Scope 3 · Digital & IT · Cambridge CBECI
Bitcoin Emissions Calculator — Cambridge CBECI & Digiconomist, with Uncertainty Bands
Audit-grade CO₂e for Bitcoin exposure, computed live from the Cambridge Bitcoin Electricity Consumption Index and the Digiconomist Bitcoin Energy Consumption Index — the two authoritative, and openly disagreeing, sources. Choose Cambridge, Digiconomist, or the range between them; pick whether you are accounting holdings, mining electricity, or per-transaction; and read a central figure with an explicit uncertainty band. Because in Bitcoin accounting, a single point estimate with no band and no named model behind it is a warning sign, not a number.
Two models, both live and switchable. Every figure on this page is read live from MasterBrain, which carries the headline value and the uncertainty bound for each metric from both governing sources. The Cambridge CBECI/CDMIR estimate is bottom-up — it builds network electricity from hardware efficiency and hashrate, then applies a mining-map grid weighting. The Digiconomist BECI estimate is economic — it infers energy from miner revenue. They disagree by design, and the calculator lets you select either, or plot the Range between them.
From electricity to emissions. The network carbon figure is not a GreenCalculus grid conversion. It uses Cambridge’s own mining-mix weighting — per-country and per-province grid mix multiplied by life-cycle grid factors — yielding an effective intensity of 0.288 kgCO₂e/kWh for the Cambridge model against Digiconomist’s higher static 0.558 kgCO₂e/kWh. The mix is global and shifts as mining migrates between jurisdictions, which is why the same terawatt-hour of consumption can carry materially different carbon depending on where the machines run.
Three accounting bases. Holdings apportions a share of the network’s annual emissions to the coins you hold (your share of circulating supply × annual network CO₂e) — the basis a treasury holder or fund uses for financed emissions. Mining electricity multiplies the megawatt-hours your operation consumes by the mining-mix intensity — the basis a miner uses for operational Scope 2. Per-transaction is offered but flagged: it is a downstream allocation, not a marginal impact, and Cambridge declines to publish a per-transaction headline at all.
Two consumption bases, never blended. The consumption headline (170.42 TWh/yr) comes from CBECI’s live best-guess model; the emissions headline (39.8 Mt CO₂e/yr) sits on Cambridge’s survey-derived ~138 TWh base. These are different denominators by construction — the calculator cites each as published and does not divide one by the other to back out an intensity, because doing so would manufacture a figure neither source endorses.
Bitcoin proof-of-work only. This calculator models Bitcoin, which uses energy-intensive proof-of-work. Ethereum after its 2022 move to proof-of-stake, and other proof-of-stake chains, carry a profile orders of magnitude lower and are not modelled here. The result panel states this explicitly so a reader never mistakes a Bitcoin figure for a whole-crypto one.
Boundary. The figures cover use-phase electricity only. Excluded, per Cambridge’s own boundary statement: the embodied carbon of mining hardware, mining e-waste (roughly 2.3 kt/yr from fleet decommissioning), cooling water, data-centre overhead beyond IT load, and transmission losses. Cambridge estimates these other life-cycle stages at under 1% of the operational total and is explicit that the index “cannot be considered a life-cycle assessment.”
Blank = dated default (≈19.87M, mid-2026). Holdings basis only.
Choose an estimate basis and enter a value above to calculate
Results appear instantly. The uncertainty band (Cambridge ↔ Digiconomist), live network context, scope mapping, key insights and the full audit trail appear after calculation.
Results are indicative Bitcoin emissions estimates built from the Cambridge Centre for Alternative Finance (CBECI / Digital Mining Industry Report 2025) and Digiconomist (Bitcoin Energy Consumption Index) figures in MasterBrain, against a dated network snapshot — not real-time or metered data. The two methodologies diverge substantially (Cambridge is the academically conservative, survey-based estimate; Digiconomist is a higher economic-model estimate); the Range view shows the full published band and is the most honest treatment of the uncertainty. The boundary is direct mining electricity only — it excludes embodied mining hardware (ASIC manufacturing), e-waste, cooling water, data-centre overhead beyond IT load, and grid transmission losses. Per-transaction figures are a downstream allocation, not a marginal impact: mining electricity is largely independent of transaction count, and Cambridge explicitly declines to publish a per-transaction headline. Holdings use proportional network-share allocation (CCRI 2021) against a dated circulating-supply constant. For your own mining operation, the Mining electricity basis with your measured grid factor is the most defensible. This is a transparent estimate; reconcile against your own records and, where material, proceed to third-party verification.
You have seen the number: one Bitcoin transaction equals the carbon of a transatlantic flight, or running a fridge for a year, or some other vivid analogy. It travels well because it sounds precise. It is also one of the most misleading figures in climate reporting — not because Bitcoin’s footprint is small, but because the “per-transaction” framing measures something that does not physically exist.
This calculator replaces the viral number with the two figures professionals actually use — Cambridge and Digiconomist — shows you where they disagree, and lets you account holdings, mining, or transactions with an honest uncertainty band on every result.
The Bitcoin network consumes roughly 170 TWh a year and emits about 39.8 Mt CO2e on the Cambridge estimate. There is no single per-transaction footprint: mining energy is almost independent of transaction count, so per-transaction figures are allocations, not measurements.
Two live models
Cambridge CBECI/CDMIR (bottom-up) and Digiconomist BECI (economic). Switchable; Range plots the band between them.
Three accounting bases
Holdings (share of network) · Mining electricity (kWh × intensity) · Per-transaction (flagged allocation).
Grid basis
Cambridge’s own mining-mix weighting — 0.288 kgCO₂e/kWh. Not a GreenCalculus grid read. Digiconomist static 0.558.
Scope mapping
Holder → Scope 3 / financed (Cat 15). Miner → Scope 2. Consumer of mined service → Scope 3.
Boundary
Use-phase electricity only. Hardware, e-waste, cooling water, T&D excluded. Not a life-cycle assessment.
Uncertainty
Every metric carries a band. Network: 32.9–114.03 Mt CO₂e/yr. Bands are the point, not a footnote.
Bitcoin’s energy is consumed by mining — the competitive computation that secures the network — and that energy is almost entirely independent of how many transactions ride in each block. Whether the network processes 200,000 transactions a day or 600,000, the miners run at the same intensity, chasing the same block reward. So dividing total network energy by transaction count produces a number that moves when transaction volume changes but reflects no real causal link. Cambridge declines to publish a per-transaction headline for exactly this reason. The calculator offers the basis because people search for it — but flags it, and steers you toward holdings or mining electricity, which map to something physical.
See the Cambridge CBECI Basis →
What “Bitcoin Emissions” Covers — Network, Not Transaction
Bitcoin emissions are the greenhouse gases from the electricity that powers Bitcoin mining — the global fleet of specialised machines competing to add blocks to the chain. The footprint is a property of the network as a whole, not of any single payment, and accounting it honestly means starting from network energy and allocating down, rather than starting from a transaction and multiplying up.
The Energy → Emissions Chain
- Hashrate and hardware. Miners run application-specific machines (ASICs) at a combined hashrate near 796 EH/s. Hardware efficiency sets how much power that computation draws.
- Network electricity. Power draw across the fleet, annualised, gives network consumption — about 170 TWh/yr on the Cambridge live model.
- Grid mix. That electricity is converted to CO₂e using the mix of grids where mining actually happens — not a single global average, but a weighted blend that shifts as miners relocate.
Why “Per-Transaction” Is a Derived, Not Physical, Unit
A transaction does not cause mining energy. Mining runs continuously to win block rewards, and a block carries whatever transactions are queued — from a handful to several thousand. Dividing network energy by transaction count therefore yields a figure that changes with block fullness and with the rise of inscriptions like Ordinals, without any change in the underlying energy. It is an allocation, useful only with that caveat attached, which is why this calculator de-emphasises it.
Scope Mapping — Where Bitcoin Lands in Your Inventory
The right scope depends on your relationship to the network. A company holding Bitcoin on its balance sheet reports the associated emissions as Scope 3 Category 15 (investments) — the financed-emissions logic, the same family the PCAF standard governs for other asset classes. A company mining Bitcoin reports the electricity it draws as Scope 2, like any other operational power consumption. A company paying for a mined service treats it as a purchased Scope 3 service. This calculator covers all three bases and labels the scope for each.
Included vs Excluded
| Included in this calculator | Excluded — account separately or not in boundary |
|---|---|
| Mining use-phase electricity (network consumption, annualised) | Embodied carbon of ASIC mining hardware manufacture |
| Network CO₂e via Cambridge mining-mix weighting (or Digiconomist) | Mining e-waste — roughly 2.3 kt/yr from fleet decommissioning |
| Holdings allocation (share of circulating supply × network CO₂e) | Cooling water and data-centre overhead beyond IT load |
| Mining-electricity basis (your MWh × mining-mix intensity) | Transmission and distribution losses |
| Per-transaction allocation (flagged, both models) | Ethereum and other proof-of-stake chains (different profile entirely) |
The Two Models — Cambridge CBECI vs Digiconomist
Almost every Bitcoin-energy figure in circulation traces to one of two sources. They are built on opposite logics, they disagree by a wide margin, and understanding why is the difference between quoting a number and understanding it. The calculator carries both live and lets you switch.
Cambridge CBECI — Bottom-Up, Hardware-Weighted
The Cambridge Bitcoin Electricity Consumption Index builds consumption from the ground up: it models the mix of mining hardware in service, weights each by efficiency, and derives a plausible range of power draw for the observed hashrate. Its emissions companion, the Cambridge Digital Mining Industry Report (CDMIR), applies a mining-map grid weighting to convert that power into CO₂e. It is the figure most academic and policy work uses, and it deliberately reports a wide range rather than a single point — the live best-guess is 170.42 TWh/yr consumption and 39.8 Mt CO₂e/yr emissions.
Digiconomist BECI — Economic, Top-Down
The Digiconomist Bitcoin Energy Consumption Index works backwards from money: it assumes miners spend a predictable share of their revenue on electricity, then converts that spend into energy at an assumed price. Because miner revenue is high, this economic approach tends to produce higher consumption and higher emissions than Cambridge, and it applies a higher static grid intensity (0.558 kgCO₂e/kWh, anchored to a post-China-migration 2021 baseline). It is the source behind most of the dramatic per-transaction headlines.
Cambridge CBECI / CDMIR
Bottom-up from hardware efficiency × hashrate, mining-map grid weighting. Network: 170.42 TWh/yr · 39.8 Mt CO₂e/yr · 0.288 kgCO₂e/kWh effective. Wide published range. Source: CCAF, 2025.
Digiconomist BECI
Top-down from miner revenue → electricity spend → energy. Higher consumption, static 0.558 kgCO₂e/kWh. Per-transaction headline 488 kg/tx. Source: Digiconomist (prose citation; no standards page).
For corporate reporting and disclosure, the Cambridge estimate is the more widely accepted and conservative choice, and it is the calculator’s default. Digiconomist is useful as a high-end sensitivity bound — if your figure holds up under Digiconomist’s harsher assumptions, it is robust. The Range mode plots both so you can report the central estimate with the spread visible, which is the most defensible presentation for an auditor or a disclosure reviewer.
One subtlety the calculator preserves and that careless reporting destroys: Cambridge’s consumption headline and its emissions headline sit on different terawatt-hour bases. The 170.42 TWh is the live consumption model; the 39.8 Mt is derived from a survey-based ~138 TWh figure. Dividing the emissions by the consumption to “check” the intensity gives a wrong answer, because the two were never meant to share a denominator. Cite each as published.
From Electricity to Emissions — The Mining-Mix Grid Problem
Two operations consuming identical electricity can have very different footprints, because Bitcoin’s carbon depends entirely on where the machines run. This is the single largest source of legitimate disagreement between estimates, and it is why a terawatt-hour figure alone tells you nothing about emissions.
Location Is Everything
Mining is mobile: machines follow cheap power, which means the global mining mix shifts as jurisdictions open, ban, or subsidise. After China’s 2021 mining ban, hashrate migrated heavily to the United States, Kazakhstan, and Russia, changing the blended grid intensity overnight. Cambridge’s approach captures this by weighting per-country and per-province grid mixes against the observed distribution of mining; Digiconomist applies a single static factor anchored to a post-ban snapshot. Neither is wrong — they answer slightly different questions — but they diverge precisely because the mix is contested.
| Grid approach | What it represents | Effective intensity |
|---|---|---|
| Cambridge mining-mix (default) | Per-country/province grid mix × life-cycle factors, weighted to mining distribution | 0.288 kgCO₂e/kWh |
| Digiconomist static | Single factor anchored to the post-China-migration 2021 baseline | 0.558 kgCO₂e/kWh |
| Sustainable share (context) | Proportion of mining electricity from low-carbon sources | 52.4% (renewables 42.6% + nuclear 9.8%) |
The sustainable-electricity share is worth dwelling on. On the Cambridge mix, 52.4% of mining electricity is low-carbon — 42.6% renewables (hydro 23.4%, wind 15.4%, solar 3.2%, other 0.5%) plus 9.8% nuclear — with the fossil balance of 47.6% dominated by gas (38.2%) over coal (8.9%). That a majority of mining runs on low-carbon power is why the effective intensity (0.288) sits well below a coal-heavy grid, and why the location of mining matters more than its quantity.
Bitcoin Network Emissions — Annual and Per-Transaction
Two framings dominate Bitcoin-emissions reporting: the annual network total (defensible) and the per-transaction figure (popular but flawed). The calculator surfaces both, with very different weight.
Annualised Network Footprint
On the Cambridge estimate the network draws 170.42 TWh/yr and emits 39.8 Mt CO₂e/yr, with a published uncertainty band running from 32.9 to 114.03 Mt — itself a measure of how contested the figure is. The consumption alone is comparable to a mid-sized industrialised country. This is the number to anchor on for any network-scale claim, and the calculator’s context strip displays it on every result regardless of which basis you choose.
Per-Transaction — Why the Band Is Enormous
The live per-transaction headline is 488.07 (AR6) kgCO₂e [GreenCalculus cryptocurrency.bitcoin.tx_emission · DIGICONOMIST BTC · v2026.203] on the Digiconomist basis — but the honest range runs from roughly 110 kgCO₂e (Cambridge-implied, dividing 39.8 Mt across some 360 million annual transactions including Ordinals-driven volume) to over 700 kgCO₂e at Digiconomist’s historical peak. A spread that wide across the same year is the clearest possible signal that the metric is an artefact of the denominator, not a stable property of a payment.
| Per-transaction basis | Figure | What drives it |
|---|---|---|
| Cambridge-implied (low) | ~110 kgCO₂e/tx | 39.8 Mt ÷ ~360M annual tx (incl. Ordinals) |
| Digiconomist (headline) | 488.07 kgCO₂e/tx | Higher consumption ÷ on-chain tx count |
| Digiconomist peak (high) | ~712 kgCO₂e/tx | 2025 historical peak |
The calculator shows a non-dismissible banner whenever you select the per-transaction basis: it is a downstream allocation, not a marginal impact. Adding one transaction to a block adds almost no energy; the block would be mined regardless. Layer-2 systems like Lightning batch thousands of payments under a single on-chain settlement, which collapses any per-transaction figure further. If you must report per-transaction, report it as an allocation with the model and band named — never as a measured cost of a payment.
Proof-of-Work vs Proof-of-Stake — Why Bitcoin Is the Outlier
Bitcoin’s footprint is a consequence of one design choice: proof-of-work, which secures the network through energy-intensive competitive computation. It is not representative of “crypto” as a whole, and conflating the two is a common reporting error.
The Ethereum Contrast
Ethereum, the second-largest network, used proof-of-work until September 2022, when “the Merge” switched it to proof-of-stake — which secures the network through staked capital rather than computation. The change cut Ethereum’s energy use by roughly 99.95%. A single network, the same year, moved from a Bitcoin-like profile to a negligible one by changing its consensus mechanism. This is why a Bitcoin figure must never be generalised to other assets.
Bitcoin — proof-of-work
Security from competitive mining. ~170 TWh/yr, ~39.8 Mt CO₂e/yr. Energy is the design, not a side effect. This calculator’s sole scope.
Ethereum & PoS chains
Security from staked capital. Post-Merge Ethereum cut energy ~99.95%. Footprint orders of magnitude lower — modelled differently, not here.
For asset selection and financed-emissions reporting this distinction is decisive: a portfolio’s crypto carbon is dominated by its proof-of-work exposure, and a holding of a proof-of-stake asset carries a fraction of the footprint per dollar. This calculator is Bitcoin (proof-of-work) only, and the result panel says so directly so a reader never extrapolates a Bitcoin number across a mixed portfolio. A separate proof-of-stake basis is feasible — MasterBrain already carries Ethereum and other proof-of-stake chains — but is out of scope for this version.
Worked Examples — Holdings and Mining, Step by Step
Two examples reproduce the calculator’s exact output for the two bases a corporate user is most likely to need: a treasury holder accounting Bitcoin on the balance sheet, and a miner accounting the electricity it draws. The first is the calculator’s default boot state.
Example 1 — Holding 1 BTC (the default boot state)
This is exactly what the calculator shows on load: one Bitcoin held, Cambridge model, financed-emissions framing. The holding’s footprint is its share of the network’s annual emissions.
One Bitcoin is 1 ÷ 19,870,000 of circulating supply — a share of 5.033×10⁻⁸. Applied to the Cambridge network total of 39.8 Mt CO₂e/yr, the holding’s annual footprint is 2.00 tCO₂e, with a band of 1.66–5.74 tCO₂e (the Cambridge central estimate sits at 2.00, the Digiconomist high at 5.74). The network context strip shows alongside: 170.42 TWh/yr consumption, 39.8 Mt/yr emissions, 52.4% sustainable electricity, 796 EH/s hashrate.
| Step | Calculation basis | Result |
|---|---|---|
| Supply share | 1 BTC ÷ 19,870,000 circulating | 5.033×10⁻⁸ |
| Network driver | Cambridge network_annual_co2e | 39.8 Mt CO₂e/yr |
| Allocation | 39.8 Mt × 5.033×10⁻⁸ (× 10⁶ t/Mt) | 2.00 tCO₂e |
| Band | low 32.9 Mt → 1.66 t · high 114.03 Mt → 5.74 t | 1.66–5.74 tCO₂e |
Audit trail note: All carbon values are live MasterBrain reads (cryptocurrency.bitcoin.network_annual_co2e, MB v2026.13, snapshot 2026-05-21, source CCAF_CBECI_2025). The single hardcoded input is circulating supply (19,870,000) — a dated constant, advanced-editable in the tool and disclosed as such. The 2.00 tCO₂e figure is the engine’s default-boot output recorded on the review date; scope is Scope 3 Category 15 (financed emissions, the holdings basis).
Example 2 — A Small Mining Operation Drawing 1 GWh/yr
This example uses the mining-electricity basis, which carries the kilowatt-hour step a holder’s calculation skips. A modest operation consuming 1 GWh (1,000 MWh) per year, on the Cambridge mining-mix intensity:
| Step | Calculation basis | Result |
|---|---|---|
| Electricity drawn | 1 GWh = 1,000,000 kWh | 1,000,000 kWh |
| Mining-mix intensity | Cambridge mining_mix_carbon_intensity | 0.288 kgCO₂e/kWh |
| Operational emissions | 1,000,000 × 0.288 ÷ 1000 | 288 tCO₂e |
| Band | Cambridge 0.288 → 288 t · Digiconomist 0.558 → 558 t | 238–558 tCO₂e |
Audit trail note: Mining-mix intensity is a live MasterBrain read (cryptocurrency.bitcoin.mining_mix_carbon_intensity, value 0.288, band 0.238–0.558, source CCAF_CBECI_2025; Digiconomist high bound 0.558). Bar widths are scaled to the Digiconomist high (558 t) and reconcile with the table. Scope is Scope 2 (location-based) for the miner’s own electricity. The result scales linearly — a 10 GWh/yr farm is 2,880 tCO₂e on the same basis.
Standards and Methodology — How These Numbers Are Grounded
The calculator’s credibility rests on two published indices and a clear scope mapping. The references fall into three groups.
Accounting Standard
The footprint is reported under the GHG Protocol Corporate Standard and, for holdings, the Scope 3 Standard as Category 15 investments — the financed-emissions treatment the PCAF standard governs. Mining electricity is Scope 2. The consumer of a mined service reports Scope 3. The calculator labels the scope for whichever basis you select.
Network Energy and Emissions Basis
Consumption and emissions follow the Cambridge Bitcoin Electricity Consumption Index and its mining-industry report (CCAF, 2025) by default, with the Digiconomist Bitcoin Energy Consumption Index available as a higher-end alternative model — cited in prose, as it has no dedicated standards page. For an independent cross-check, the Crypto Carbon Ratings Institute publishes peer-reviewed network-emissions ratings using a comparable bottom-up method. The full derivation, including the bottom-up hardware-efficiency build, the mining-map grid weighting, and the two-terawatt-hour-base subtlety, is documented in the Bitcoin emissions methodology page.
Boundary and Grid Basis
The carbon conversion uses Cambridge’s mining-mix weighting rather than a GreenCalculus grid read, so it already reflects where mining occurs. The boundary is use-phase electricity only; hardware manufacture, e-waste, cooling water, and transmission losses are excluded, consistent with Cambridge’s statement that the index is not a life-cycle assessment.
Relationship to the Other Digital Models
Bitcoin is the outlier among the energy-intensive digital models. The AI compute calculator models the other headline high-power workload, the data-centre PUE calculator supplies the facility-overhead multiplier that mining and hosting share, and the cloud compute calculator frames the general instance-allocation case. All rest on the same grid-intensity lineage, so the estimates stay mutually consistent. See also the end-user devices calculator and the IT asset and e-waste calculator for adjacent digital workloads.
Accounting for Bitcoin in a Corporate Inventory
The viral per-transaction figure is useless for corporate reporting; what a sustainability or finance team actually needs is a defensible treatment for one of three relationships to Bitcoin. The calculator’s three bases map directly to them.
Holder / treasury exposure
Bitcoin on the balance sheet is Scope 3 Category 15 — financed emissions. Allocate your share of network CO₂e by share of circulating supply. The default basis.
Miner / operational
Electricity drawn to mine is Scope 2, location-based. Use the mining-electricity basis with your actual MWh — far more accurate than a network allocation.
Disclosure exposure
Material crypto holdings may need disclosure under CSRD ESRS E1, IFRS S2, or the SEC climate rules. Report the model, band, and basis — not a single number.
The recurring mistake is to treat a Bitcoin holding as carbon-free because “the network would run anyway.” Under financed-emissions accounting, that argument fails for the same reason it fails for an equity holding in a fossil producer: attribution is by ownership share, not by marginal causation. If you hold the asset, you account a share of its emissions. The calculator’s holdings basis implements exactly this, and links the result to the scope and category an auditor will expect.
Audit Checklist — Common Crypto-Emissions Reporting Errors
The errors below are the recurring sources of overstated, understated, or unverifiable Bitcoin figures. The calculator guards against most; viral statistics guard against none.
- Quoting a per-transaction figure as a measured cost. The most common and most damaging error. Per-transaction is an allocation that moves with block fullness, not a marginal impact. Report network totals, or holdings/mining bases, with the model named.
- Naming no model. “Bitcoin emits X” is meaningless without Cambridge or Digiconomist attached — they differ by nearly 2× on intensity. Always state the source and the band.
- Generalising Bitcoin to “crypto”. Proof-of-stake chains carry a fraction of the footprint. A Bitcoin figure applied to a mixed portfolio overstates everything but the proof-of-work share.
- Back-deriving an intensity from mismatched bases. Cambridge’s consumption (170 TWh) and emissions (39.8 Mt, on ~138 TWh) sit on different denominators. Dividing them manufactures a figure neither source endorses.
- Ignoring the grid mix. Identical electricity carries different carbon depending on where mining runs. A single static global factor misses the migration that drives real change.
- Treating holdings as carbon-free. Financed-emissions accounting attributes by ownership share, not marginal causation. A balance-sheet holding carries a footprint.
- Quoting a point value with no band. The network estimate spans 32.9–114.03 Mt CO₂e/yr. A single figure without a band overstates the available precision.
- Claiming a life-cycle figure. These numbers are use-phase electricity only. Hardware, e-waste, and cooling are excluded — do not present the result as a full LCA.
Data Sources, Factor Provenance, and Uncertainty
Factor Provenance — Live MasterBrain
Every figure on this page is a live MasterBrain v2026.203 read from the cryptocurrency.bitcoin.* keyspace, snapshot 2026-05-21:
- Network consumption and emissions — 170.42 TWh/yr and 39.8 Mt CO₂e/yr, from the Cambridge CBECI and CDMIR 2025 (source id CCAF_CBECI_2025).
- Mining-mix intensity — 0.288 kgCO₂e/kWh (Cambridge), with the Digiconomist 0.558 as the high bound.
- Per-transaction and per-kWh — 488.07 kgCO₂e/tx and 875.05 kWh/tx (source id DIGICONOMIST_BTC).
- Hashrate and sustainable share — 796 EH/s and 52.4%, shown on the context strip.
The one value that is not a live read is circulating supply (~19.87 million), a dated constant that the tool exposes as advanced-editable and that the disclaimer flags. Grid factors are deliberately not read here — the carbon conversion is internal to the Cambridge mining-mix weighting.
Uncertainty
| Input | Nature of uncertainty |
|---|---|
| Network emissions | Published band 32.9–114.03 Mt CO₂e/yr. Model choice (Cambridge vs Digiconomist) is the largest driver. |
| Network consumption | Band 82.70–315.45 TWh/yr. Hardware-mix assumptions dominate. |
| Mining-mix intensity | 0.238–0.558 kgCO₂e/kWh. Depends on where mining runs — shifts with migration. |
| Per-transaction | 110–712 kgCO₂e/tx. An allocation artefact of the denominator, not a stable value. |
| Circulating supply | Dated constant (~19.87M); rises slowly with issuance until the 21M cap. |
| Sustainable share | 52.4% central; survey-based and revised periodically. |
Update Schedule
Bitcoin factors refresh with each MasterBrain release as Cambridge and Digiconomist publish new figures. The data version badge in the calculator footer always reflects the live MasterBrain version in use.
Frequently Asked Questions
About 39.8 Mt CO₂e a year on the Cambridge estimate, from roughly 170 TWh of electricity — comparable to a mid-sized country. The published band runs from 32.9 to 114.03 Mt, reflecting genuine disagreement between models and grid-mix assumptions. The figure depends heavily on which source you use (Cambridge is lower and more widely accepted; Digiconomist is higher) and on where mining is concentrated. The calculator shows both models and the band between them.
There is no honest single answer, because mining energy is almost independent of transaction count. Per-transaction figures range from about 110 to over 700 kgCO₂e depending on the model and the year, which tells you the metric is an allocation, not a measured cost. Adding a transaction to a block adds almost no energy — the block is mined regardless. Cambridge declines to publish a per-transaction headline for this reason. Account holdings or mining electricity instead, which map to something physical.
They use opposite methods. Cambridge builds energy bottom-up from mining hardware efficiency and hashrate, then weights emissions by where mining actually occurs. Digiconomist works top-down from miner revenue, assuming a fixed share is spent on electricity. The economic approach tends to produce higher figures, and Digiconomist also applies a higher static grid intensity (0.558 vs 0.288 kgCO₂e/kWh). Both are legitimate; Cambridge is the more conservative and widely cited, which is why the calculator defaults to it.
Partly. On the Cambridge mining mix, about 52.4% of mining electricity is low-carbon — roughly 42.6% renewables (hydro, wind, solar) plus 9.8% nuclear — with the remaining 47.6% fossil, mostly natural gas. That majority-low-carbon mix is why the effective intensity (0.288 kgCO₂e/kWh) sits below a coal-heavy grid. The share shifts as mining migrates between jurisdictions, which is the single biggest lever on Bitcoin’s real-world footprint.
As Scope 3 Category 15 — financed emissions. Your holding’s footprint is its share of the network’s annual emissions: your coins divided by circulating supply, multiplied by network CO₂e. One BTC works out to about 2.00 tCO₂e a year on the Cambridge estimate, with a band of 1.66–5.74 t. This is the calculator’s default basis. The “the network would run anyway” argument does not exempt a holding, because financed-emissions accounting attributes by ownership share, not marginal causation.
No longer. Ethereum used energy-intensive proof-of-work until September 2022, when it switched to proof-of-stake — cutting its energy use by roughly 99.95%. Bitcoin remains proof-of-work, so its footprint is orders of magnitude larger. This is why a Bitcoin figure must never be generalised to “crypto”: proof-of-stake chains carry a tiny fraction of the footprint. This calculator models Bitcoin only.
It depends on your relationship to the network. Holding Bitcoin is Scope 3 Category 15 (financed emissions). Mining Bitcoin is Scope 2 — the electricity you draw, like any operational power. Paying for a mined service is Scope 3 as a purchased service. The calculator labels the scope for whichever basis you select, so the figure maps cleanly into a GHG Protocol inventory.
Because the honest uncertainty is wide. The network estimate alone spans 32.9 to 114.03 Mt CO₂e/yr across models and assumptions, and the two leading sources disagree by nearly 2× on grid intensity. A single point value would imply a precision that does not exist. The calculator reports a central estimate with the Cambridge and Digiconomist bounds plotted, so you can present the spread an auditor will expect rather than a false-precision figure.
Methodology Notes and Limitations
Two live models, switchable. Cambridge CBECI/CDMIR (bottom-up) is the default; Digiconomist BECI (economic) is the high-end alternative. Every figure is a live MasterBrain read of both models’ values and bounds; Range plots the band between them.
Network, not transaction. Bitcoin’s energy is a property of the mining network and is nearly independent of transaction count. Per-transaction figures are downstream allocations, flagged as such and de-emphasised.
Mining-mix carbon, not a grid read. The carbon conversion uses Cambridge’s per-country/province mining-map weighting (0.288 kgCO₂e/kWh), not a GreenCalculus grid factor. The figure already reflects where mining runs and shifts as mining migrates.
Two terawatt-hour bases, never blended. Consumption (170.42 TWh, live model) and emissions (39.8 Mt, ~138 TWh survey base) sit on different denominators by construction. Each is cited as published; the calculator does not back-derive an intensity from them.
Use-phase boundary. Electricity only. Hardware manufacture, e-waste (~2.3 kt/yr), cooling water, data-centre overhead, and transmission losses are excluded. This is not a life-cycle assessment, consistent with Cambridge’s own statement.
Bitcoin proof-of-work only. Ethereum and proof-of-stake chains carry a far lower profile and are not modelled here. Circulating supply (~19.87M) is the single hardcoded, advanced-editable input.