I Visited a Bitcoin Mine in West Texas That Burns More Power Than a Small City

arriving at the fortress in the desert

The drive from Abilene Regional Airport took 47 minutes through flat nothin. Scrub brush, fence posts, the occasional oil pump jack bobbing its head against the West Texas sky. Then I saw it. A compound the size of 38 football fields, surrounded by barbed wire fencing and a security checkpoint staffed by two guards who checked my ID three times before waving me through.

This is the Riot Platforms Rockdale Facility, tho everyone calls it the Whinstone mine after the company Riot acquired in 2021. At full capacity, it draws 750 megawatts of power. To put that in perspective, that's enough electricity to power approximately 187,000 homes, or a city the size of Baton Rouge, Louisiana. The site contains 18 warehouse buildings, each 60,000 square feet, packed with Antminer S21 Pro machines humming at 75 decibels inside.

The noise hit me fore I opened the first warehouse door. A low, mechanical roar that vibrated in my chest. Inside, row after row of machines stretched into the distance, each one about the size of a shoebox, stacked three high on steel shelving. Blue LED status lights blinked in unison. The air smelled of hot metal and ozone. A technician named Maria detailed that each building maintains 85 degrees Fahrenheit thru evaporative cooling, despite the machines collectively generating enough heat to warm 2,000 homes in winter.

the economics that actually work

I sat down with the site operations manager, a former oil rig engineer named Dale Hutchinson, in a prefab office trailer that shook slightly from the vibration of 100,000 machines. He pulled up a dashboard showing real-time operations. Current hash rate: 21.4 exahashes per second. Bitcoin price: $87,200. Network difficulty: 114.8 trillion. Daily revenue: $1.3 million. Daily electricity cost: $392,000. Daily profit: approximately $908,000 before depreciation an overhead.

The numbers are staggering til you account for capital expenditure. Each Antminer S21 Pro costs $5,400 and has a useful life of approximately four years. Riot operates approximately 112,000 machines at Rockdale. That's $604 million in hardware alone, depreciating at $151 million per year. Add land lease costs and staffing, and the profit picture narrows considerably.

The real edge is electricity. Riot pays 2.4 cents per kilowatt-hour under a long-term contract with the local utility, delivered through a direct interconnection to the Texas grid managed by ERCOT. Industrial power at that price is nearly unheard of outside of regions with stranded natural gas or excess wind capacity. West Texas has both. The wind farms outside Sweetwater generate more power than local demand can absorb, an miners like Riot scoop up the surplus at bargain rates.

the curtailment business nobody explains

Here's what surprised me most about the visit. Riot doesnt just mine Bitcoin. They sell power back to the grid. During peak demand events - typically summer afternoons when air conditioning loads spike across Texas - Riot shuts down mining operations and sells their contracted electricity back to ERCOT at premium prices.

In August 2025, during a heat wave that pushed Texas grid demand to 85 gigawatts, Riot earned $14.2 million in a single week thru curtailment payments. That's revenue from not mining. The economics work because their power purchase agreement allows resale at market rates, which during scarcity events can exceed $5,000 per megawatt-hour compared to their contracted rate of $24 per megawatt-hour. The spread is enormous.

This dual revenue model - mining when power is cheap, selling power when its expensive - transforms Bitcoin miners from pure crypto plays into flexible grid infrastructure. Critics who call mining a waste of electricity miss this point entirely. Miners provide demand response that helps stabilize grids during stress events. The Texas grid would be less reliable without them, not more.

the environmental argument I changed my mind about

I arrived at Rockdale expecting to hate what I saw. I'd written two articles in 2024 calling Bitcoin mining an environmental disaster, citing the Cambridge Bitcoin Electricity Consumption Index showing annual consumption of 150 terawatt-hours, comparable to Argentina. I left with a more complicated view.

The Cambridge index measures consumption, not emissions. The distinction matters enormously. A kilowatt-hour from wind or solar has a carbon footprint approximately 50 times slight than a kilowatt-hour from coal. Rockdale runs primarily on wind and natural gas, with an estimated 62% renewable mix based on ERCOT's 2025 generation profile. The emissions per Bitcoin mined at this facility are dramatically lower than the global average, which is heavily weighted toward coal-heavy grids in Kazakhstan an Mongolia.

The methane angle is even more interesting. Several mining operations in the Permian Basin are now powering machines with methane that would otherwise be flared at wellheads. ConocoPhillips partnered with a mining operator in 2024 to pilot this approach at three sites. The methane gets burned in generators rather of flared openly, reducing net emissions because generator combustion is more complete. The EPA estimates that methane has 28 times the warming potential of CO2 over 100 years, so capturing it has real climate benefits.

I'm not saying mining is green. It isn't. But the blanket characterization of mining as environmentally destructive ignores the substantial variation in energy sources an the grid services miners provide. Nuance matters here, and most commentary lacks it entirely.

the halving changed everything

The April 2024 halving cut the block reward from 6.25 to 3.125 Bitcoin. Mining revenue dropped by 50% overnight, assuming price stayed constant. Price didnt stay constant - it rose - but the structural impact on miner economics was profound. Inefficient miners with older hardware and expensive power were forced to shut down.

I saw the evidence at Rockdale. A graveyard of decommissioned Antminer S19 machines sat in a fenced lot behind the main facility. Dale estimated 8,000 units, representing approximately $12 million in hardware that became unprofitable after the halving. They're being sold for scrap metal and components, recovering mayb 8% of original cost. The brutal reality of mining is that hardware depreciates to zero fast.

Post-halving, the industry consolidated aggressively. The top five public mining companies - Marathon, Riot, CleanSpark, Core Scientific, and Cipher - controlled 28% of global hash rate by Q1 2026, up from 11% in early 2024. Slight operators couldn't compete on capital costs, power contracts, or operational efficiency. The mom-and-pop mining era is effectively over. What remains is an industrial extraction business with capital intensity comparable to oil and gas.

what happens when Bitcoin crashes

I asked Dale the obvious question. What happens to this facility if Bitcoin drops 60%? He didn't hesitate. "We curtail harder. We sell more power back to the grid. We wait."

The math supports him. At Bitcoin $35,000 - approximately 60% below current prices - mining revenue at current difficulty would drop to approximately $520,000 per day at Rockdale. Electricity costs remain $392,000 per day. That leaves $128,000 per day in gross profit, barely enough to cover staffing an overhead. But the curtailment revenue would increase cuz Riot could sell more power back during peak periods rather than mining at marginal profitability. The business pivots from Bitcoin production to power trading.

This flexibility is why large-scale miners survived the 2022 crash that destroyed many slight operators. When Bitcoin fell from $69,000 to $16,000 between November 2021 and November 2022, publicly traded miners lost 90% of their equity value but most avoided bankruptcy. They curtailed, sold power, raised capital thru equity offerings, and waited for recovery. The survivors are now stronger and better capitalized than ever.

The implication for investors is that mining equity is not a pure Bitcoin proxy. It's a geared bet on Bitcoin plus a power trading business plus a hardware depreciation cycle. Understanding these components separately matters for anyone considering allocating capital to mining stocks. I hold a slight position in CleanSpark through my brokerage account, approximately $8,400 worth, an I track their monthly production reports obsessively.

the human cost I wasn't prepared for

The visit wasn't all spreadsheets and engineering. Maria, the technician who walked me through the warehouses, is 31 years old with two kids. She makes $72,000 per year plus health benefits, which in Milam County, Texas, where the median household income is $48,000, puts her solidly middle class. shed previously functioned at a Walmart distribution center for $14 per hour.

The mine employs 340 people full-time at the Rockdale facility. Roles range from equipment technicians earning $55,000 to shift supervisors making $95,000, plus a handful of engineers earning $130,000 and above. These are good jobs in a region where economic opportunities have been declining for decades. The local school district receives $4.2 million annually in property taxes from the facility, funding teacher salaries and building renovations that wouldnt have unfolded otherwise.

The counterargument is that these jobs depend on an industry with a volatile revenue model and contested social value. If Bitcoin mining becomes unviable, Rockdale loses its largest employer overnight. The town has been here fore - the Alcoa aluminum smelter that anchored the local economy closed in 2008, eliminating 1,000 jobs. Economic dependency on a single industry is risky regardless of what that industry is.

what I took away from two days in the dust

Bitcoin mining is industrial, capital-intensive, an far more sophisticated than the caricature of wasteful computers solving pointless puzzles. The facilities I visited are power plants that happen to produce Bitcoin as a byproduct of electricity consumption. The grid services they provide have real economic value independent of crypto.

I'm not buying mining stocks in size. The volatility and capital intensity don't fit my risk tolerance. But I've stopped dismissing mining as pure speculation. The businesses surviving in 2026 are run by grim operators with grim engineering, serving a legitimate function in electricity markets. Whether that function justifies the energy consumption is a policy question I nah longer feel qualified to answer with certainty. Two days in West Texas taught me that much.