How Much Electricity Does Bitcoin Use Globally? (2025 Update)

The headline number (September 2025)

• Annual electricity use: ~138 TWh (annualised)

• Share of global electricity: ~0.54%

• Average continuous load: ~15.8 GW (138 TWh / 8,760 hours)

These figures come from the Cambridge Digital Mining Industry Report (April 2025), which synthesizes survey data and the updated CBECI methodology. Cambridge Judge Business School

Why trust Cambridge? The CBECI is the most-cited academic tracker, publishing a best-guess with lower/upper bounds and a transparent methodology page. In 2023 they revised assumptions to avoid periodic overestimation. ccaf.io+1

Why the estimate moves

1. Price & difficulty: higher prices attract hashpower until difficulty rises, changing total watts used.

2. Hardware efficiency: newer ASICs do more hashes per joule, bending power demand down for a given hashrate.

3. Network economics: miners switch on/off based on electricity costs and revenue.
   These drivers are the basis of the IEA’s and Cambridge’s explanations of Bitcoin’s energy use dynamics. IEA

How Cambridge measures it (in plain English)

• Bottom-up techno-economic model: build a basket of real-world miners, assume rational operators run gear that’s profitable at current prices, and compute watts needed network-wide.

• Outputs: a daily power demand (instantaneous) and an annualised consumption (TWh/yr).

• Context: Cambridge also publishes notes on bounds, efficiency trends, and comparisons, plus a public change log. ccaf.io+1

Put 138 TWh in context

• Cambridge translates that to ~0.54% of global electricity—a small slice in percentage terms but the size of a mid-sized country’s power use. Cambridge Judge Business School

• Global electricity demand is itself growing quickly through 2027, so Bitcoin’s share can shift even if mining’s absolute use is flat. IEA

Common misconceptions (quick takes)

• “Energy per transaction”: misleading. Most mining cost is to secure blocks; energy doesn’t scale linearly with the number of transactions. Focus on network security cost, not per-tx averages. (See Cambridge’s methodology and commentary on framing.) ccaf.io

• “One true number”: there isn’t one. Good trackers publish ranges; Cambridge has even revised historical figures when evidence improved. Cambridge Judge Business School

• “All estimates agree”: they don’t. Some trackers (e.g., Digiconomist) often publish higher numbers due to different assumptions; methodology matters. Digiconomist

What could change the footprint next

• Hardware cycles: new ASICs with better J/TH reduce watts per hash.

• Energy sourcing: shifts toward curtailed renewables, hydro seasons, or gas-flare mitigation change emissions even if TWh stays similar.

• Macro & policy: power-price shocks, carbon markets, and grid-service programs can push miners to different regions or demand-response roles. (Cambridge’s 2025 study tracks mix and sourcing trends.) Cambridge Judge Business School

Methodology notes & sources (for your readers)

• Primary: Cambridge Digital Mining Industry Report (Apr 2025); ~138 TWh and ~0.54% share. Cambridge Judge Business School

• Method & bounds: CBECI methodology + change log (explains how estimates are built and updated). ccaf.io+1

• Background explainer: IEA on what drives Bitcoin’s energy use. IEA

• Global demand context: IEA Electricity 2025 outlook for overall consumption growth. IEA

Keep exploring with Coin Miner

• Blog home: https://coinminer.com.au/blogs/crypto-mining-blog

• Mining calculator: https://coinminer.com.au/pages/mining-calculator
  

Educational content only. Figures reflect sources available as of September 29, 2025 (AEST) and may change as networks and methodologies update.