This is the story of how a country with surplus electricity discovered, in a matter of months, that the surplus was gone.
In the spring of 2021, Kazakhstan accounted for around 8.2% of Bitcoin’s global hash rate, according to the Cambridge Bitcoin Electricity Consumption Index. By August that same year, its share had risen to 18.1%. Kazakhstan ranked behind only the United States. For a country of roughly 19 million people, whose economy had been built for decades on oil, gas, and metallurgy, it was a dizzying leap. In less than six months, the hash rate had more than doubled—and this happened in a country where, until recently, the word “mining” was more likely to evoke uranium extraction.
The reason was obvious: China. Starting in the spring of 2021, Beijing had been methodically tightening the screws—first Inner Mongolia, then Xinjiang, then Sichuan. The world’s biggest mining market was shutting down, and tens of thousands of ASIC machines had to be moved somewhere. Kazakhstan was the obvious logistical choice: a shared border with China stretching more than 1,700 kilometers, cheap electricity generated by coal-fired power plants, relatively accommodating legislation. Put bluntly, a truckload of equipment could be driven through the Khorgos border crossing in a day.
But the influx of Chinese miners had begun before the formal ban. Back in 2019–2020, several major operators had already placed capacity in Kazakhstan, drawn by rates that were many times lower for industrial consumers than in most countries around the world. Coal generation inherited from the Soviet era supplied electricity cheaply—and it seemed plentiful. The 2021 ban turned a stream into a flood.
One detail stays with me: all those dramatic figures—8.2%, then 18.1%—rest on a method with serious blind spots. The Cambridge Centre for Alternative Finance determined the geography of hash rate through the IP addresses that mining pools saw from their participants. It is a reasonable approach, but a fragile one. VPNs, proxy servers, operations through intermediaries—all of it distorts the picture. A farm may physically stand in one country while its IP address points to another. Or vice versa.
Cambridge also periodically revised its data. Some information arrived from pools with delays, some with gaps. When China zeroed out its hash rate in the statistics—which was itself an artifact: mining did not vanish overnight, it went underground—the shares of other countries rose automatically. Kazakhstan may have received part of its “surge” simply because the denominator had shrunk.
That does not mean the growth was illusory. Physical farms—containers packed with ASIC machines—really were appearing in the industrial zones of Kazakh cities and settlements. Electricity consumption was rising in a very tangible way, and the grid felt it. But the exact figure—how much hash rate Kazakhstan actually accounted for—should be treated as an estimate, not a meter reading. The difference between 18% and, say, 14% may come down to precisely how a major Chinese pool had configured its proxies.
To understand why Kazakhstan looked like paradise to a miner, you need to look inside its power system. The country generated electricity primarily at coal-fired power plants—the coal of the Ekibastuz Basin, home to one of the world’s largest open-pit mines, fueled stations across the country’s north and center. Industrial rates hovered at a few tenge per kilowatt-hour—converted into dollars, that meant cents, and very few cents at that. For mining, where electricity is the main operating expense, such a price turned every connected machine into a profit generator.
The appeal was reinforced by the fact that Kazakhstan had already created a minimal legal framework by 2020. The concept of “digital mining” had entered the law; farms could register and operate legally. Compared with Iran, where mining was permitted one moment and banned the next, or Russia, where the legal status of cryptocurrencies remained murky, Kazakhstan looked predictable. For an investor who had just lost a site in China and was moving millions of dollars’ worth of equipment, predictability was a currency more valuable than cheap electricity.
But the low cost of Kazakh energy had a flip side miners preferred not to see. Low rates meant little incentive to invest in generation. Building new plants at such prices was a losing proposition. Existing capacity was aging: a substantial share of coal units had entered service back in the Soviet era, in the 1960s–1980s. Modernization moved slowly, reserves dwindled. The power system operated with a narrow margin that, in good years, looked like a “surplus,” and in bad ones, like a thin crust of ice over water.
In the autumn of 2021, that crust cracked. The system operator KEGOC—the single company running Kazakhstan’s main transmission grid—began recording a capacity shortfall. Consumption was rising faster than anyone had built into their forecasts. Mining farms, each drawing anywhere from several to dozens of megawatts, collectively added a load comparable to that of a small city. And there were many such cities, and they appeared quickly.
Kazakhstan was forced to increase electricity imports from Russia—through the interconnected power system inherited from Soviet times. At the same time, KEGOC began restricting electricity supply to some industrial consumers, including legally registered mining businesses. The restrictions came through reduced power allocations: a farm designed for 50 megawatts might have its limit cut to 30 or lower. For a business where idle equipment means direct losses, it hurt.
Geography made matters worse. Kazakhstan’s power system has historically been split into three zones—northern, southern, and western—linked by limited transmission flows. The country’s south had an energy deficit even without mining: Almaty and its surroundings consumed more than local plants produced. Some farms, according to available reports, were located precisely in the south—closer to the Chinese border, but farther from the main coal stations. Every additional megawatt in the south put pressure on already overloaded transmission lines.
During the winter of 2021–2022, several regions of Kazakhstan experienced power outages. Establishing a direct causal link between mining and the blackouts is difficult—the system was also under pressure from abnormally cold weather, accidents at aging plants, and rising household consumption. But mining became a politically convenient and technically sound explanation: it had indeed added hundreds of megawatts to peak demand that had not been in the balance sheet.
One of the murkiest chapters in this story is the divide between official mining and the shadow kind. Kazakh authorities repeatedly said that registered farms were only the tip of the iceberg. Beneath them lay a mass of illegal operations: containers filled with equipment connected to the grid through front companies, through contracts for supposedly industrial consumption, through direct taps. Estimates varied, but officials said illegal mining might consume more electricity than legal mining.
It is virtually impossible to verify those estimates from the outside. An illegal farm files no reports, does not pay for over-limit consumption, does not answer the regulator’s requests. It just devours electricity—and that consumption gets smeared across the statistics as “network losses” or “unaccounted-for growth in industrial load.” For a power system running at the edge, the difference between accounted-for and unaccounted-for megawatts is the difference between a manageable situation and an accident.
I’ll admit it: I am struck by the speed with which a parallel infrastructure rose virtually from nothing. Chinese operators, experienced in working in gray zones back home, reproduced familiar schemes on new territory. Local intermediaries provided sites, contracts, connections. The business model was elementary: buy electricity at Kazakh rates, mine bitcoin at the global price, pocket the difference. As long as bitcoin was worth $40,000–60,000—and in November 2021 it approached $69,000—the margin justified any risk.
Kazakh law enforcement conducted raids, seized equipment, shut down farms. But the scale of the task—finding and cutting power to hundreds of scattered sites across a territory the size of Western Europe—made the fight resemble trying to catch water with a sieve.
On January 1, 2022, a law took effect introducing an additional charge for electricity consumed for digital mining: 1 tenge per kilowatt-hour on top of the base rate. The sum seems small—at the exchange rate of the time, around $0.0023. But for a large farm consuming dozens of megawatts around the clock, the extra payment added up to significant numbers. More importantly, the signal was unambiguous: the free banquet was over.
After that, the regime tightened. The authorities discussed and introduced higher rates, restrictions on connecting new capacity, registration requirements. Some farms began winding down or relocating—to Russia, Central Asia, Paraguay. Kazakhstan’s hash rate in global statistics began to decline: by 2022, the country’s share had fallen noticeably from its August peak.
January 2022 brought Kazakhstan another shock—mass protests that began with rising prices for liquefied gas and grew into a political crisis. During the unrest, the authorities shut down the internet across the country for several days. Bitcoin’s global hash rate dipped noticeably on those days—indirect confirmation that Kazakh capacity really was significant. An internet shutdown is a blunt instrument, but it produced a peculiar real-world experiment: when Kazakhstan went offline, the Bitcoin network felt it.
After the January events, the push for tighter controls became part of a broader agenda—controlling resources, restoring order, demonstrating sovereignty over domestic infrastructure. Mining went from a symbol of technological openness to a symbol of uncontrolled consumption.
The central paradox of Kazakhstan’s story lies in why the system was so unprepared for the miners’ arrival. A country with vast reserves of coal, oil, and gas, with installed power plant capacity formally sufficient to cover domestic demand, found itself in deficit within a few months. This is not a story about mining as evil: decades of low rates had made it impossible to renew generation, the grid had remained designed around Soviet industrial geography and was never rebuilt, and there were simply no mechanisms for responding quickly to a sudden surge in demand.
In essence, miners discovered a hole in the pricing system—electricity cost less than it took to produce it once equipment wear and the need for investment were factored in. They converted that price anomaly into bitcoins and took the profits out of the country. Kazakhstan received taxes—from legal farms—jobs—few, because mining is not labor-intensive—and a burden on the grid that exposed every problem that had been piling up.
It seems important to me that this scenario is not unique. Similar stories have unfolded in Iran, in individual U.S. states, in Abkhazia—everywhere cheap energy attracted mining faster than regulators could grasp the scale of it. Kazakhstan simply became the largest and fastest example.
By 2023, Kazakhstan’s share of global hash rate had fallen to a few percent, according to various estimates. Some farms had left, some had been shut down, some continued operating under the new regulatory conditions—with higher payments and restrictions. The power system received no magical cure: problems with generation and grids remained, and the government began discussing the construction of new capacity, including a nuclear power plant. Whether there will be enough time and money for that before the next winter peak remains an open question—and the answer depends on variables far more complicated than the price of bitcoin.