal to achieve net negative emissions. Sounds like a dream, but actually producing power or heat from this process has so far proven too expensive to really make sense. There are only a few so-called BECCS facilities operating in the U.S. today, and they’re all just ethanol fuel refineries with carbon capture and storage technology tacked on.
But the advances in 3D printing and computer modeling that allowed the SpaceX team to build an increasingly simple and cheap rocket engine have allowed Arbor to move quickly into this new market, Hartwig explained. “A lot of the technology that we had really pioneered over the last decade — in reactor design, combustion devices, turbo machinery, all for rocket propulsion — all that technology has really quite immediate application in this space of biomass conversion and power generation.”
Arbor’s method is poised to be a whole lot sleeker and cheaper than the BECCS plants of today, enabling both more carbon sequestration and actual electricity production, all by utilizing what Hartwig fondly refers to as a “vegetarian rocket engine.” Because there’s no air in space, astronauts have to bring pure oxygen onboard, which the rocket engines use to burn fuel and propel themselves into the stratosphere and beyond. Arbor simply subs out the rocket fuel for biomass. When that biomass is combusted with pure oxygen, the resulting exhaust consists of just CO2 and water. As the exhaust cools, the water condenses out, and what’s left is a stream of pure carbon dioxide that’s ready to be injected deep underground for permanent storage. All of the energy required to operate Arbor’s system is generated by the biomass combustion itself.
“Arbor is the first to bring forward a technology that can provide clean baseload energy in a very compact form,” Clea Kolster, a partner and Head of Science at Lowercarbon Capital told me. Lowercarbon is an investor in Arbor, alongside other climate tech-focused venture capital firms including Gigascale Capital and Voyager Ventures, but the company has not yet disclosed how much it’s raised.
Last month, Arbor signed a deal with Microsoft to deliver 25,000 tons of permanent carbon dioxide removal to the tech giant starting in 2027, when the startup’s first commercial project is expected to come online. As a part of the deal, Arbor will also generate 5 megawatts of clean electricity per year, enough to power about 4,000 U.S. homes. And just a few days ago, the Department of Energy announced that Arbor is one of 11 projects to receive a combined total of $58.5 million to help develop the domestic carbon removal industry.
Arbor’s current plan is to source biomass from forestry waste, much of which is generated by forest thinning operations intended to prevent destructive wildfires. Hartwig told me that for every ton of organic waste, Arbor can produce about one megawatt hour of electricity, which is in line with current efficiency standards, plus about 1.8 tons of carbon removal. “We look at being as efficient, if not a little more efficient than a traditional bioenergy power plant that does not have carbon capture on it,” he explained.
The company’s carbon removal price targets are also extremely competitive — in the $50 to $100 per ton range, Hartwig said. Compare that to something like direct air capture, which today exceeds $600 per ton, or enhanced rock weathering, which is usually upwards of $300 per ton. “The power and carbon removal they can offer comes at prices that meet nearly unlimited demand,”Mike Schroepfer, the founder of Gigascale Capital and former CTO of Meta, told me via email. Arbor benefits from the fact that the electricity it produces and sells can help offset the cost of the carbon removal, and vice versa. So if the company succeeds in hitting its cost and efficiency targets, Hartwig said, this “quickly becomes a case for, why wouldn’t you just deploy these everywhere?”
Initial customers will likely be (no surprise here) the Microsofts, Googles and Metas of the world — hyperscalers with growing data center needs and ambitious emissions targets. “What Arbor unlocks is basically the ability for hyperscalers to stop needing to sacrifice their net zero goals for AI,” Kolster told me. And instead of languishing in the interminable grid interconnection queue, Hartwig said that providing power directly to customers could ensure rapid, early deployment. “We see it as being quicker to power behind-the-meter applications, because you don’t have to go through the process of connecting to the grid,” he told me. Long-term though, he said grid connection will be vital, since Arbor can provide baseload power whereas intermittent renewables cannot.
All of this could serve as a much cheaper alternative, to say, re-opening shuttered nuclear facilities, as Microsoft also recently committed to doing at Three Mile Island. “It’s great, we should be doing that,” Kolster said of this nuclear deal, “but there’s actually a limited pool of options to do that, and unfortunately, there is still community pushback.”
Currently, Arbor is working to build out its pilot plant in San Bernardino, California, which Hartwig told me will turn on this December. And by 2030, the company plans to have its first commercial plant operating at scale, generating 100 megawatts of electricity while removing nearly 2 megatons of CO2 every year. “To put it in perspective: In 2023, the U.S. added roughly 9 gigawatts of gas power to the grid, which generates 18 to 23 megatons of CO2 a year,” Schroepfer wrote to me. So having just one Arbor facility removing 2 megatons would make a real dent. The first plant will be located in Louisiana, where Arbor will also be working with an as-yet-unnamed partner to do the carbon storage.
The company’s carbon credits will be verified with the credit certification platform Isometric, which is also backed by Lowercarbon and thought to have the most stringent standards in the industry. Hartwig told me that Arbor worked hand-in-hand with Isometric to develop the protocol for “biogenic carbon capture and storage,” as the company is the first Isometric-approved supplier to use this standard.
But Hartwig also said that government support hasn’t yet caught up to the tech’s potential. While the Inflation Reduction Act provides direct air capture companies with $180 per ton of carbon dioxide removed, technology such as Arbor’s only qualifies for $85 per ton. It’s not nothing — more than the zero dollars enhanced rock weathering companies such as Lithos or bio-oil sequestration companies such as Charm are getting. “But at the same time, we’re treated the same as if we’re sequestering CO2 emissions from a natural gas plant or a coal plant,” Hartwig told me, as opposed to getting paid for actual CO2 removal.
“I think we are definitely going to need government procurement or involvement to actually hit one, five, 10 gigatons per year of carbon removal,” Hartwig said. Globally, scientists estimate that we’ll need up to 10 gigatons of annual CO2 removal by 2050 in order to limit global warming to 1.5 degrees Celsius. “Even at $100 per ton, 10 gigatons of carbon removal is still a pretty hefty price tag,” Hartwig told me. A $100 billion price tag, to be exact. “We definitely need more players than just Microsoft.”
Hurricane OscarAccuWeather
Ember