You’re out of free articles.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
Sign In or Create an Account.
By continuing, you agree to the Terms of Service and acknowledge our Privacy Policy
Welcome to Heatmap
Thank you for registering with Heatmap. Climate change is one of the greatest challenges of our lives, a force reshaping our economy, our politics, and our culture. We hope to be your trusted, friendly, and insightful guide to that transformation. Please enjoy your free articles. You can check your profile here .
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Subscribe to get unlimited Access
Hey, you are out of free articles but you are only a few clicks away from full access. Subscribe below and take advantage of our introductory offer.
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Create Your Account
Please Enter Your Password
Forgot your password?
Please enter the email address you use for your account so we can send you a link to reset your password:
Implementing the new rules could mean reshaping the entire U.S. energy system.

The most generous, lucrative, and all-around lavish subsidy in President Joe Biden’s climate law, the Inflation Reduction Act, is the new tax credit for clean hydrogen production. Under the policy, a company can get a bounty of up to $3 for each kilogram of hydrogen made with clean electricity that it produces and sells. There are few legal limits to what a company can earn.
So it figures, then, that this subsidy has been the subject of maybe the most acrimonious, dramatic, hair-tearing fight over the law so far, one that saw snoozy lobbyists and power plant operators take out Spotify spots and full-page New York Times ads in order to make their point.
On Friday, the first phase of that battle ended — and the side supported by most environmental groups claimed a provisional victory. The Biden administration proposed strict rules governing the tax credit, designed to ensure that only zero-carbon electricity meeting rigorous standards can be used to make subsidized hydrogen. The rules, which some industry groups allege could stunt the field in its infancy, will have far-reaching consequences not only for hydrogen itself, but for how America’s power grid prepares for an age of abundant, zero-carbon electricity. It will create a system for organizing clean electricity that could soon determine how companies, consumers, and the federal government buy and sell that electricity — even when it has nothing to do with hydrogen.
But all of that is in the future. Now, to get the highest value of the tax credit, companies must — like other subsidies in the law — demonstrate that they paid a prevailing wage and took advantage of local apprenticeship programs.
They also must demonstrate that they used clean, zero-carbon electricity to power their electrolyzers, the energy-hungry machines that pull hydrogen out of water or other molecules. And defining clean electricity has proven to be an enormous challenge. However the Biden administration chose to define it, someone was going to be left out — or let in.
Consider just one hypothetical. Pretend you own a fancy new electrolyzer. If you buy power for it from a wind farm that’s already hooked up to the grid, then another power plant will have to replace the electrons that you’re now using. That marginal electricity will probably have to come from a coal or natural gas power plant, meaning that it will need to burn extra fuel, meaning it will release extra carbon pollution. Does that mean that the electricity that you bought is actually clean? And if not, do you still get the tax credit?
Earlier this year, climate groups proposed that any clean electricity used to make hydrogen had to meet three requirements: It had to come from a truly new source of power on the grid; it had to generate power at the same time that it was used; and it had to be produced on essentially the same grid where it was used. The Biden administration largely adopted those requirements in Friday’s proposal. On a briefing call with reporters ahead of the rule's release, Deputy Secretary of the Treasury Wally Adeyemo was effusive about the new rule’s benefits. “We’ve developed a structure that will drive innovation and create good-paying jobs in this emerging industry while strengthening our energy security and reducing emissions in hard-to-transition sectors of the economy,” he said.
Not everyone feels that way. Senator Joe Manchin, who provided a key vote for the IRA, told Bloomberg that the draft is “horrible” and promised that “we are fighting it.”
“It doesn’t do anything the bill does. They basically made it 10 times more stringent for hydrogen,” he said. The trade group for the nuclear industry has also expressed its “disappointment,” arguing, more or less correctly, that the proposal “effectively eliminates all existing clean energy from qualifying” for the credit.
But debate about the proposal has not quite run on green vs. industry lines. Air Products, the world’s largest hydrogen producer, has backed the administration’s approach, as have half a dozen other hydrogen companies. So has Synergetic, a hydrogen developer that recently left the trade group the American Clean Power Association to protest its laxer stance. “Consumer groups are behind these rules, and environmental justice has also come out to express support,” Rachel Fakhry, a policy director at the Natural Resource Defense Council, told me.
The excessive focus on the hydrogen tax credit has been, in one sense, surprising. If you care most about cutting carbon pollution in the near-term, the hydrogen tax credit is unlikely to be the most important part of the IRA. Other policies — such as the clean electricity tax credit, which could add vast amounts of new wind and solar to the grid, or new subsidies for electric vehicles — will likely reduce greenhouse gas pollution by far more in the next decade.
But a clean hydrogen industry could soon be crucial to the climate fight. Hydrogen could eventually be used to fuel medium- and heavy-duty trucks, which are responsible for roughly a quarter of the country’s transportation emissions.
It could also decarbonize the production of steel, chemicals, and fertilizer, all of which require fossil fuels today. These are a looming climate problem: By the middle of this decade, heavy industry will pollute the climate more than any other sector of the American economy, according to the Rhodium Group, an independent research firm.
Yet this does not explain why the hydrogen tax credit attracted so much attention. It became a big fight, in short, because it stood the biggest chance of backfiring. Because the tax credit is so generous, incentivizing hydrogen companies to use more and more power, it risked gobbling up too much electricity and distorting the country’s power markets. In the disaster-movie scenario, the tax credit could wind up like the federal government’s ethanol subsidies, which have cost billions while doing nothing to help the climate.
The hydrogen tax credit “has been the most challenging piece of policy that we’ve had to contend with,” John Podesta, the White House adviser in charge of implementing the IRA, told me on the sidelines of COP28 in Dubai earlier this month.
He described the administration as balancing between two extremes. On the one hand, overly strict rules could cause companies to invest more in so-called “blue hydrogen,” which is produced by separating natural gas and capturing the resulting carbon. Yet overly loose rules could cause emissions to balloon and power prices to soar.
“We could kind of blow it in either direction, I think,” he said.
This hasn’t always been seen as a problem. Since the IRA passed last year, the clean hydrogen tax credit has stood out for its extreme generosity, which goes far beyond what is contemplated by other tax credits in the law.
Once the Treasury Department decides that a hydrogen project qualifies for the tax credit, for instance, then that project can receive credits for the next 10 years. For five of those years, it can even get that money as a direct payment from the government, rather than as a tax cut. What’s more, projects can qualify for the tax credit as long as they begin construction by 2033. That means the tax credit will still be used well into the 2040s, even if Congress does not extend it.
Almost no other policy in the law spends federal dollars so lavishly or directly. Manchin, who negotiated the final text of the IRA with Senate Majority Leader Chuck Schumer, has long championed the hydrogen industry and seen it as a way to use fossil-fuel assets, such as pipelines, in the energy transition.
Soon after the IRA passed, however, climate advocates realized that this generosity could pose risks to the rest of the law. In the summer of 2022, Wilson Ricks, an engineering Ph.D. student at Princeton, was interning for the Department of Energy, studying how to measure the climate impact of hydrogen produced by electrolysis.
Ricks had already concluded that the “lifecycle” of the electricity used to make hydrogen mattered: If electricity from a nuclear power plant was sent to an electrolyzer instead of the power grid, thereby forcing a natural-gas plant to turn on and send power to the grid instead, then so-called “clean hydrogen” could actually result in more climate pollution than the traditional approach of using natural gas to make hydrogen.
Then the IRA passed, and “potentially hundreds of billions of dollars hinged on that question,” he told me. In January, Ricks and his colleagues at Princeton’s ZERO Lab published a study urging the Biden administration to adopt stringent guidelines for the tax credit. Without hourly matching, they concluded, the subsidy could wreak havoc in the country’s electricity markets.
Ricks wasn’t the only expert suddenly worried about what a giant new hydrogen subsidy could do to electricity markets. Nearly a year earlier, Taylor Sloane, an energy developer for the utility and power company AES, virtually predicted the hydrogen fight in a Medium post.
“The reason it matters that we get these rules right is that we don’t want to have an environmental backlash against green hydrogen in a few years demonstrating how it actually increases emissions,” he wrote. “Getting the rules right from the start will ensure more stable long-term growth of green hydrogen.”
Ultimately, the administration decided that nearly all clean electricity used to produce hydrogen must meet three requirements — largely inherited from the climate groups’ proposals. They also mirror hydrogen regulations already adopted in the European Union.
First, the electricity must come from a relatively new source of zero-carbon power, such as a wind or nuclear plant: You can’t use electrons that once would have powered homes or cars to power an electrolyzer.
Second, the electricity must be produced at roughly the same time that it is used to make hydrogen: You can’t buy cheap solar power at noon and claim that you’re using it to make hydrogen at midnight.
Finally, the electricity must have been made on the same power grid that the electrolyzer itself is using: You can’t buy wind power in Iowa and claim that you’re using it to make hydrogen in Massachusetts.
Today, no power company in the country has a way of certifying that its electricity meets all three requirements of the new hydrogen rule — and none has any way of selling it, either. So the rules also require local power grids to set up and sell “energy attribute certificates,” or EACs, which certify that a given kilowatt-hour of electricity was produced on a certain grid, at a certain time, and using a certain source of clean energy.
Utilities and grid managers have until 2028 to launch this new system; until then, hydrogen companies can keep using the existing system of renewable energy credits, or RECs, which certify only that zero-carbon electricity was generated during a certain year.
Although this new system of EACs may sound like so much bureaucratic legerdemain, it could eventually become more important than the hydrogen tax credit itself, because it could all but reshape how the country’s electricity systems work.
Right now, even though the availability of clean energy rises and falls throughout the day — solar panels make more power at noon than at midnight, for instance — there is no way to buy or sell claims to that power. By creating a systematic way to describe and sell an hour of clean electricity, EACs could actually create a market for 24/7 clean electricity.
The existence of that system could alter corporate sustainability pledges, climate-friendly government orders, and even how companies measure their own progress toward meeting their Paris Agreement goals. Even though hundreds of American companies say that they buy their electricity from zero-carbon sources, only Google, Microsoft, and a few other companies have committed to buying 24/7 clean electricity.
“I know the administration faced absurd amounts of pressure given how lucrative this is,” Ricks told me. “But it seems like they pretty much held firm and went with the science.”
That said, the proposal kicks two issues down the road. It asks companies whether it should allow any exceptions to the general rule requiring that clean electricity come from clean sources. Some nuclear power plant operators, for instance, have argued that electricity from a nuclear plant should count toward the credit if the plant would otherwise be slated to shut down.
That decision could shape other administration priorities. Two of the government’s seven proposed “hydrogen hubs,” new industrial facilities funded by the bipartisan infrastructure law, are planning to use nuclear power to generate clean hydrogen. Under the current rules, these hubs may not qualify for the generous hydrogen tax credit, even though they could still earn billions in other subsidies.
The proposal also asks for advice about how to count so-called renewable natural gas, which is captured methane released from cows or landfills. Some environmentalists worry that the rules for this technology, if poorly drafted, could allow companies to engage in aggressive carbon accounting that does not align with reality. But so far, the Biden administration seems to have little appetite for that approach.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
Plus, Google and Amazon report on what hyperscaling has done to their emissions.
There’s an interesting new report out today from the progressive think tank Groundwork Collaborative that makes a case for how Democrats can harness the artificial intelligence and data center boom to help the power grid — while also cutting costs for electricity customers.
But first, some news. We’ve known for some time now that artificial intelligence is transforming America’s biggest technology companies, turning them into major energy consumers and even quasi-industrial firms. Now we have even more evidence that it’s driving up their carbon emissions, too.
Google and Amazon released their annual sustainability reports yesterday, and both show huge surges in their energy use and climate pollution. Google’s greenhouse gas pollution grew by 18% last year, its largest year-over-year jump on record, and its energy use leapt by 37%. The company’s energy use rose by more than a quarter last year; it now uses roughly 3.5 times as much energy as it did before the pandemic.
Amazon’s climate pollution, meanwhile, increased by more than 16%, surging by the equivalent of more than 10 million metric tons of carbon dioxide. Emissions from its purchased electricity increased 34% since last year. If you feel like you’re seeing more Rivian-made Amazon delivery vans on the road, you’re not wrong: The company claims it deployed an additional 21,000 last year.
What’s driving this surge? The AI boom, of course. “Our AI infrastructure buildout is currently accelerating faster than the grid is decarbonizing,” Kate Brandt, Google’s chief sustainability officer, said in a statement.
What to do about it? That’s what Groundwork’s report is about.
“How do we bring down costs now? How do we bring down costs in the long term? And how can we make those two things mutually reinforcing?” Grayson Flood, the report’s author and a former policy adviser to Representative Alexandria Ocasio-Cortez, told me. “We wanted to be pretty direct about addressing what we see as a broken incentive structure within the system, particularly for interregional transmission.”
The report outlines a few novel ideas about how to lower prices immediately, in part to get through a coming multi-year “crunch,” during which the power grid in some regions will be maximally constrained while utilities work to bring new power plants online:
The report also imagines several policy ideas to help build out the grid. One of them is a Grid Trust Fund, a new federal bank account funded through an excise tax on data centers and other large electricity loads.
The government has often turned to funds like these to support infrastructure that creates a natural monopoly at national scale, Flood said. “The interstate highway is the most notorious example, but you can look at airports, you can look at seaports — they have these types of trust funds. There’s a lot of precedent for this in the tax code, and they tend to be financed with excise taxes on some sort of corresponding usage of the infrastructure.”
Under his scheme, the new excise tax would fall on big power users like data centers or crypto miners that don’t generate many permanent local jobs — in other words, aluminum smelters, steel mills, and semiconductor fabs would be exempt from it. But even just taxing electricity for large loads at 1 or 1.5 cents per kilowatt-hour, he said, could throw off more than $100 billion in a decade. That money could then be used to fund new transmission projects, technical assistance for utilities, ratepayer relief, or economic development.
That trust fund would be partly overseen by a National Power Authority, a new government corporation modeled on the Tennessee Valley Authority or the Energy Department’s existing power marketing administrations. This authority would have limited powers and would be partly inspired by Texas’ successful effort to centrally plan transmission lines in order to expand its electricity market.
The new authority would plan and develop interregional transmission, linking far-flung regions of the country to create new power markets. It would also have the power to build new 24/7 zero-carbon electricity power plants with high up-front capital costs, such as new geothermal projects, offshore wind farms, or nuclear plants.
“People talk about the power grid as a platform,” Flood said. But “right now, the grid is not functioning as a backbone and platform, it’s functioning as a bottleneck.”
The goal of the report, he said, is to ask: “How do we build [the power grid] as a backbone to support the growth of private markets, whether that’s in renewable energy generation, or an AI data center, or a new hospital that’s showing up?”
It’s an interesting document. Many energy wonks have proposed plans to shift some of the costs of expanding the electricity system out of the ratebase — that is, out of customers’ power bills — and onto the tax base, which is funded in a more progressive way. (I recently argued for a national, publicly funded grid buildout in The New York Times.) The new Groundwork report, in essence, tries to reframe those ideas for an era of populist politics — and one in which Americans are suspicious of data centers, as Heatmap’s polling has shown.
In its fusion of populist and pro-growth attitudes, this new set of proposals reminds me of New York City Mayor Zohran Mamdani’s attempt to freeze the rent for some tenants while passing major supply-side reforms allowing new housing construction. That effort has won Mamdani praise from many housing advocates in New York (even as some remain dubious about his de facto rent freeze). Whether that kind of politics works at a national level remains to be seen.
The bill is part of a package now sitting on Governor Mikie Sherrill’s desk.
Data center politics are continuing to evolve rapidly, and almost always in the direction of increasing costs and restrictions for data center development.
In New Jersey, which has become ground zero for the political backlash to high electricity prices, a gaggle of bills relating to data centers and electricity prices just hit the desk of newly elected Governor Mikie Sherrill, including a large load tariff bill, a water and energy reporting bill, and a bill to scale back tax credits available to data center projects.
All of these pieces of legislation are consistent with national and local trends (federal regulators are encouraging regional electricity markets to come up with large load tariffs, for example), with tax credits getting an especially close look in statehouses across the country.
Thirty-eight states have “ dedicated tax incentives for data centers,” according to an April study by the National Conference on State Legislatures. These often include exemptions from sales taxes for data center equipment like servers and routers, or property tax abatements for newly constructed data centers.
In Virginia, which last year elected Sherrill’s former House colleague Abigail Spanberger as governor, the sales tax exemption has become a hot issue of political contestation, as powerful Virginia State Senator Louise Lucas has come out in opposition to it. A budget deal recently reached in the state’s General Assembly included a tax on data center electricity consumption, while the data center tax exemption question will be kicked to a working group for now, according to the Virginia Mercury.
The New Jersey bill currently on the governor’s desk targets a tax credit program called Next New Jersey, which has some $500 million to disburse for tax credits. Half of that has been allocated for a CoreWeave data center project on the site of an existing laboratory, State Senator Joseph Cryan told me. The remaining $250 million would be used to bolster a number of existing state programs.
“The reason for eliminating it was, frankly, because people are outraged over the amount of money CoreWeave got,” Cryan said.
CoreWeave did not respond to a request for comment. A Sherill spokesperson didn’t comment on the record about when or whether the bills would be signed.
New Jersey and Virginia’s examination of tax credits comes after another state with a Democratic governor, Illinois, paused tax incentives for data centers that had been worth almost $1 billion in the first five years of this decade.
The turn against tax incentives for data centers comes as the public is increasingly wary of the latter and their perceived effect on electricity prices. This turn in sentiment has forced governors — like, say, Indiana Governor Mike Braun — to pivot away from their typical cheerleading for new businesses.
“States are very focused on attracting industries of the future, attracting jobs for their residents, attracting business,” Justin Balik, a former economic development official in New Jersey and vice president for states at the climate group Evergreen Action, told me. But, he asked, “Does the economic development strategy for a state reflect its other policy priorities?”
New Jersey itself is an example of how quickly the politics of economic development can turn. When the bill establishing the Next New Jersey program passed in 2024, then-Governor Phil Murphy trumpeted the bill for “capitalizing on this moment to ensure we establish ourselves as a frontrunner in generative AI innovation.”
“AI has already started to revolutionize our everyday lives, and New Jersey is capitalizing on this moment to ensure we establish ourselves as a frontrunner in generative AI innovation,” Murphy said in a statement typical of the more boosterist era of, uhhh, two years ago. “AI will be a transformative industry that will change lives and grow our economy and New Jersey is ready to take the lead.”
That was in July 2024. Now it’s July 2026. Electricity bills in New Jersey have gone up from $108 per month in May 2024 to $140 this past May, according to the Heatmap-MIT Electricity Price Hub, while rates have gone up some 38%. And while it’s often difficult to attribute electricity rate hikes directly to data center development — or even determine whether data centers raise rates at all — New Jersey, which is part of the PJM Interconnection electricity market, is almost certainly seeing hikes due to data center construction. PJM has struggled to bring on new generation or adequate transmission, and its own market monitor said in March that “data center load growth is the primary reason for recent and expected capacity market conditions, including total forecast load growth, the tight supply and demand balance, and high prices.”
The conditions have forced lawmakers to reconsider their typical bias toward economic development, Balik told me. “I think we’re seeing a moment where there’s a reckoning with the energy affordability conversation,” he said, “Where folks are rightly saying, hey, we care about clean energy in some cases, and in a lot of cases we care about energy affordability. Does our economic development strategy match those priorities, or are these two things at odds with each other?”
Cryan, the state senator, put it more bluntly: “The reason for doing it was to show the public that we hear their outrage and can do something about it,” he said. “The governor and the legislature have heard the voices of the people of New Jersey.”
What the heck is “surficial mineralization”?
According to one of the world’s leading carbon removal buyers, the sector’s future lies in piles of industrial waste.
When Frontier, the Stripe-led coalition of carbon removal supporters, announced its latest $915 million funding commitment, it took the opportunity to lay out the five technologies it views as most promising. I was familiar with four of them — ocean alkalinity enhancement, biomass carbon removal and storage, enhanced rock weathering, and direct air capture. Heatmap has covered them all. But the name on the very top of the list stumped me: surficial mineralization.
It sounds technical, and like all methods of carbon removal, it is — sort of. The idea is to take advantage of the tailings ponds and slag heaps left behind by the mining and steelmaking industries. These piles of calcium- or magnesium-rich debris naturally capture and store carbon from the air — not enough to change the trajectory of our warming planet without any human intervention, but managed well, they could one day capture carbon at a significant scale.
How significant, exactly? While there’s been very little action in the space to date, Frontier says surficial mineralization has the potential to remove over 10 gigatons of carbon from the atmosphere per year — as much or more than any other pathway — at an eventual cost of $80 to $120 per ton. That would put it among the cheapest approaches on Frontier’s list, in part because those heaps of industrial waste alone could absorb anywhere from a gigaton to 4 gigatons of carbon before there’s a need to mine rocks solely for carbon removal purposes.
“The beauty of surficial mineralization is twofold,” Hannah Bebbington Valori, who heads the Frontier coalition, told me. “One, we are working with an abundant source of highly reactive rock, and so there is a significant opportunity for carbon dioxide drawdown. And two, it is carbonating in place, and so sufficient mineralization technologies can be considered closed system approaches, and have generally more straightforward measurement reporting and verification infrastructure.”
At a chemical level, the process resembles other carbon removal pathways Frontier champions, such as enhanced rock weathering and ocean alkalinity enhancement. All three rely on alkaline minerals reacting with moisture and ambient carbon dioxide to form stable carbonate compounds that permanently lock away the gas. The difference is exactly where this reaction takes place: While surficial mineralization contains it to waste piles at industrial sites, the other approaches disperse the reaction across open, difficult-to-monitor systems such as farmland soils and the ocean.
That makes measurement, reporting, and verification — known as MRV — far more challenging and expensive for ocean- and soil-based systems, as scientists must track carbon uptake across ecologically complex environments where countless biological and chemical processes are unfolding simultaneously. These intersecting processes makes it difficult to demonstrate that human intervention was responsible for any given ton of carbon removed, as opposed to natural variability. MRV for these pathways thus relies heavily on modeling, which can never provide the same level of certainty as direct measurement.
Surficial mineralization, however, can be measured much more directly. On-site sensors continuously monitor CO2 concentrations above mine tailings or steel slag, providing a real-time signal of how quickly and to what degree the materials are drawing down carbon. Scientists can then validate these measurements in the lab by comparing physical samples of the material taken before and after the reaction, quantifying exactly how much solid carbonate formed as a result of various engineered interventions. The primary tool for this is X-ray diffraction — a well-established geological technique that identifies a sample’s mineral composition like a chemical fingerprint, making it possible to directly measure how much carbon the material locked away.
Don’t mistake the relative simplicity of the MRV framework for evidence that surficial mineralization is a proven carbon removal pathway — the reality is far from it. While mineralization may look simpler than, say, direct air capture, which typically uses giant fans and specialized sorbents to pull CO2 from the air, there are very few companies working in this space today. All are extremely early stage, and the time and capital required to secure feedstock partnerships, gain site access, and acquire necessary industrial equipment remain significant barriers to getting these projects off the ground.
Why is this heavy equipment needed in the first place? Because these waste piles won’t do much carbon capture work if they’re simply left untouched. That’s because the minerals at the pile’s surface will begin to slowly carbonate, eventually becoming fully saturated and acting as a seal that blocks carbon from reaching the reactive minerals below. As yet there’s no consensus on how to most quickly and cost-effectively break through this natural process to maximize carbon uptake — companies are testing a range of approaches, from crushing and spreading material to maximize air exposure (similar to enhanced rock weathering) to actively churning piles of waste to constantly reveal fresh reactive surfaces.
“Understanding exactly what is the best system to use to maximize your carbon removal efficiency and minimize your cost — this is what we need real-world deployment to do, and to understand,” Bebbington Valori told me.
One of the seed-stage startups Frontier has supported with a small pre-purchase agreement, Arca, spun out of the University of British Columbia to commercialize its approach to carbon removal from mine tailings. The company’s focus is ultramafic waste — magnesium- and iron-rich rock that locks away carbon dioxide as stable magnesium carbonate. “My pathway for interest on that was knowing that there was already about 2 billion tons of ultramafic mine waste sitting on the surface of the Earth in Canada alone,” Greg Dipple, Arca’s co-founder and head of science, told me.
Arca proposes to increase the surface area available for carbon capture in two ways. The first is by using customized robots to continuously till and churn tailings piles, constantly exposing fresh feedstock to the air to maximize carbon uptake before the next layer of tailings is deposited on top. That strategy, Dipple told me, “can give us a five- to 10-fold increase in the rate of CO2 capture” at active mine sites.
It successfully demonstrated this approach in an 18-month pilot project with Australian mining giant BHP at an active mine in the country's Northern Goldfields region where Arca says it increased the tailings’ mineralization rate by an order of magnitude. But the startup plans to push the efficacy of its tech further through what it calls “mineral activation.” This technique uses industrial-scale microwaves to heat the minerals rapidly enough to drive off the water that’s chemically bound within their crystal structure. This essentially blows apart the minerals from the inside out, exposing fresh magnesium-rich surfaces primed to react with carbon dioxide. The expected result is faster mineralization and more carbon captured per ton of mine tailings — but the startup has yet to test it in the field.
“Essentially we’re making microwave popcorn out of silicate minerals,” Dipple explained. “The microwaves cause the release of that water in the same way that when you make popcorn, you’re essentially boiling the water out of the center of the kernel, and that’s what blows the kernel up and creates this high surface area.” The idea is to eventually integrate this step into the mine’s tailings processing stream, with minerals moving through the giant microwave before they’re deposited at the storage facility.
Dipple told me that mineral activation will be a core part of Arca’s future projects, including those intended to fulfill the company’s 10-year carbon removal offtake agreement with Microsoft. Signed last October, the deal calls for Arca to deliver nearly 300,000 metric tons of carbon removal to the software giant.
While no other startup in the space has landed an offtake agreement of that scale, several have secured early backing from Frontier through pre-purchase agreements. One of them, Karbonetiq, is working to capture carbon from steel slag, the calcium-rich byproduct of steel production that accumulates in large piles at processing sites. Like the magnesium-rich minerals in mine tailings, calcium compounds in steel slag naturally react with moisture and carbon dioxide to form a stable calcium carbonate — a.k.a. limestone — permanently locking up the CO2.
Unlike mine tailings however, slag doesn’t begin as a fine powder. Instead, the molten byproducts poured off from high-temperature steel furnaces cool into chunks the size of large rocks, leaving only their outer surfaces exposed to the air and able to react with CO2. Karbonetiq’s strategy is essentially to crush and disperse those rocks to increase their reactive surface area. As the company’s commercial vice president, Luke Rondel, explained, “We crush [the slag] down so you get smaller particle sizes. We then spread that out in a field of material, and we till that material with a tractor and plow, which is just turning over new surfaces.”
Each pathway has its advantages — while Arca’s magnesium-rich mine tailings are the most abundant feedstock, Rondel told me that the calcium-based reactions in slag happen significantly faster. For its part, Frontier hopes to test and evaluate a range of approaches at its new Surficial Mineralization Hub in Quebec, which it announced at the end of April. Located at a former asbestos mine, the hub will give participating startups access to “10,000 tons of serpentinite tailings and space for pilot scale testing,” Bebbington Valori told me, as well as local labs with specialized equipment.
This should eliminate some of the hurdles facing the nascent sector, chief among them being access to the right kinds of reactive rocks. Small startups “really need to either partner with large academic labs or with large mining companies to get access to that feedstock,” Bebbington Valori told me — a difficult and expensive proposition for a company that’s just getting off the ground.
While Frontier has yet to announce the cohort of participating startups, both Arca and Karbonetiq told me they hope to test their technology there, with the latter planning what would be one of its first mine tailings pilots through the program. Ultimately the goal is to generate the proof points needed to give both the startups and Frontier a clearer roadmap for which approaches can realistically scale — and what kind of support they’ll need to get there.
It certainly won’t be a straightforward process — bringing new technology into old-school industries never is — and the economics will only start to pencil if their operations reach meaningful scale. In theory, mining companies could benefit from hosting surficial mineralization projects, whether through site access fees, outsourcing elements of waste management, or even critical minerals recovery. Miners could even develop and scale the technology themselves, if they so desire. But the sector has historically been reluctant to adopt new tech. “The classic quote is, in mining you always want to be No. 2, you don’t want to be the first one,” Dipple told me. “You don’t want to put up a $2 billion plant that doesn’t work.”
So like nearly everything in the carbon removal space, early execution is falling to the startups that aren’t afraid of a little risk. “They’re watching for sure,” Dipple said of the mining industry at large. “But they want to be No. 2. We’re going to have to be No. 1.”