Cryptocurrency Mining
Bitcoin mining consumes roughly 150 TWh of electricity per year, comparable to Poland or Argentina. After China banned mining in mid-2021, the network rebalanced within months. Cambridge's last country map, from January 2022, put the US at 38%, followed by China at 21% despite the ban, then Kazakhstan, Canada and Russia. Texas is the largest US hub, attracted by cheap power and a permissive regulatory environment.
Key insights
Energy intensity is structural, not incidental
Bitcoin's Proof-of-Work consensus mechanism is intentionally energy-intensive — the security model requires mining hardware to be expensive to operate. Total energy use scales roughly with Bitcoin price (high prices = more mining profitable = more miners join until margins compress). Estimated network energy use grew about 5× from 2017 to 2024. Ethereum's move to Proof-of-Stake in 2022 reduced its energy use 99%+; Bitcoin has not done this and is unlikely to.
China's ban scrambled the geography
In 2020, China hosted about 65% of global hash rate. The June 2021 mining ban (motivated by energy targets and crypto-trading restrictions) forced wholesale migration. By late 2021 the US had become the largest location, but China did not disappear. Texas, in particular, attracted miners with cheap wind/gas power and ERCOT's interruptible-load market design. Kazakhstan surged and then hit power grid problems. Cambridge's January 2022 data also showed China back at about 21%, as underground mining resurfaced.
Renewable mining: real but partial
Mining at sites with otherwise-curtailed renewable energy (Texas wind, Paraguay hydro, Iceland geothermal) is genuinely renewable-driven. But much US mining runs on grid power that's significantly fossil-derived. Mining advocates argue miners help renewable economics by providing interruptible demand; critics note the alternative use of that electricity for other purposes is forgone. The 'green mining' debate has not been settled empirically.
Estimated Bitcoin mining electricity use 2017–2024
TWh per year, Cambridge Bitcoin Electricity Consumption Index
Show data table
| TWh/year | |
|---|---|
| 2017 | 30 |
| 2018 | 55 |
| 2019 | 70 |
| 2020 | 75 |
| 2021 | 90 |
| 2022 | 105 |
| 2023 | 135 |
| 2024 | 150 |
Key Finding: Energy use has risen steadily, broadly following mining revenue, though the annual figures here smooth over 2022's price collapse. Cambridge's live index read about 138 TWh in mid-2026 on secondary reports.
Bitcoin hash rate distribution by country (Jan 2022, latest Cambridge map)
% of global hash rate
Show data table
| % global hash rate | |
|---|---|
| USA | 37.8 |
| Mainland China | 21.1 |
| Kazakhstan | 13.2 |
| Canada | 6.5 |
| Russia | 4.7 |
| Other | 16.7 |
Key Finding: The US led with 37.8%, but China still held 21.1% after the ban. Kazakhstan (13.2%), Canada (6.5%) and Russia (4.7%) followed. Cambridge stopped updating this map, so more recent country shares are estimates.
Methodology & caveats
Estimating energy use
CBECI's methodology starts from network hash rate, calculates the mix of mining hardware in operation (efficiency varies from older ASICs ~50 J/TH to current Antminer S21 ~17 J/TH), estimates idle/active percentages, and infers electricity consumption. Uncertainty bands are ±25% but the central estimate is well-calibrated against independent operator disclosures.
Hash rate distribution caveats
Hash rate geographic distribution is estimated from mining-pool IP traffic — captures where mining pool servers are, not necessarily where miners are physically. VPN use complicates this. This matters for China: pool data showed about 0% in mid-2021 and about 21% by January 2022, which Cambridge attributed to underground mining that had been hidden behind VPNs and foreign proxies.
Comparison to other industries
150 TWh is similar to gold mining electricity use (~130 TWh/year) and about a third of global data centers (415 TWh in 2024). As share of world electricity (~28,500 TWh) it's ~0.5%. As share of world CO₂ emissions, it depends on electricity mix where mining occurs — estimates range from 0.1% to 0.4%.
Why estimates of Bitcoin's electricity use differ
Nobody meters global mining. Cambridge infers power use from the network hash rate and the assumed efficiency of the machines running, so its range is wide: a mid-2026 reading reported at about 138 TWh sat between lower and upper bounds of roughly 74 and 242 TWh. Digiconomist uses an economic method based on miner revenue and tends to give higher numbers. A headline figure stated without its range suggests more precision than exists.
The main caveat on geography
Cambridge's map relied on data from a handful of mining pools and has not been updated since January 2022, so every current country ranking is an estimate from company filings or surveys. The China episode is the warning: its share read as zero, then about 21% once the VPN effect became visible. Treat the US lead as well supported and exact shares as uncertain. The April 2024 halving also matters, because it cut miners' block reward and pushed out older, less efficient machines.
Cite this page
GlobeChart. “Cryptocurrency Mining.” Last updated 2026-10-31. https://globechart.com/technology/cryptocurrency-mining/
Please also credit the original publisher named under each chart: Cambridge Bitcoin Electricity Consumption Index; CBECI Bitcoin Mining Map. Spotted an error? Tell us.