Bitcoin and Energy Consumption: What Is Documented and What Is Disputed

Why Bitcoin's energy consumption concerns businesses too

Anyone in a business who thinks about crypto assets quickly hears two statements. One is that Bitcoin uses as much electricity as an entire country. The other is that the network runs mostly on renewable energy. Both are put forward with great conviction, and both contain a kernel of truth.

This article therefore asks: What is documented about Bitcoin's energy consumption, what is disputed, and what does it mean for a business that accepts or holds Bitcoin? To answer it, we first clarify why Bitcoin uses electricity at all, then look at the best available figures, weigh two readings, and finally derive a grid for your own assessment.

The thinking model: proof of work as a security budget

Bitcoin is a public transaction register, the blockchain, maintained on many computers at once. Who may append a new block of transactions is decided by a computing competition, known as mining. Participants try out millions of numbers per second until one finds a solution that meets a condition set by the network. This verifiable effort is called proof of work. The winner receives newly created bitcoin and the transaction fees.

The electricity is not a by-product but the purpose. Anyone who wanted to falsify the blockchain after the fact would have to muster more computing power than all honest participants combined. The more energy there is in the network, the more expensive an attack would be. I call this the security budget: the reward for miners determines how much they can spend on electricity and equipment, and thus how well the network is protected.

An important property follows from this. Electricity consumption does not depend on the number of payments but on the value of the reward. If the price rises, more mining pays off. Roughly every four years, the number of new bitcoin per block halves, to 3.125 since 2024, as Bitcoin in fifteen years describes. Whether consumption falls in the long run therefore depends mainly on the price and on transaction fees.

An analogy makes this tangible. A safe is not secure because there is a lot of money in it but because it is expensive to break open. Bitcoin builds the safe out of computing work. The price for this is energy, and it is paid anew every day.

Bitcoin does not use electricity despite its security but because of it. Anyone who criticizes the one has to talk about the other.

What is documented: consumption, electricity mix, emissions

The most frequently cited source is the Cambridge Centre for Alternative Finance at the University of Cambridge. For its report of April 2025, it surveyed 49 mining companies from 23 countries that together cover around 48 percent of global mining activity. Around 75 percent of the reported activity was in the United States.

Compared with 2022, the mix has shifted considerably. The share of sustainable sources rose from 37.6 to 52.4 percent, coal fell from 36.6 to 8.9 percent, and natural gas rose from 25 to 38.2 percent. One major reason is geographic: after China banned mining in 2021, a large part of the activity moved to the United States. Xiao and co-authors (2023) estimate mining's earlier contribution to emissions in China at around 78 million tonnes of CO₂ in total for 2017 to 2021.

The figures have an important limitation. The electricity mix is based on self-reported information from the surveyed companies, which cover about half of the network, and US companies are overrepresented. It is a snapshot, not an audited overall balance.

What is disputed: estimates, points of reference, attribution

Ibañez and Freier (2023) compiled the state of research in a review. Their first finding: estimates diverge widely because nobody knows exactly which devices are running and on which electricity. In 2022, the White House Office of Science and Technology Policy gave a range of 72 to 185 terawatt hours per year.

The second point deserves particular attention because it is the one most often misused in everyday debate. Ibañez and Freier show that there are several ways to attribute emissions: per transaction, per bitcoin held, or a mix of both. Because consumption depends on the security budget and not on the number of payments, the figure 'energy per transfer' says little about what an additional payment actually causes. It is nevertheless a widespread argument.

For comparison: other blockchains use proof of stake instead of proof of work, a method in which participants deposit crypto assets as collateral instead of spending computing power. Ethereum switched to this method in 2022 and, according to the Ethereum Foundation, cut its electricity consumption by more than 99 percent. No such switch is planned for Bitcoin.

Two readings: waste or a building block of the energy transition?

Ibañez and Freier (2023) mainly examine the second reading. They conclude that mining with renewable energy could, under certain conditions, have a net relieving effect on power grids because, as a flexible load, it absorbs surplus electricity and switches off in times of scarcity. But they stress that this requires adjustments in practice and that mining competes with other flexible consumers such as battery storage.

I consider both readings too simple when they stand alone. It is documented that consumption is high and the mix has become cleaner. It is not documented that mining accelerates the energy transition on a large scale. Whether it does is decided site by site. For the assessment, what matters is therefore less the overall figure than the question of where the electricity comes from in the specific case.

What this means for a business

A single Bitcoin payment causes only a small additional consumption for a business because the network runs regardless. The question still reaches the business, in three ways.

1. Customers and banks ask. Anyone who has to document sustainability in tenders will be asked how crypto assets fit into their own balance. What such questions look like is described in ESG ratings.
2. Statements must be accurate. Anyone who accepts Bitcoin and advertises sustainability at the same time needs evidence. Blanket statements such as 'climate-neutral payment' are risky, as Greenwashing shows.
3. Providers must disclose information. The EU regulation MiCA requires providers of crypto-asset services to publish information on the main impacts of the respective consensus mechanism on climate and the environment. You can request this information from your provider.

Whether crypto payments are worthwhile for a business at all is a separate question, which Crypto as a means of payment answers with a four-question filter. Anyone who wants to hold Bitcoin will find the trade-offs in Bitcoin on the balance sheet.

The documented-or-disputed grid

From the thinking model and the findings I derive a simple grid that I call the documented-or-disputed grid. It helps you classify any statement about energy consumption before you adopt it or pass it on.

1. Check the source

   Does the figure come from a traceable survey such as Cambridge's, from an estimate with a range, or from a provider's advertising?

2. Check the point of reference

   Is consumption given per transaction, per year, or in comparison with a country or an industry? Does the point of reference fit the question?

3. Check the electricity mix

   Is the stated share of clean energy based on self-reports, grid data, or certificates?

4. Check the date

   Does the figure date from before 2021, when a large part of mining took place in China, or from today?

Transferability and limits

The dispute over Bitcoin's energy consumption shows a pattern that appears with many new technologies: a large figure becomes a symbol, and both sides choose the point of reference that supports their view. Anyone who has to decide then needs clarity above all about which question they want to answer.

The limits of this assessment are clear. The Cambridge figures are based on surveys and models, not on complete measurements. More recent projections point to a further increase but were preliminary at the time of writing. The review by Ibañez and Freier relies in part on industry reports. The figures will continue to change with price, technology, and locations.

Conclusion: high, cleaner, and still disputed

To the guiding question: it is documented that, according to the Cambridge survey, the Bitcoin network uses around 138 terawatt hours a year, about 0.5 percent of global electricity, and that the share of sustainable sources among surveyed miners has risen to just over half. The security budget explains why consumption depends on the value of the reward and not on the number of payments. The right point of reference, the reliability of the electricity mix, and whether mining can support the energy transition remain disputed.

For a business this means: adopt neither the country comparison nor the green power promise without checking. Check source, point of reference, electricity mix, and date, and make only statements you can back up. This article is not investment advice.