It’s the talk of the crypto-sphere: How fast will Chinese Bitcoin miners—all and sundry recently ejected from what was formerly the world’s largest nation for minting coins—find new homes, and where will they land?
Overnight, the Chinese exit banished half the competition for the producers who never stopped cranking in the rest of the world, hugely expanding their profits. The quicker the displaced Chinese can relocate, the more of that bounty they’ll reap for themselves, bounty greatly enhanced by Bitcoin’s one-third leap in price (to $46,000) since Beijing’s late-June crackdown.
The Chinese earthquake also spotlights the environmental issue that’s crucial to Bitcoin’s future: If most of the displaced Chinese miners choose locales where they’ll run on fossil fuels, the industry’s already gigantic carbon footprint, the size of Greece’s before the lockdown, could grow far vaster. In part, that’s because the hydroelectric plants in Sichuan and Yunnan provinces hosted by far the world’s largest share of green production before the exodus. Now, miners that once drew their power from China’s raging rivers are moving to regions offering cheap natural gas, fuel oil, and even coal.
BIT Mining, one of China’s largest producers, along with numerous others from the diaspora, are powering up on coal in Kazakhstan. At an old Alcoa aluminum smelting plant in Rockdale, Texas, giant expatriate Bitmain installed banks of computers extending the length of three football fields that will run on natural gas––one of Bitmain’s executives proudly posed by the towering racks in a ten-gallon hat. A Nevada energy company called Black Rock Petroleum recently announced a deal with a group of Chinese producers that, if it happens, would move 1 million mining machines, and a huge share of the uprooted industry, to Canada’s oil-rich Alberta province.
The Chinese migration is obscuring another trend that’s at least as important: The mother lode in Bitcoin, boosted by the sudden shrinkage in mining capacity and jump in prices, is luring U.S. and other non-Chinese entrepreneurs, mostly backed by venture capital and independent power suppliers. Where these startups go will play a lead role in establishing whether Bitcoin leans green, or spews ever greater quantities of greenhouse gases. And no startup underscores that urgent emissions question more than Stronghold Digital Mining, a new U.S. miner that expects to build a very profitable business around one of the world’s dirtiest energy sources.
Fat Bitcoin profits lure startups
Private equity firm Atlas Holdings is farming Bitcoin at a onetime coal plant retrofitted for natural gas on the shores of the scenic Finger Lakes; the operation has raised fierce opposition from residents who claim that hot water discharged from the facility is warming Lake Seneca and endangering aquatic life. One local complained that the quest for Bitcoin has turned the glacial waters into a “hot tub.” In Kentucky, the largest coal-producing state in the nation, the legislature just passed tax incentives that will benefit such Bitcoin miners as Blockware, which is planning a new plant in Paducah that will run on a blend of hydro, wind, and of course coal. Wyoming has joined Kentucky in offering tax benefits. Crusoe Energy Systems is an operator in the Powder River Basin that takes excess natural gas that’s now flared and uses it for mining. Crusoe is now seeking $100 million to $125 million in loans to rechannel the energy that’s now burned and wasted, creating a flaming eyesore for nearby communities.
The flood of new ventures encompasses several that are greenhouse-gas-free. In Texas, Manhattan Solar Partners has entered a joint venture with crypto mining outfits BIT5IVE and GMINE for erecting new renewable power plants. Even nuclear power producers are entering the mix. Nuclear producer Energy Harbor is partnering with crypto miner Standard Power on a state-of-the-art facility in Ohio, and Talen Energy signed a deal with producer TeraWulf for a plant next to Talen’s twin reactors in Berwick, Pa., that could cost $400 million.
It’s hard to predict the environmental impact of the Chinese flight and the rest-of-the-world gold rush. The predominance of private players and lack of industrywide data obscures how the evolving balance between nuclear, fossil fuels, and green sources is playing out. The test: At these prices, miners can make incredibly high profits mining Bitcoin. The lower the cost of their energy feedstock, the higher those profits. Right now, they’re often finding the cheapest, most reliable sources in regions that offer plentiful natural gas—and, as we’ll see, they’re also turning to more polluting coal and waste coal. In important cases, it’s state subsidies that are making those non-green supplies so irresistibly cheap. The danger is that we’re in an arms race where the industry expands well beyond its size when China was online, generating far more electricity than ever to exploit what’s right now possibly the most lucrative major business on the planet.
“At a price [per Bitcoin] of $47,000, the power utilized by miners would expand by 50% over the level before the China crackdown,” says Alex de Vries, the Dutch economist whose website Digiconomist tracks Bitcoin’s carbon footprint. A sizable business getting much larger and at the same time more dependent on fossil fuels would hobble the drive toward the goal of net-zero carbon emissions by 2050, embraced by the U.S., European Union, and dozens of other nations.
Stronghold taps “waste coal”
No newcomer better exemplifies the challenge than Stronghold Digital Mining. The Western Pennsylvania startup plans to become one of the world’s biggest Bitcoin producers and promises to generate the richest of margins––by deploying the one of the dirtiest forms of energy now generating electricity in America. That energy source is waste coal, dumped by mines as long as a century or more ago, which produces half-again as much CO2 as regular coal, and three to four times the carbon tonnage from natural gas.
The Commonwealth of Pennsylvania, however, strongly supports the industry. The state even rates its “coal refuse” plants in the same environmentally friendly category as hydropower. The reason: The facilities supposedly offset the negatives from high emissions with the benefits of removing the black carbon mountains that feed their boilers. State government in Harrisburg mandates credits that greatly reduce Stronghold’s fuel expense. In fact, Stronghold claims that it can produce “at what we believe to be some of the lowest costs in the industry.” It’s unclear how that’s possible, since coal refuse can’t generate electricity for heating homes and powering plants at prices that come close to competing with the Keystone State’s plentiful natural gas.
Still, even if costs turn out higher than Stronghold’s predictions, its model for generating Bitcoin with waste coal looks so fantastically profitable that it could well attract more miners, further reviving what was a dying field. According to Fortune’s analysis of figures in Stronghold’s July 27 filing for an IPO, the venture when fully up and running expects to clock over half-a-billion dollars in annual revenue, and operating profits in the 90% range, assuming Bitcoin remains at or near its current price of around $46,000. The excellent record of Stronghold’s founders and the credible projections in the S-1 for its offering that incorporate the big benefit of the subsidies make a convincing case that Stronghold will hit, at least in the short run, margins that dwarf those at such top earners as Apple, Microsoft, AMD, or LVMH.
Stronghold’s cofounders are Bill Spence, a mining engineer and waste-coal veteran who owned and ran its flagship plant, and CEO Greg Beard, a Wall Street star drawn by the venture’s towering prospects to the wilds of northwestern Pennsylvania. According to his bio on Stronghold’s home page, Beard served from 2010 to 2020 as senior partner and global head of natural resources at investment colossus Apollo Global Management.
In late June, Stronghold announced that it had raised $105 million in equity via two funding rounds. A month later, it filed for a $100 million IPO. Stronghold is only a few months old, formed on March 19 of this year, according to the S-1. It currently owns one plant, the 650-acre Scrubgrass facility in tiny Kennerdell, Pa., about an hour’s drive from Youngstown, Ohio. That’s the facility where Spence first mined Bitcoin. The company is in contract to purchase a second facility, called Panther Creek, in the state’s eastern tier just north of Allentown. Stronghold has also reached a deal to acquire an undisclosed plant—reportedly the Northampton Generation Co., located about 25 miles from Panther Creek. Altogether, the three plants pack 277 megawatts in generating power, about the punch of a small gas-fired plant.
Stronghold didn’t respond to two emails that I sent requesting interviews for the story. Nor did the company respond to a third that presented my projections of revenues and profits based on information in the S-1, and asked for comments or clarifications.
Bitcoin revives a dying business
Bitcoin mining is handing a lifeline to an industry spiraling downward. From the mid-19th century to the 1970s, mines in Pennsylvania routinely separated the low-quality coal—scraps mixed with rocks, often with high sulfur or ash content—from the good stuff. Children called “breaker boys” seated at conveyor belts or chutes would pick out the slate and rock by hand, no gloves permitted—brutal work that often left their fingers bleeding. The mines dumped the waste coal in the fields and villages near their plants to build mountains as high as 100 feet, many standing to this day. Or they would empty their truckloads to backfill entire valleys. Even now, over 800 piles dot the former coal regions, covering a surface area 250 times the size of the baseball field at PNC Park, home to the Pittsburgh Pirates.
Rain mixed with the sulfur forms a toxic runoff called acid mine drainage, or AMD, that leaches into rivers and aquifers that furnish drinking water for communities. Campfires and lightning sometimes ignite the coal hills and plains. Many will flame for weeks, cloaking nearby homes in black smoke. The piles didn’t just scar the countryside. The eyesores blighted rural villages as well. In Swoyersville, near Scranton, a towering bank of waste coal looms over homes that seem to stand in its foothills.
It wasn’t until 1978 that federal legislation banned coal mines from dumping their refuse aboveground, instead requiring that they burn or bury it in landfills. The federal Public Utility Regulatory Policy Act (PURPA) took aim by requiring local utilities to purchase power from newly built plants specifically equipped with technology for burning waste coal. Under PURPA, the utilities awarded long-term power agreements that guaranteed new facilities above-market prices that covered their construction and operating costs, and provided an attractive profit to boot.
The coal refuse industry was born. In the years that followed through the early 1990s, 11 plants sprouted in Pennsylvania, mostly built by large independent power companies such as General Electric and Foster Wheeler drawn by the price-guaranteed contracts typically running 20 to 30 years. The concept was basic: The facilities either made their own deals with the towns or owners of the land where the piles were located, or turned to “primary fuel suppliers,” independents who secured rights to the sites, scooped the mountains with excavators, and trucked the coal to the newly built plants. Those outfits would burn the coal in specialized boilers powerful enough to pulverize rock, generating electricity they sold to the regional utilities.
For well over a decade, the purchase power agreements worked well. Burning waste coal provided the revenue for cleaning up waste coal. “The goal wasn’t primarily to furnish electricity,” says Jaret Gibbons, director of the coal refuse industry group Appalachian Region Independent Power Producers Association (ARIPPA). “The purpose was environmental, to generate the cash to clean up these piles polluting rivers and streams.” In the past 30 years, the industry has removed 220 million tons of coal refuse, eliminating around half the mounds strewn across Pennsylvania. The burning process leaves a powdery byproduct that producers spread over the work sites as fertilizer; the ash loosens the hard texture of the soil, allowing plants to sprout quickly. In place of scarred landscapes, the industry left bike trails, open spaces, and parks. Recycling waste coal for profit rescued coal burial sites that are now home to baseball fields and warehouses.
Starting around 2013, two forces converged to rock the industry. First, the purchase power agreements began expiring. Since Pennsylvania had partially deregulated its power industry several years earlier, the utilities were no longer obligated to buy power from coal refuse producers at often-elevated prices. That shift forced the industry to sell electricity on the open market to utilities on the gigantic, highly competitive PJM grid that spans 13 states from North Carolina to New Jersey to Illinois. Second, the fracking boom, and especially the vast supplies of natural gas flowing from Pennsylvania’s Marcellus shale fields, pushed the price of that superabundant source so low that waste coal couldn’t compete. “Natural gas prices dropped sharply over the past decade,” says Gibbons. “Waste coal was competitive when the utilities were paying well over 4 cents a kilowatt hour or more at wholesale for waste-coal production, and natural gas was about the same price. But natural gas prices fell to 2 cents. That was unheard-of a decade ago.”
Since 2018, five Pennsylvania plants have closed down (a sixth shuttered years before). The big owners mainly bailed. Many sold at extremely low prices to the primary fuel suppliers that dig and deliver the coal. For example, the Scrubgrass feedstock contractor Bill Spence reportedly purchased the plant in 2017, turning it into a vertically integrated operation. Even so, diving energy prices forced the industry to severely curb its output of electricity. Last year, the business ran at just one-seventh of its total capacity, limiting operations mainly to periods of peak demand in the summer and winter. In 2020 according to ARIPPA’s numbers, Scrubgrass went from producing 700,000 megawatt hours at the peak to 5,900 in 2020.
The industry’s decline throttled the cleanup campaign. Last year, the producers removed just 10 million tons, around half the total in the flush years. At the old peak rate, says Gibbons, it would have taken 20 more years to erase the remaining piles. At the current pace, the timeline would extend to 2061.
State credits should boost Stronghold’s profits
Pennsylvania supports the industry with two forms of aid, one of which will be a major factor in lowering Stronghold’s mining costs. The first is the Coal Refuse and Reclamation Tax Credit. Enacted in 2016, it today provides a total of $20 million a year in payments to producers. The larger the share of the tonnage that an individual producer collects and burns each year, the more of the $20 million that operator pockets. The awards are capped at a maximum of $4.4 million for any one plant. Hence, though it could provide a nice extra boost, the plan won’t be a major contributor to Stronghold’s Bitcoin mining profits.
On the other hand, the second program promises a bonanza due to a recent change in its design. In force since 2004, the state’s Alternative Energy Portfolio Standards (AEPS) divides sources into two “tiers” that qualify for differing subsidies. Tier 1 encompasses wind and solar. Under Tier 2 are hydropower, municipal solid waste colloquially known as garbage, and coal refuse. It’s Pennsylvania’s version of the Renewables Portfolio Standards or RPS plans across the country; the Keystone State uses “alternative” instead of “renewables” to bring waste coal and other non-wind or solar sources under the Tier 2 umbrella. That the latter is deemed as environmentally beneficial as hydro is abhorred by green energy advocates. Recently, the AEPS has provided extra cash that’s helped keep reclamation efforts chugging, albeit at slower speed.
Under the program, the Tier 2 players receive credits based on their output. Every megawatt hour they produce creates a corresponding AEPS credit. The more megawatts the owner of a coal refuse plant generates, the more credits it amasses and sells to the utilities. The utilities are obliged to purchase credits equivalent to 10% of the retail electricity that they sell in their territory. (The number for Tier 1 is 8%.) The more electricity a utility provides to light homes and run data centers in the Scranton area, for example, the more credits it must buy to fill its quota.
Until recently, the utilities were free to purchase credits anywhere in a gigantic, multistate market. The credits are traded in an open marketplace, with prices set by supply and demand. Prior to last year, they fetched, on average, about 25 cents for each mWh the plant produces annually. The reason they were cheap: Utilities could buy credits from any Tier 2 producer in the entire 13-state region covered by the PJM grid. They could shop in a huge, highly competitive arena that kept prices low. A utility in Scranton could buy credits from a solid-waste outfit in North Carolina, say.
In November of 2020, Harrisburg changed the rules to buoy the flagging fortunes of waste coal. The legislature now mandates that Pennsylvania utilities purchase Tier 2 credits exclusively from the hydro, solid waste, and coal refuse enterprises operating within its borders. Since Pennsylvania is a much smaller market than that covered by all of PJM, prices are jumping. In early 2020, before the restrictions took hold, credits were selling for around 25 cents each. Late last year, they started spiking in anticipation of the shift, jumping to $1 to $2. Incredibly, the credits are now trading in the $20 range, an 80-fold leap from early 2020. “Closing the borders is driving up prices and subsidies,” says Rob Altenburg of environmental group PennFuture.
The rich prices created by the new regime could prove a windfall for Stronghold. At the old peak production, its three plants were generating almost 2.2 million mWh a year. At the current price of around $20 per mWh, that’s $44 million in extra revenue. As we’ll see, the newly enhanced AEPS are a big help in getting to what Stronghold predicts as Bitcoin mining costs that will rank among the industry’s best.
Ambitious plans in an IPO filing
In its S-1 IPO filing, Stronghold unveils what must rank among the industry’s most ambitious blueprints for expansion. In this year’s first quarter, the Scrubgrass plant generated only about $1.1 million from mining Bitcoin and “hosting” or running power for other producers whose machines are parked at the plant. That’s even less than the combined revenues that Scrubgrass and soon-to-be-acquired Panther Creek pocketed by selling electricity to the grid, a number that’s in turn a fraction of past years.
But cofounders Spence and Beard plan to pack the three plants with so many state-of-the-art “miners” that they’ll soon be back to the peak production days of guaranteed pricing. The S-1 provides rich detail on their mining costs and the projected size of their operation. It states that as of today, Stronghold owns just 1,800 “miners,” but has ordered another 27,300, and pledges to use part of the proceeds from the $100 million IPO to buy 27,900 more. When that firepower’s in place by December 2022, says the S-1, Stronghold will be generating 5,300 petahashes per second in pursuit of Bitcoin. (A petahash is one quadrillion hashes.)
The way the Bitcoin algorithm works, a miner gains the share of newly issued coins that’s equal to its proportion of the Bitcoin network’s total computing power. The hashes are random codes, one of which matches the hash that releases the award of 6.25 Bitcoin every 10 minutes. A producer that generates 1% of the hashes typically gets 1% of the Bitcoin. In the S-1, Stronghold “assumes” a size of 150,000 petahashes per second for the Bitcoin ecosystem. Due to the flight from China, it’s actually running well below that now at around 119,000. But using Stronghold’s figure, at 5,300 petahashes, it would by the close of next year control 3.5% of what might be called Bitcoin production capacity.
How does that share translate into revenues? Obviously, that will depend on the wildly volatile cryptocurrency’s future price. At around $46,000 on Aug. 12, the total value of the 328,500 Bitcoin issued each year, plus transactions fees paid to the miners, is $15.5 billion. Hence, Stronghold would be collecting roughly $540 million a year in revenue (3.5% of $15.5 billion). The startup would be winning 33 coins a day worth around $1.5 million.
Naturally, its profits will depend on the cost of mining each one of those 33 coins. Here’s where the numbers get stratospheric. In the S-1, Stronghold discloses that it’s aiming to “dedicate” 204,000 kilowatt hours of electricity to minting Bitcoin. For the Scrubgrass facility, Stronghold puts production costs per kWH at just 1.8 cents. And that’s an all-in number. It represents “all of our expected costs of generating power (including the cost of procuring fuel, operations and maintenance expenses, and plant general and administrative expenses).” All told, Stronghold says that at the Scrubgrass plant it has “a cost to mine of less than $3,000 per Bitcoin.”
I’m making the reasonable assumption that production costs at the other two plants will resemble those at Scrubgrass. Stronghold doesn’t say otherwise. At 204,000 kilowatts per hour, Stronghold would be spending roughly $88,000 a day or $32 million a year on production costs. It also reveals that sans state subsidies, its expense would be more than twice the 1.8 cents at 3.8 cents, for a total of $68 million. So it appears that the jump in the value of credits it sells to utilities, due to the recent change limiting the market to Pennsylvania, is a huge factor in lowering its mining costs. The S-1 displays a chart showing where Stronghold’s 1.8 cents stands versus the number for six “comparable companies.” Their costs per kWh range from a low of 2.2 cents, to a high of 4.0 cents. The study appears to bolster Stronghold’s case that its production costs are best, or nearly the best, in the entire domain of Bitcoin.
By my estimates, the expense that the S-1 calls “less than $3,000” per Bitcoin is actually around $2,700. If the scenario plays out as expressed in the S-1, and Bitcoin’s price stays at $46,000 (admittedly a big “if”), Stronghold by late next year would be making gross mining profits of $540 million in sales less $32 million in expenses, or $508 million. That’s a margin of 94%.
Of course, Stronghold will be shouldering a lot more costs, chiefly for its investment in the plants (whose expense is counted in depreciation) and corporate overhead. For its IPO filing, Stronghold commissioned an independent study from PA Consulting that calculates all expenses for Scrubgrass, apparently a proxy for its future full operation, excluding SG&A. The measure’s called “leveled cost of energy,” or LCOE. It mainly differs from the 1.8 cents in production costs by adding capital expenditures, notably “initial capital investment,” encompassing what Stronghold paid for the Scrubgrass plant. Amazingly, tacking on that item raises the expense per kWh to just 2.1 cents. That would amount to an additional $6 million a year when the entire operation is up and running.
As for overhead, the S-1 contains a pro forma showing the combined SG&A expense for Scrubgrass and Panther Creek for 2020 and 2019. In the latter, a much more active year, the figure was $5.6 million. Northampton, the third facility, is about 30% larger than the other two. Adding Northampton, a high-end number would seem to be $10 million a year.
Let’s add it up. A good estimate of total operating expenses is the $32 million for production, $6 million for capital investment, and the $10 million for overhead. That comes to $48 million. At annual revenues of $540 million, Stronghold would book operating profits, or earnings before taxes and interest, of $492 million, or a margin of 91%. In comparison, in their last fiscal years Pfizer had an operating margin of 18%; Apple and Alphabet both achieved 24%; and Microsoft registered 37%.
Stronghold’s projections raise a big question. If Scrubgrass can generate electricity at a cost of 1.8 cents to produce Bitcoin, it seems that it could have been producing power for utilities at that cost all along. If so, the operation could have competed with natural gas and kept running at full capacity. Instead, it mostly shut down. Altenburg of environmental group PennFuture believes that Scrubgrass’s costs are much higher that the 3.8 cents before subsidies. “Getting to 1.8 cents, to put it charitably, is highly optimistic without massive additional subsidies,” he says. But even at a much higher number, say 7 cents, Stronghold would still be producing at 5 cents after the state credits. Its production costs would go from a projected $31 million to $86 million. Even so, Stronghold would be garnering 80% operating margins.
“Everywhere else on the grid gas is cheaper than coal,” says Altenburg. “That applies even with the subsidies. So it raises the question of why Stronghold is mining with waste coal instead of natural gas.”
What will mining with coal refuse mean for the environment?
Stronghold and the champions of the clean-to-burn solution claim that Bitcoin mining’s arrival is a prime example of how the leading crypto can improve the environment. Gibbons, chief of the industry’s association ARIPPA, likes the Stronghold venture because it will help revive a mission now in retreat. “It’s another opportunity to inject private funding into the waste-coal cleanup process,” he says. “If plants aren’t running, they aren’t doing their job of cleaning up the polluting piles. If using Bitcoin as a vehicle to help fund those operations allows them to operate more, it’s a win for the commonwealth and the environment.”
Gibbons points out that the plants are effective in eliminating such pollutants as mercury, sulfur dioxide, and the nitrous oxide that contributes to smog. Waste coal meets state and federal permitting requirements for all those emissions. What about the industry’s heavy CO2 emissions? The waste-coal heaps also send pollutants into the air we breathe that are often more damaging than airborne carbon. “Simply looking at the plants overlooks the totality of their operations,” notes Gibbons. When the piles ignite into flame or smolder beneath the surface, they unleash methane that’s four times as potent as CO2, as well as carcinogens such as mercury. “The plants collect the waste coal that throw off those dangerous emissions, and captures them when the same coal is burned in their boilers,” he says.
Altenburg doesn’t buy that trading one form of pollution for another, especially with Bitcoin poised to put one of them on the fast track, will advance the Keystone State’s or the nation’s energy goals. Quite the contrary. “You’re trading pollution in land and water for pollution in the air. Less pollution is always better,” he says. “You hear from the industry that emissions are declining because natural gas is replacing coal. Burning more waste coal goes in the other direction.” Hitting state and national carbon targets is hard enough, he adds, without burning more fossil fuels when the target is using less. “If all the waste-coal plants go to Bitcoin, it would make reaching reasonable carbon targets in Pennsylvania impossible. Waste coal pollutes far more than the regular coal plants that are disappearing. So we’re backpedaling in encouraging the burning of more coal refuse.”
Altenburg favors a state campaign to remove the coal piles or in some cases shield the ground with beach grasses and other vegetation. “The industry argues that the value you get from cleaning land and water is worth the damage from polluting the air,” he declares. Sure, the water quality of the town’s streams improve, Altenburg acknowledges. The dark side: “But in the next town a number of kids will get asthma attacks and be hospitalized for respiratory issues. The industry never talks about the other side of the equation.”
Whatever the arguments from both camps, one point is obvious. Waste coal is simply the one of the dirtiest of fossil fuels and sends enormous volumes of carbon into the atmosphere. Recent studies calculate that waste coal emits almost 1.1 tons of carbon per megawatt hour generated, versus 0.35 tons for natural gas and 0.8 tons (or 30% less) for regular coal. A reasonable estimate is that the three Stronghold plants would throw off well over 3 million tons of carbon a year, three times the volumes for a gas-fired plant of similar size. On the other hand, Stronghold’s big expansion will doubtless hasten the disappearance of tokens of a bygone Industrial Age still ravaging the scenic hills and historic hamlets of Pennsylvania.
The Stronghold case is a great illustration of how the combination of a Bitcoin price that’s multiplied fivefold in the past year, and the sudden shutdown of half the mining industry, has boosted profitability to such high levels that we’ll see immense expansion of mining in the months or years ahead. As the Chinese relocate and competition rises, those gigantic profits will shrink. Still, rapid growth in production will lead to a rise in Bitcoin’s carbon footprint. The Stronghold example shows how the lure of a cash bonanza from the booming Bitcoin business, and government policies favoring the old fossil fuel plants that the explosive sector is helping to revive, can coalesce to undermine America’s vision of a green future.
Subscribe to Fortune Daily to get essential business stories straight to your inbox each morning.
