Bitcoin Energy Consumption 2026: 138 TWh (Cambridge CBECI)

As of July 2026, Bitcoin uses roughly 140 to 205 terawatt-hours (TWh) of electricity per year, about as much as a mid-size country such as Poland or Thailand, and around half a percent of total global electricity consumption. The exact figure depends on the model, but every credible estimate puts annualized Bitcoin electricity use in the mid-hundreds of TWh.

The range reflects two respected sources that model the network differently, not sloppiness. The Cambridge Bitcoin Electricity Consumption Index (CBECI) sits at the lower end (a live estimate of 138.2 TWh as of July 29, 2026); Digiconomist sits near the top (204.4 TWh). This page explains how the number is measured, why the two estimates diverge, how it has changed over time, what fuels it, and why the popular "per-transaction" figure is misleading.

You cannot meter Bitcoin directly, so researchers estimate it. The Cambridge Centre for Alternative Finance builds the CBECI by taking the network's total hashrate (the collective computing power securing the chain), mapping it to the mix of mining hardware likely in use, and multiplying by each machine's power draw, then annualizing as if that power level held constant for a full year.

Because the hardware mix is uncertain, Cambridge reports a range rather than a single number:

Cambridge's 2025 sustainability study, based on a survey representing about 48% of global mining activity, estimated Bitcoin's annual electricity consumption at roughly 138 TWh, about 0.5% of global consumption. Digiconomist puts the figure closer to 197 TWh.

The two estimates diverge because they model the network differently. Cambridge works bottom-up from observed hashrate and a probable mix of mining hardware, then takes a mid-point. Digiconomist works from mining economics, assuming miners spend a roughly fixed share of their revenue on electricity, so its estimate climbs faster when Bitcoin's price is high. Neither is "wrong"; they answer slightly different questions. Quoting a sourced range (roughly 140-205 TWh as of July 2026) is more honest than a single decimal, and anyone citing "Bitcoin uses X TWh" should say which model they mean.

The number also moves with the market. Preliminary Cambridge data reported by crypto.news put annualized mining demand near 190 TWh in December 2025, at the peak of the last bull market, before the 2026 drawdown pushed less-efficient miners offline and the live CBECI estimate eased back toward the high 130s.

Abstract TWh figures are hard to picture, so put them next to whole countries:

• Bitcoin's ~140-205 TWh/year is comparable to the total annual electricity consumption of a country like Poland (lower estimate) to Thailand (higher estimate). Digiconomist explicitly benchmarks its ~204 TWh figure against Thailand.
• It is roughly twice Finland's national electricity demand (~85-95 TWh).
• It is a small slice, about half a percent, of the world's total electricity use.

So Bitcoin uses a meaningful amount of power in absolute terms, but a small share globally.

Bitcoin's energy footprint has grown with its price and hashrate, but not linearly. It dips during bear markets when unprofitable miners switch off, and climbs during bull runs. The long-run direction is up, roughly tracking adoption: as more people own Bitcoin and its price rises, more computing power competes to secure the network each cycle. Efficiency gains cut the other way: newer ASIC miners do far more hashing per watt than older models, which softens the growth in energy use relative to the growth in hashrate.

Bitcoin mining's fuel mix has shifted meaningfully toward lower-carbon sources. The Cambridge 2025 study found that sustainable energy sources now account for 52.4% of Bitcoin mining (42.6% renewables plus 9.8% nuclear), up from about 37.6% in 2022.

The full breakdown from that study:

Natural gas has overtaken coal as the single largest source (coal fell from 36.6% in 2022 to 8.9%), and miners increasingly use otherwise-flared or stranded gas that would be wasted anyway.

You will often see headlines like "one Bitcoin transaction uses as much energy as a household for X weeks." That framing is misleading. As Bitcoin Magazine and the International Energy Agency both explain, Bitcoin's energy use is driven by mining, not by transaction throughput.

The network consumes the same power whether it processes one transaction or one thousand in a block. Mining energy scales with Bitcoin's price and the resulting competition among miners, not with how many payments settle. Dividing total energy by transaction count produces a number that swings wildly and says nothing about the marginal cost of a payment, especially since Layer-2 solutions batch huge numbers of transfers off-chain.

Comparing Bitcoin to a whole country is popular but not very useful. A more apples-to-apples comparison is to the industries Bitcoin partly competes with. One widely cited analysis estimated:

By this estimate, Bitcoin uses less than half the energy of either the banking system or gold mining. The caveat cuts both ways: Bitcoin's energy use is transparent and easy to measure, while banking and gold footprints are far harder to pin down, which makes Bitcoin an easy target for scrutiny that comparable industries rarely face.

Roughly 140-205 TWh annually as of July 2026, depending on the model. The Cambridge CBECI live estimate is 138.2 TWh, while Digiconomist estimates about 204 TWh. That is around half a percent of global electricity consumption.

138.2 TWh, the CBECI live best-guess estimate as of July 29, 2026, published at ccaf.io/cbnsi/cbeci. Cambridge pairs that central figure with a theoretical lower bound near 74 TWh and an upper bound near 242 TWh, because the exact mix of mining hardware on the network cannot be observed directly. The index recalculates every 24 hours, so the live figure drifts with hashrate.

Cambridge's 2025 sustainability study, built on a survey covering about 48% of global mining activity, estimated Bitcoin's annual electricity consumption at roughly 138 TWh, around 0.5% of world consumption. Preliminary CCAF data reported by crypto.news put annualized demand near 190 TWh in December 2025, at the peak of the bull market, before the 2026 drawdown pushed less-efficient miners offline.

The CBECI annualized figure of 138.2 TWh works out to roughly 15.8 gigawatts of continuous power draw (138.2 TWh divided by the 8,760 hours in a year). Cambridge publishes that live power demand in gigawatts alongside the annualized TWh number, and the two are just the same measurement expressed over different time windows.

There is no official figure, because no authority meters the Bitcoin network. The CBECI is the closest thing to a reference standard: it is maintained by the Cambridge Centre for Alternative Finance at the University of Cambridge's Judge Business School, it publishes its methodology openly, and it is the estimate most cited by researchers and regulators. Digiconomist is the main alternative and runs higher, at 204.4 TWh.

They model the network from opposite directions. Cambridge works bottom-up from observed hashrate and a probable mix of mining hardware, then takes a mid-point. Digiconomist works from mining economics, assuming miners spend a roughly fixed share of revenue on electricity, so its estimate climbs faster when Bitcoin's price is high. Neither is wrong; quoting the sourced range of roughly 140-205 TWh is more honest than a single decimal.

The CBECI is the most widely cited academic estimate of Bitcoin's electricity use, maintained by the Cambridge Centre for Alternative Finance at the University of Cambridge's Judge Business School. The official site is ccaf.io/cbnsi/cbeci; it updates every 24 hours and publishes a best-guess annualized figure (138.2 TWh as of July 29, 2026) alongside theoretical lower and upper bounds (~74 and ~242 TWh) and the network's live power demand in gigawatts.

It depends on the fuel mix. The Cambridge 2025 study found 52.4% of Bitcoin mining now runs on sustainable energy (renewables plus nuclear), with natural gas replacing coal as the top fossil source. Bitcoin also uses less than half the energy of the banking or gold sectors by some estimates.

Because Bitcoin's energy use comes from mining, not transaction volume. As Bitcoin Magazine and the IEA note, the network draws the same power regardless of how many payments it settles, so dividing total energy by transactions produces a misleading figure.

No. Spending or sending Bitcoin adds a transaction to a block that would be mined anyway, so the marginal energy cost of a payment is essentially zero. On efficient chains like Solana, the per-transaction energy footprint is negligible by design.

Bitcoin's ~140-205 TWh/year is comparable to a mid-size country like Poland or Thailand, and about twice Finland's national demand, but still only about half a percent of global electricity.

If your goal is simply to use crypto for everyday purchases, the energy question is largely beside the point. A single payment adds no meaningful load to the network, and stablecoins on efficient chains settle for a fraction of a cent. A crypto debit card lets you pay with crypto anywhere Visa and Mastercard are accepted, so you can spend Bitcoin like cash without ever touching a mining rig. SolCard settles top-ups over Solana in seconds, making the per-payment energy footprint effectively nil.

• Cambridge Bitcoin Electricity Consumption Index (CBECI)
• Cambridge CBECI Methodology
• Cambridge Judge Business School -- Sustainable energy rising in Bitcoin mining (2025)
• Digiconomist -- Bitcoin Energy Consumption Index
• crypto.news -- Hydropower overtakes gas as Bitcoin mining power use jumps 38% (CCAF preliminary data)
• IEA -- Bitcoin energy use, mined the gap
• Bitcoin Magazine -- Energy Per Transaction Is a Misleading Metric