Bitcoin mining consumes roughly as much electricity each year as a mid-sized country. That comparison is accurate, and it is also incomplete. Bitcoin’s environmental impact spans energy consumption, carbon emissions, water use, and hardware waste, and the data behind each of those has shifted considerably since most people formed their view of the issue.
This article covers what the most recent research actually shows, where the numbers come from, why they differ, and what is genuinely contested versus what is settled.
Use the multichain Bitcoin.com Wallet app, trusted by millions to safely and easily send, receive, buy, sell, trade, and manage Bitcoin and the most popular cryptocurrencies.
Key Takeaways
- Bitcoin mining consumes roughly 0.5% of global electricity, comparable to a mid-sized country, though estimates vary by methodology
- More than half of Bitcoin’s electricity now comes from sustainable sources, up from around a third in 2022, according to the Cambridge Centre for Alternative Finance (CCAF)
- Coal’s share of mining energy has fallen from 36.6% to 8.9% since 2022; natural gas is now the single largest individual source
- Bitcoin’s environmental footprint extends beyond energy: water consumption and hardware waste are real, measurable impacts that receive far less coverage
- Bitcoin’s energy use is tied to its price and miner competition, not to how many transactions the network processes
- Mining hardware has improved roughly 7x in energy efficiency over the past decade, meaning network growth consistently outpaces energy growth
- A growing number of mining companies are converting infrastructure to AI data centers, reducing Bitcoin’s direct mining footprint but raising new questions about whether the environmental benefit is real
The two inner black circles are identical in size. Context changes how we perceive them, the same way that choosing what to compare Bitcoin’s energy use to can make an identical figure read as alarming or unremarkable.
Why Bitcoin Uses Energy in the First Place
Bitcoin secures its network through a process called proof of work (PoW). To add a new batch of transactions to the blockchain, specialized computers called miners race to solve a cryptographic puzzle. The first machine to find the correct answer wins the block reward and any transaction fees in that block. Every other miner’s work on that round is discarded.
The energy expenditure is deliberate. It is what makes the Bitcoin ledger expensive to falsify. Anyone trying to rewrite the chain would need to redo the computational work of the entire honest network, a cost that scales with Bitcoin’s price and the hardware invested in mining globally.
This structure has a direct consequence for how to read bitcoin energy consumption figures: Bitcoin’s electricity use correlates with its price, not with how many transactions it processes. The network consumes roughly the same amount of bitcoin mining electricity whether it confirms one transaction or one million in a given period. Energy use is driven by miner competition for the block reward, which is tied to Bitcoin’s market value. Most headlines get this relationship wrong, and it matters for interpreting every statistic that follows.
How Big Is Bitcoin’s Environmental Impact?
Bitcoin’s environmental impact covers three measurable categories: greenhouse gas emissions from electricity generation, water consumption from cooling systems and power plants, and bitcoin e-waste from obsolete mining hardware.
CategoryEstimated Scale (2025)Comparable BenchmarkAnnual electricity consumption~138 TWhPoland or Argentina (annual use)Annual CO2 emissions~39.8 Mt CO2eSlovakia’s total national emissionsAnnual water consumption~2,772 gigalitersSwitzerland’s total annual water useAnnual electronic waste~20.75 kilotonnesDisputed; see below
The most comprehensive recent dataset is the Cambridge Digital Mining Industry Report (April 2025), produced by the Cambridge Centre for Alternative Finance (CCAF). The CCAF surveyed 49 mining firms across 23 countries, covering 48% of global Bitcoin hashrate by self-report. Their estimate for annual electricity consumption is 138 terawatt-hours (TWh), or roughly 0.5% of global electricity generation. The Digiconomist Bitcoin Energy Consumption Index, which uses a different methodology, puts consumption higher, at around 175 TWh. These are not contradictory figures. They reflect different assumptions about average hardware efficiency across the full network. Both sit in the same order of magnitude, and both represent a significant real-world footprint.
Carbon Emissions
The CCAF’s April 2025 report puts network-wide bitcoin greenhouse gas emissions at 39.8 megatonnes of CO2-equivalent (MtCO2e), comparable to Slovakia’s total national output. This reflects a 24% year-over-year improvement in hardware efficiency among surveyed miners and a cleaner energy mix than in previous years. Other indices report higher figures. Digiconomist’s 2025 estimate lands closer to 98 million metric tons, comparable to Qatar. The gap reflects genuine methodological differences and incomplete visibility into the full network, particularly the unsampled miners in Russia, Central Asia, and underground China.
A Note on Per-Transaction Figures
You have probably seen claims like “one Bitcoin transaction uses as much electricity as two weeks of a typical household’s consumption.” These figures divide the network’s total energy use by its total transaction count. The problem is that Bitcoin’s energy use does not scale with transaction volume. The network consumes the same electricity whether it processes ten transactions or ten million in a given block period. Energy is determined by miner competition for the block reward, which is a function of Bitcoin’s price and hardware investment, not payment activity. Of the twelve most recent peer-reviewed studies on Bitcoin and energy, eleven had stopped using the per-transaction metric, because it produces comparisons that are technically derived but practically misleading. A 2025 study in Scientific Reports and analyses from the LSE Business Review both note this methodological shift. The meaningful figures are total network consumption and total associated emissions.
Water Consumption
Water consumption is the least-covered part of Bitcoin’s environmental footprint. Mining operations use water directly through liquid cooling at data centers, and indirectly through the thermal power plants that generate their electricity. The Digiconomist 2025 index estimates annual consumption at approximately 2,772 gigaliters, roughly equivalent to Switzerland’s total annual water use. This figure rarely appears in mainstream coverage, but it is a real and measurable impact.
Bitcoin E-Waste
ASIC mining hardware becomes economically obsolete as each more efficient generation arrives. The chips are purpose-built for Bitcoin’s SHA-256 algorithm and cannot be repurposed. When operators retire old machines, the hardware typically becomes scrap. Digiconomist’s 2025 index estimates annual bitcoin e-waste at approximately 20.75 kilotonnes. Some industry researchers have disputed this, citing incorrect assumptions about hardware lifespans in the underlying model. The precise figure is genuinely unsettled, but hardware waste is a real component of Bitcoin’s environmental costs and one that receives less scrutiny than it deserves.
Where Bitcoin Mining Actually Happens
Geography is one of the most consequential factors in Bitcoin’s carbon footprint, because the carbon intensity of electricity varies dramatically by region. An operation running on Icelandic geothermal power emits essentially nothing per bitcoin mined. The same machine on a Kazakhstani coal grid emits orders of magnitude more.
The distribution of mining activity changed fundamentally in 2021, when China banned cryptocurrency mining, eliminating roughly 65% of global hashrate almost overnight. That capacity relocated primarily to the United States, Kazakhstan, Russia, and other regions.
Country / RegionApprox. Hashrate Share (2025 to 2026)Primary Energy SourceEmissions ProfileUSA (Texas, Georgia, Kentucky)37 to 40%Mixed: gas, wind, nuclear, coalModerate; improving with renewablesRussia15 to 17%Predominantly fossil fuelsHigh carbon intensityKazakhstan~14%Predominantly coalHigh carbon intensityCanada (Quebec, British Columbia)~9%Predominantly hydroelectricVery lowParaguay~4%Near-100% hydroelectric surplusVery lowUnderground China10 to 12%Coal-heavy gridHigh carbon intensityIceland and Nordic countriesSmall shareGeothermal and hydroNear-zero
One important caveat from the Cambridge Bitcoin Electricity Consumption Index itself: the 52.4% sustainable energy figure applies only to the 48% of global hashrate covered by the survey. The unsampled portion, concentrated in Russia, China, and Central Asia, almost certainly skews more fossil-heavy. The sustainable energy share should be read as a floor estimate for the surveyed population, not a definitive global average.
What the data does confirm clearly: coal’s share of Bitcoin mining energy fell from 36.6% in 2022 to 8.9% in 2025. Natural gas is now the single largest individual source at 38.2%, while renewables and nuclear together account for 52.4% of the sampled mix (hydropower 23.4%, wind 15.4%, solar 3.2%, nuclear 9.8%). The industry that once ran heavily on Chinese coal has undergone a genuine structural shift, even if cryptocurrency sustainability advocates sometimes overstate how complete that shift is.
The Hardware Efficiency Story
Raw consumption numbers are difficult to interpret without understanding the trajectory of mining hardware.
ASIC miners (application-specific integrated circuit machines) are the only hardware used for Bitcoin mining today. Their efficiency is measured in joules per terahash (J/TH): how much electrical energy a machine consumes to perform one trillion SHA-256 calculations. Lower is better.
The original Bitmain Antminer S9, released in 2016, operated at around 98 J/TH. By 2026, the most efficient commercially available machines reach 13 to 15 J/TH. The Antminer S21 XP, Bitmain’s current flagship air-cooled model, operates at approximately 13.5 J/TH. The Antminer S21 Pro runs at roughly 15 J/TH. That is a roughly 7x improvement in energy efficiency over a decade, as documented by Spark.money’s 2026 mining economics analysis and manufacturer specifications.
The practical consequence is significant. Bitcoin’s network hashrate crossed 800 exahashes per second (EH/s) in Q1 2026, representing approximately 35% year-over-year growth. Over the same period, energy consumption grew by only an estimated 10 to 15%. The network became computationally much stronger while consuming proportionally less additional electricity, because newer machines replaced older ones. Projections built on older efficiency assumptions consistently overestimate how fast consumption grows. This does not make bitcoin mining energy use trivial, but it does mean the relationship between network growth and energy growth is not linear.
The Grid-Balancing Argument
One substantive argument in favour of Bitcoin mining’s role in the energy system deserves honest treatment: the claim that miners can act as a stabilizing force on electricity grids, particularly those with high renewable penetration.
Bitcoin mining is one of the most interruptible large-scale electricity loads in existence. Unlike a factory or a hospital, a mining operation can cut its entire power draw within seconds with no impact on any product or service. Other miners on the network pick up the computational work immediately. This makes miners natural candidates for demand-response programs, where grid operators pay large consumers to curtail usage during peak stress events.
In Texas, the Electric Reliability Council of Texas (ERCOT) has documented Bitcoin miners providing demand response and frequency regulation on an ongoing basis. During the July 2022 heatwave, miners curtailed consumption significantly, freeing capacity for residential and commercial use when the grid was under pressure.
Beyond curtailment, miners have a structural incentive to seek the cheapest available electricity. The cheapest electricity tends to be power that would otherwise go to waste: surplus hydroelectric generation in Paraguay (where the Itaipu and Yacyretá dams produce more than the country can consume), curtailed wind energy in West Texas (where transmission bottlenecks force generators to dump output they cannot sell), and flared natural gas at oil drilling sites (where methane is burned as a waste product rather than transported to market). A 2023 study published in ScienceDirect found evidence that Bitcoin mining can absorb excess energy, help balance grids, and support bitcoin renewable energy integration under specific conditions.
The limits of this argument are equally real. Mining operations in Russia, Kazakhstan, and underground China are not participating in renewables integration or grid balancing programs. The argument applies to a subset of miners in deregulated, renewable-heavy markets. It does not describe the full network, and it does not make Bitcoin carbon-neutral.
From Renewable Energy to Bitcoin Mining: A Conversation with CleanSpark CEO Zach Bradford
CleanSpark started as a renewable energy company before becoming one of the fastest-growing Bitcoin miners in the United States. In this interview, CEO Zach Bradford explains the operational philosophy behind scaling a mining business that treats energy strategy as a core competency, not an afterthought.
The AI Pivot and What It Means for Bitcoin’s Footprint
A significant structural shift is underway in the mining industry that has direct implications for Bitcoin’s environmental impact. Public mining companies including Core Scientific, IREN, TeraWulf, and Bitfarms have been converting their data center infrastructure from Bitcoin mining to artificial intelligence (AI) and high-performance computing (HPC) workloads. As of mid-2026, listed miners have announced more than $70 billion in cumulative AI and HPC contracts. Core Scientific secured roughly $10 billion through CoreWeave. IREN signed a $9.7 billion deal with Microsoft. TeraWulf has stated its intention to exit Bitcoin mining entirely.
On one level this reduces Bitcoin’s direct environmental footprint, because less infrastructure is dedicated to proof-of-work mining. Facilities being repurposed rather than decommissioned also reduces hardware waste, since existing power infrastructure and buildings are reused rather than scrapped.
The picture is more complicated than that, though. AI data centers require consistent, uninterrupted power, which is fundamentally different from the interruptible load model that made Bitcoin miners useful for grid balancing. A mining facility that could curtail its entire draw within seconds during a grid stress event becomes far less flexible once it is hosting AI inference workloads with uptime guarantees. As Spark.money’s 2026 energy analysis notes, as mining companies transition to AI hosting, their value as flexible grid load may diminish even as their total energy consumption increases.
The net environmental effect of the pivot is genuinely uncertain. Less Bitcoin mining means a smaller Bitcoin-specific footprint. But AI data centers are not carbon-neutral, require consistent rather than flexible power, and are growing rapidly in their own right. Whether the infrastructure repurposing reduces total emissions or simply redistributes them is a question the data does not yet clearly answer.
The Bigger Picture
Bitcoin’s environmental impact is real and significant. The network consumes electricity at the scale of a mid-sized country, emits tens of millions of tonnes of CO2 annually, and generates hardware waste that most coverage overlooks entirely. Those facts are not in dispute.
What has changed is the context around them. The energy mix is cleaner than it was three years ago, hardware is more efficient than it was five years ago, and the industry is restructuring in ways that will continue to shift the picture. The most accurate thing that can be said about Bitcoin’s environmental footprint right now is that it is improving, it remains substantial, and following the data rather than the headlines is the only reliable way to keep up with it.
