How Much Electricity Does Crypto Mining Really Use?
What crypto mining actually consumes, where the power comes from, and why efficient hosting decides who profits.
Crypto mining's electricity appetite is the most argued-about number in the industry — and the most misunderstood. The honest answer in 2026 is this: the Bitcoin network draws roughly 170–180 terawatt-hours a year, about 0.7–0.8% of all electricity generated on Earth, with more than half of that power now coming from renewable and recovered sources. This article breaks down exactly where that figure comes from, why "per-transaction" energy stats are misleading, how the grid impact compares to data centers and households, and — the part that actually decides whether a miner makes money — what a single kilowatt-hour costs depending on where the machine is plugged in.
Key takeaways
- ✓ Bitcoin mining uses ~170–180 TWh/year (Cambridge CBECI, Feb 2026) — comparable to Thailand or Vietnam, and roughly 0.7–0.8% of global electricity.
- ✓ Energy use tracks hashrate and hardware efficiency, NOT the number of transactions — "1,200 kWh per transaction" is a flawed metric.
- ✓ Over 52% of the network now runs on sustainable energy (hydro is the single largest source at ~23%), up from 37% in 2022.
- ✓ Profitability is decided at the meter: the same machine that loses money at $0.12/kWh prints cash at $0.04/kWh.
- ✓ OneMiners hosts on a 2,163 MW network at an average 7-year fixed rate of $0.0480/kWh — from $0.0364/kWh in Nigeria — which is why efficiency, not idealism, is the real energy story.
The real number: how much electricity crypto mining uses in 2026
Start with the headline figure, because everything else is context. The Cambridge Centre for Alternative Finance, whose Bitcoin Electricity Consumption Index (CBECI) is the most cited academic estimate in the industry, places Bitcoin's annualized electricity draw at roughly 170–180 TWh as of early 2026. That is approximately 0.7–0.8% of total global electricity generation — a measurable slice, but a small one. Put differently, the entire Bitcoin network consumes about as much power in a year as a mid-sized nation such as Thailand or Vietnam, and slightly more than the Netherlands' national consumption.
It is worth being precise about why estimates differ. CBECI uses a bottom-up methodology: it models the mix of mining hardware in service, each machine's energy efficiency in joules per terahash, the global hashrate, and regional electricity prices, then produces a best-guess band rather than a single hard number. Earlier 2025 snapshots from the same index sat closer to 138 TWh; the rise into the 170–180 TWh range reflects a higher network hashrate after the 2024 halving cycle and continued fleet expansion. Anyone quoting a single decimal-point figure for Bitcoin's consumption is overstating the precision — the honest framing is a range, and the range is what we use here.
Crucially, 'crypto mining' in 2026 is mostly Bitcoin mining. Ethereum, once the second-largest energy consumer, moved to proof-of-stake in 2022 and cut its electricity use by an estimated 99.9% overnight. That single design change removed the second-biggest sink from the equation, which is why the energy debate today is essentially a Bitcoin debate. You can model your own machine's draw against live network conditions with the OneMiners crypto mining calculators before you ever commit to hardware.
| Power source / site | 7-yr fixed rate ($/kWh) | Est. annual power bill | Verdict |
|---|---|---|---|
| Home / U.S. residential | $0.1500 | ~$6,480 | Usually unprofitable |
| OneMiners USA regional | $0.0455 | ~$1,966 | Strong margin |
| OneMiners Ethiopia (hydro) | $0.0399 | ~$1,724 | Excellent + clean |
| OneMiners Nigeria (cheapest active) | $0.0364 | ~$1,573 | Best-in-class |
Why 'energy per transaction' is the wrong way to measure it
You have almost certainly seen the claim that a single Bitcoin transaction 'uses 1,200 kWh' or even 1,338 kWh — enough to power an average home for weeks, or the equivalent of roughly 100,000 Visa swipes. The arithmetic is real, but the metric is misleading, and understanding why is the single most important concept in this entire topic.
Bitcoin's electricity consumption is driven by hashrate — the total computational effort securing the network — and that effort is essentially fixed by the protocol's difficulty adjustment regardless of how many transactions are in a block. Miners burn the same energy whether a block carries 2,000 transactions or 4,000. So dividing total network energy by the number of transactions produces a number that swings wildly based on mempool activity and Layer-2 batching, while telling you nothing about the actual physics. It is like measuring a city's electricity use by dividing it by the number of emails sent that day.
The defensible way to frame Bitcoin's energy is per unit of security or per unit of value settled, not per transaction. The network spends energy to make rewriting history economically impossible, and the Lightning Network already settles millions of off-chain payments that never touch the base layer's energy budget. When you read a viral 'per-transaction' statistic, treat it as a red flag for shallow analysis — the deeper question is always cost per kilowatt-hour and joules per terahash, which is exactly what a serious operator optimizes. Our how-it-works guide walks through how that security-energy relationship actually functions.
Where the power actually comes from: the renewable shift
The 'energy hog' label that dominated headlines from 2019 to 2022 is now factually out of date, and the data is unambiguous about why. According to the Cambridge Judge Business School's 2025 sustainability study, the sustainable-energy share of Bitcoin mining has crossed 52%, up from 37% in 2022. Hydroelectric power alone accounts for roughly 23% of the network's energy mix — the single largest source — followed by a growing contribution from wind, solar, nuclear, and recovered gas.
The recovered-gas story is the most underreported part of the picture. Oil fields routinely flare or vent natural gas that has no pipeline to reach, burning methane uselessly into the atmosphere. Mining operations now park modular data centers directly on well pads and convert that stranded gas into electricity. Power Engineering International reports that combusting methane in enclosed gas engines for mining reduces its global-warming impact by as much as 90% versus venting or flaring, and the Cambridge Centre for Alternative Finance estimates methane mitigation alone offsets about 5.5% of all Bitcoin network emissions. Mining, in these cases, is not just clean — it is actively cleaner than the alternative of doing nothing.
This is why carbon intensity is falling even as total consumption rises. Industry tracking compiled by SQ Magazine and Digiconomist shows the carbon intensity per mined Bitcoin dropping to roughly 358 kg CO₂e, down from 371 kg a year earlier — the result of a fleet that keeps migrating toward cheaper, cleaner electrons. OneMiners leans into exactly this trend: our hosting network is anchored by hydro and renewable-heavy sites such as Ethiopia (40 MW at $0.0399/kWh) and cold-climate facilities in Finland and Arctic Norway, where ambient cooling slashes the overhead energy that machines waste on heat removal.
How crypto mining's footprint compares to everything else
Context turns a scary number into an understandable one. At 170–180 TWh per year, Bitcoin mining sits below several categories of consumption that rarely make headlines. Global data centers — the backbone of cloud computing, streaming, and the explosive growth of AI — consume well over 400 TWh annually and are climbing fast. Residential air conditioning, idle household electronics in standby mode, and the worldwide gold-mining industry each command comparable or larger slices of the energy pie without provoking the same scrutiny.
That is not a deflection — energy use is energy use, and it deserves honest accounting. But proportion matters when allocating concern. Bitcoin's 0.7–0.8% share of global electricity buys a globally distributed, 24/7, censorship-resistant settlement network secured by physics rather than trust. Whether that is 'worth it' is a values judgment, but the comparison frame matters: the same society that runs 400+ TWh of data centers to serve video and AI is well within reason to run a fraction of that to secure a $2-trillion-plus monetary network.
The more interesting comparison for operators is convergence. As we cover in our analysis of the industry, the same infrastructure, power contracts, and cooling expertise that built efficient Bitcoin mining is now being repurposed for AI and high-performance computing — two workloads that share an identical core problem: how to turn cheap, abundant electricity into useful computation at the lowest possible cost per kilowatt-hour.
The number that actually matters: cost per kilowatt-hour
Here is the pivot that separates a curiosity question from a business decision. How much electricity crypto mining uses globally is interesting; how much YOUR electricity costs is everything. A modern hydro-cooled ASIC such as the Antminer S23 Hydro draws on the order of 5,000+ watts. Run that machine continuously for a month and it consumes roughly 3,600 kWh. The only variable that decides whether those kilowatt-hours produce profit or losses is the price you pay per unit.
Do the math at home. At a typical U.S. residential rate of $0.15/kWh, that single machine's monthly power bill exceeds $540 — frequently more than the Bitcoin it mines is worth, which is why home mining is a losing game almost everywhere outside subsidized-power regions. Drop the rate to $0.0455/kWh, the OneMiners U.S. regional fixed rate, and the same machine's bill falls to about $164. At Nigeria's $0.0364/kWh — our cheapest active site — it falls below $135. That spread of $400+ per machine per month, multiplied across a fleet, is the entire ballgame.
This is the precise reason industrial hosting beats home mining and why electricity rate, not hardware price, is the first number a serious miner negotiates. OneMiners locks its headline rates for up to 7 years on prepaid energy, removing the single biggest variable in the profitability equation. You can browse every machine and its draw across the full ASIC catalog and model the economics against a fixed rate rather than a volatile utility bill.
Worked example: the same machine at three electricity rates
Numbers make this concrete. Take a single 5 kW-class machine running 24/7 — about 3,600 kWh per month, or roughly 43,200 kWh per year. The table and chart below hold the hardware constant and change only the electricity rate, so you can see exactly how much of the profitability question is decided before the machine even powers on.
The lesson is brutal and simple: identical hardware, identical hashrate, identical Bitcoin price — and a swing of over $5,000 per machine per year purely from where it is plugged in. At residential rates the machine is often underwater; at a fixed industrial rate it is a cash-flowing asset. That gap is BTC-price-independent, which is what makes it the most reliable lever in the entire business. It is also why OneMiners' 0% management fees and hosting locations are engineered around the rate first and everything else second.
- Home/residential ($0.15/kWh): ~$6,480/year in power — often exceeds mined revenue.
- OneMiners USA regional ($0.0455/kWh): ~$1,966/year in power — a ~$4,500 annual swing per machine.
- OneMiners Nigeria ($0.0364/kWh): ~$1,573/year in power — the lowest active-site cost on our network.
- Multiply by a 10-machine fleet and the rate choice alone is worth $40,000–$50,000 per year.
What drives a miner's electricity use: efficiency in joules per terahash
Not all consumption is equal. Two machines can draw the same wattage and earn wildly different amounts, because what matters is efficiency — the energy spent per unit of useful work, measured in joules per terahash (J/TH). A 2020-era Antminer S19 runs around 30+ J/TH; a 2026 flagship such as the S23 Hydro pushes below 10 J/TH. That means newer hardware produces two to three times the hashrate for every kilowatt-hour it burns.
This is why old machines die first when difficulty rises or the price dips: they consume too much electricity for the work they do, and their break-even electricity rate creeps upward until even cheap power can't save them. Efficiency is also why the network's total energy use can climb while its carbon intensity falls — the fleet keeps replacing inefficient units with machines that wring more security out of every electron. When you evaluate a purchase, the J/TH figure tells you the machine's long-term electricity destiny far more reliably than its sticker price.
Air-cooled flagships like the Antminer S21 XP and hydro units like the Whatsminer M63S sit at the efficient frontier in 2026, and even altcoin-focused machines such as the Antminer L9 Scrypt miner and IceRiver KS-series Kaspa units are judged on the same J/TH discipline. The most efficient machine on a fixed low rate is the combination that survives every market cycle.
Is crypto mining bad for the environment? An honest verdict
The fair answer in 2026 is: it depends entirely on the power source, and the power source is rapidly improving. A coal-fed mine in a high-carbon grid is genuinely dirty. A hydro-powered facility in Ethiopia, a wind-and-nuclear site in the U.S., or a stranded-gas operation that destroys methane it would otherwise vent can be carbon-neutral or even carbon-negative. With over 52% of the network now sustainable per Cambridge Judge Business School data and carbon intensity per coin falling year over year, the trajectory is unambiguously toward cleaner.
There is also a grid-services argument that rarely reaches mainstream coverage. Because miners are perfectly interruptible loads — they can shut off in seconds — they act as a flexible buffer that lets grids absorb more intermittent renewable capacity. In Texas, mining operations curtail during demand spikes and soak up surplus wind power overnight, improving the economics of building renewables in the first place. Mining can be a tool for the energy transition, not just a drain on it.
Where OneMiners fits this verdict is straightforward: we steer hosted hardware toward the cleanest, cheapest electrons available — hydro in Ethiopia and Paraguay, cold-climate efficiency in Finland and Norway, and large-scale renewable buildouts including a planned 780 MW U.S. expansion at $0.0399/kWh. The greenest kilowatt-hour and the cheapest kilowatt-hour are increasingly the same kilowatt-hour, which aligns profit with sustainability rather than pitting them against each other.
Why hosting beats home mining on the energy math alone
Everything above converges on one conclusion: the electricity question is won or lost at industrial scale. A home miner pays retail residential rates, eats the full cost of cooling in a non-purpose-built space, absorbs noise and heat, and shoulders 100% of downtime risk. An industrial host buys power in bulk on long-term contracts, builds purpose-engineered cooling, and spreads infrastructure cost across thousands of machines — which is precisely how the per-kWh price drops from $0.15 to under $0.05.
OneMiners is built around this economic reality. Our network spans roughly 2,163 MW across 20 sites in six countries, hashing at 176,760 PH/s, with a 95%+ uptime SLA, a 7-year hardware warranty, 0% management fees, and fully managed remote-control operation through an app. The average fixed rate across the network is $0.0480/kWh, and for buyers who want to scale without tying up capital, Buy Now Pay Later starts at 25% down. The result is that the energy-cost advantage that used to belong only to billion-dollar public miners is available to individual buyers.
That is the entire thesis of efficient mining in 2026: you do not control Bitcoin's price or the network's difficulty, but you can lock the one input that decides whether your kilowatt-hours produce profit. Compare the full hosting network and the hardware catalog, then run the numbers through the profitability calculators before you commit a single watt.
Frequently asked questions
How much electricity does Bitcoin mining use per year?
As of early 2026, the Cambridge CBECI estimates roughly 170–180 TWh annually — about 0.7–0.8% of global electricity, comparable to a country like Thailand or Vietnam. Model your own machine's draw with the OneMiners calculators.
How much electricity does one transaction use?
Headlines cite 1,200–1,338 kWh per transaction, but this is a misleading metric. Bitcoin's energy is driven by hashrate, not transaction count — miners burn the same power whether a block is full or nearly empty. See how that works in our how-it-works guide.
Is Bitcoin mining bad for the environment?
It depends on the power source, which is rapidly cleaning up. Over 52% of the network now runs on sustainable energy (Cambridge Judge Business School), hydro is the largest single source, and stranded-gas mining can cut methane's warming impact by up to 90%. OneMiners prioritizes hydro and renewable-heavy hosting sites.
What percentage of Bitcoin mining is renewable?
More than 52% of the network is powered by sustainable energy in 2026, up from 37% in 2022, with hydroelectric alone at roughly 23%. Carbon intensity per coin has fallen to about 358 kg CO₂e.
How does Bitcoin's energy use compare to data centers?
Global data centers consume well over 400 TWh annually — more than double Bitcoin's ~175 TWh — and are growing faster due to AI. The same low-cost-power expertise increasingly serves both workloads.
What electricity rate do I need for mining to be profitable?
As a rule of thumb, sub-$0.06/kWh is where modern hardware is reliably profitable, and below $0.05/kWh is excellent. OneMiners offers 7-year fixed rates from $0.0364/kWh; browse the hosting network.
How much power does an ASIC miner draw?
A modern flagship like the Antminer S23 Hydro draws 5,000+ watts, consuming roughly 3,600 kWh per month running 24/7. Efficiency, measured in joules per terahash, determines how much useful work each kWh buys.
Why is hosting cheaper than mining at home?
Industrial hosts buy power in bulk on long-term contracts and build purpose-engineered cooling, dropping the per-kWh price from ~$0.15 retail to under $0.05. That single difference is usually worth $4,000+ per machine per year. Compare options across the OneMiners hosting centers.
Does Ethereum still use as much electricity as Bitcoin?
No. Ethereum switched to proof-of-stake in 2022 and cut its energy use by an estimated 99.9%, which is why the crypto energy debate today is essentially a Bitcoin debate.
