Avoided Emissions Calculator — “Scope 4” / Carbon Handprint
Compute the signed net avoided emissions of a product or solution against a reference (counterfactual) scenario, stage by stage, under the WBCSD Avoided Emissions Guidance and WRI comparative-impact method — with rebound, attribution, and additionality adjustments, live grid and fuel factors on derived stages, and an in-tool claim-robustness meter.
What the engine computes. Avoided emissions — the “carbon handprint”, informally “Scope 4” — are the difference in life-cycle emissions between a world with your solution and a reference (counterfactual) world without it. The calculator is a life-cycle stage register: you enter one or more stages, and each stage carries a reference value (what the counterfactual would emit) and a solution value (what your offering emits). Gross avoided emissions are the sum of (reference − solution) across all stages. The result is then adjusted downward for rebound, attribution, and additionality to produce a signed net figure.
Two ways to enter each stage. Every stage row accepts either direct entry (reference tCO2e and solution tCO2e as numbers) or derive-from-energy (an electricity-grid or fuel quantity for the reference side, the solution side, or both, converted at a live MasterBrain factor: tCO2e = quantity × factor ÷ 1,000). Derived rows are the only place live values enter the calculation — grid factors read grid.<iso3>.electricity.location_based and fuels read fuels.<iso3>.<fuel>.<unit>. Derived rows are marked with a ⚡ symbol in the stage breakdown so the audit trail distinguishes them from directly-entered figures.
The reference scenario is yours, not the tool’s. There is deliberately no avoided-emissions or baseline factor family in the MasterBrain. A reference scenario is an epistemic choice — a claim about what would have happened otherwise — and the calculator never fabricates one. It prompts for a free-text reference justification, captures it into the audit trail and export, and scores the claim’s robustness partly on whether that justification is present. Choosing a defensible reference is the single most consequential input, and it rests entirely with you.
Three handprint adjustments. Gross avoided emissions overstate the real-world climate benefit unless three discounts are applied: rebound (the fraction of the saving eroded because the solution makes the service cheaper or easier to consume — a value above 100% is a “backfire” where the intervention increases emissions), attribution (the share of the saving fairly claimable by this actor rather than the whole value chain), and additionality (the share of the saving that would not have occurred anyway). Each defaults conservatively (rebound 0%, attribution and additionality 100%) and flows through a handprint waterfall: Reference → −Solution → −Rebound → −Attribution → −Additionality → Net.
GWP basis is declarative. The GWP-basis selector (AR6 GWP-100 default, or AR5 GWP-100) records and cites which basis the reported tonnes rest on; it does not re-weight numbers you have already entered in tCO2e. If your reference and solution figures were computed on different bases, reconcile them before entry — the calculator will not silently blend them.
Firewalled from your inventory — enforced, not advised. Avoided emissions never enter a Scope 1, 2, or 3 inventory and are never subtracted from a footprint. This is a WBCSD rule and it is enforced in the tool: a “do not net against your inventory” notice fires as an always-on insight and in the scope/boundary panel, and the JSON export carries nettable_against_inventory: false. The handprint sits beside the footprint; it never nets into it.
Avoided emissions (“Scope 4” / carbon handprint) are a comparative claim: the emissions society avoids because your solution is used instead of a reference, over the same function and life-cycle boundary. Quantify each life-cycle stage’s emissions under the reference scenario and under your solution in tCO₂e (with the GreenCalculus grid, fuels, LCA or material calculators, or your own data), then enter the tonnes below.
Reference = emissions of what would otherwise have been used (the counterfactual). Solution = your offering’s emissions for the same stage. Avoided = reference − solution, per stage. Enter tCO₂e directly, or set a row’s Input to “Derive from energy” to compute each side from an energy quantity × a live MasterBrain grid/fuel factor (auto-cited). Include the solution’s own embodied/production burden — counting only its use-phase saving overstates the handprint.
Handprint adjustments — rebound, attribution, additionality ⓘ
Enter your reference and solution emissions by life-cycle stage to quantify avoided emissions
Results appear instantly. A net-avoided headline, a reference→solution→net handprint waterfall, a stage-by-stage breakdown, a claim-robustness check, and the full audit trail appear after calculation. Avoided emissions are a comparative claim — reported separately from your Scope 1/2/3 inventory, never subtracted from it.
This is a comparative (“Scope 4” / avoided-emissions / carbon-handprint) estimate — the difference between a reference (counterfactual) scenario and your solution over the same functional unit and life-cycle boundary, aligned with the WBCSD Guidance on Avoided Emissions (2023) and the WRI/GHG Protocol comparative-impacts guidance (2019). It is computed from your own stage-level emissions figures. GreenCalculus does not select, supply or validate your reference scenario — the counterfactual is a justified assumption that governs the result and is your responsibility to document. Avoided emissions are reported separately from the corporate GHG inventory and must never be subtracted from Scope 1, 2 or 3. The output is not a verified GHG assertion (independent assurance is a separate exercise) and not carbon credits or offsets. All emissions are user-supplied; the result is only as robust as the reference assumption, boundary symmetry, and the rebound, attribution and additionality adjustments you apply. GWP values are read from MasterBrain at the declared basis and recorded for transparency, not applied as a multiplier to figures you have already CO₂e-weighted.
Avoided emissions are the most useful and the most abused number in corporate climate accounting. Useful, because a low-carbon product genuinely changes what the world emits, and a company that sells one deserves a way to describe that contribution. Abused, because the same number is routinely bolted onto a footprint, netted against actual emissions, and waved as evidence of a “net-negative” company — none of which it can support. The difference between the two is almost entirely a matter of discipline: an honest reference scenario, conservative adjustments, and a hard firewall against netting.
This calculator is built around that discipline. It treats avoided emissions as a comparison between your solution and a reference (counterfactual) scenario you must justify, computes the saving stage by stage, discounts it for rebound, attribution, and additionality, pressure-tests the claim against four robustness checks, and refuses to let the result net into your inventory. The number it produces is a handprint that sits beside your footprint — never inside it.
Avoided emissions (“Scope 4”) are the difference between what a reference scenario would have emitted and what your solution emits, summed across life-cycle stages and adjusted for rebound, attribution, and additionality. They are reported alongside your footprint and are never subtracted from it.
What avoided emissions (“Scope 4”) actually are
Avoided emissions are the emissions that do not happen because a product, service, or project exists — measured against a reference (counterfactual) scenario in which it does not. A high-efficiency motor that replaces a standard one, a video-conferencing service that displaces a flight, a renewable plant that displaces grid electricity: each avoids emissions that the reference world would have produced. The concept is also called the carbon handprint — the positive counterpart to the carbon footprint — and it is informally nicknamed “Scope 4”.
Why “Scope 4” is a nickname, not a GHG Protocol scope
The GHG Protocol defines exactly three scopes: Scope 1 (direct emissions), Scope 2 (purchased energy), and Scope 3 (value-chain emissions). All three are footprint categories — they count emissions a company causes. Avoided emissions are structurally different: they count emissions a company’s product helps a third party not cause, relative to a counterfactual. There is no fourth scope in the GHG Protocol, and the WBCSD guidance is explicit that avoided emissions are a distinct accounting category reported separately, not an extension of the scopes. “Scope 4” is a convenient label that has stuck; it is not a standardised scope, and treating it as one — adding it to a scopes total, netting it against Scope 1–3 — is the central error the discipline exists to prevent.
Handprint beside footprint — reference versus solution
The footprint answers “how much do we emit?” The handprint answers a different question: “how much less does the world emit because our solution exists?” Those are two separate ledgers. A company can have a large footprint and a large handprint at the same time — a wind-turbine manufacturer emits real Scope 1–3 emissions building turbines (footprint) while those turbines displace fossil generation (handprint). Both are true, both are disclosable, and neither cancels the other. The whole method rests on comparing two clearly-specified worlds: the reference (counterfactual) scenario and the solution scenario.
Throughout this page and in the calculator UI, the two worlds being compared are called the reference (counterfactual) scenario and the solution scenario. “Baseline” is a common synonym for the reference scenario; the WBCSD guidance and this tool use “reference” to avoid confusion with the emissions baseline year used in target-setting, which is an unrelated concept. Everywhere below, reference = the counterfactual world without your solution.
Avoided vs offset vs reduction vs removal
Four terms are constantly conflated, and the conflation is where most avoided-emissions greenwash begins. They occupy different places in the accounting and can claim different things. Getting the distinction right is the foundation for every number this calculator produces.
Avoided emissions (handprint)
Emissions a third party does not emit because your solution displaced a higher-carbon reference. Counterfactual by nature. Reported outside the scopes, beside the footprint. Cannot be netted against your own emissions, cannot be sold as a credit, cannot count toward a reduction target.
Reduction
An actual decrease in your own Scope 1, 2, or 3 emissions versus your own prior period. This is what science-based targets measure. It changes your footprint; avoided emissions do not. A science-based target is met by reductions, never by handprint.
Removal
Physically taking CO2 out of the atmosphere and durably storing it — afforestation, direct air capture, enhanced weathering. Removals change the atmospheric stock. Avoided emissions never do: they are emissions that were prevented, not carbon that was extracted.
Offset (credit)
A tradeable instrument representing a tonne reduced or removed elsewhere, bought to compensate for your own emissions. Offsets carry integrity criteria (ICVCM, VCMI) and are counted against a footprint. Avoided emissions are neither tradeable nor a credit — see the Carbon Offset Cost Calculator for that separate workflow.
The footprint counts what you emit. The handprint counts what the world avoids because of you. They are two ledgers, reported side by side — and a tonne on one is never a tonne off the other.
The single most common claim error is treating avoided emissions as interchangeable with any of the other three. “We avoided 50,000 tCO2e, so we’re net-negative” silently converts a counterfactual handprint into a footprint reduction — which it can never be. Avoided emissions do not reduce your Scope 1–3, do not remove atmospheric carbon, and are not credits you can retire. Keep the four concepts in separate columns.
How the calculator works — the stage register
The calculator is not a single reference-number-minus-solution-number box. It is a life-cycle stage register: you build up the comparison one stage at a time, and each stage independently carries a reference value and a solution value. This mirrors how a real life-cycle comparison is structured — production, distribution, use phase, maintenance, and end-of-life can each shift the balance in a different direction, and a claim that looks strong on one stage can collapse once another is included.
Reference and solution scenarios per stage
Every stage row asks for two numbers: what the reference (counterfactual) scenario emits at that stage, and what the solution scenario emits at the same stage. Gross avoided emissions are the sum of (reference − solution) across all stages. Stage types follow the standard life-cycle vocabulary — production embodied (A1–A3), distribution (A4), use-phase, maintenance, end-of-life (C), and other — so the comparison stays legible to anyone reading a life-cycle assessment. Because each stage is explicit, the tool can see when a claim rests entirely on one favourable stage while ignoring an unfavourable one.
Direct entry versus derive-from-energy
Each stage is entered one of two ways:
- Direct. You enter reference tCO2e and solution tCO2e as numbers — appropriate when you already have life-cycle figures from an LCA, an EPD, or a supplier disclosure.
- Derive from energy. You enter an energy quantity for the reference side, the solution side, or both, and the tool converts it at a live MasterBrain factor: tCO2e = quantity × factor ÷ 1,000. Electricity reads a grid factor by country (
grid.<iso3>.electricity.location_based); fuels read a fuel factor by fuel, region, and unit (fuels.<iso3>.<fuel>.<unit>). Derived rows are the only place live values enter the calculation, and they are marked with ⚡ in the stage breakdown.
A common and correct pattern is to derive the reference side (the grid electricity a solar system displaces, say) and enter the solution side directly (the system’s embodied carbon from its EPD). The tool handles asymmetric stages like this natively — a stage can have a derived reference and a direct solution, or a reference value and a zero solution value, or vice versa.
The three handprint adjustments — the waterfall
Gross avoided emissions are an upper bound. Three adjustments bring the figure down to a defensible net claim, and the tool shows them as a waterfall so the erosion from gross to net is visible:
| Adjustment | Range / default | What it captures |
|---|---|---|
| Rebound | ≥0%, default 0% | The share of the saving eroded because the solution makes a service cheaper or easier, increasing consumption. Above 100% is a “backfire” — the intervention raises emissions overall, and the tool flips the result to a net-burden mode. |
| Attribution | 0–100%, default 100% | The share of the saving fairly claimable by this actor rather than spread across the whole value chain. Lower it when several parties legitimately contribute to the same avoided tonne. |
| Additionality | 0–100%, default 100% | The share of the saving that would not have happened anyway. Lower it where regulation or market trends would have driven part of the switch regardless of your solution. |
The waterfall runs Reference → −Solution → −Rebound → −Attribution → −Additionality → Net. The hero output is signed: a positive net figure renders as “Avoided”, a negative one as “Net burden”, each with the GWP basis on the mode badge. The GWP-basis selector itself is declarative — it records and cites AR6 or AR5 GWP-100 but does not re-weight tonnes you have already entered.
Choosing a defensible reference scenario
The reference scenario is where an avoided-emissions claim lives or dies. It is a statement about what would have happened otherwise — and because the counterfactual world is not observable, it is always a judgement. The calculator deliberately does not choose it for you: there is no reference or baseline factor family in the MasterBrain, because fabricating a “standard” counterfactual would manufacture a false sense of objectivity. The tool structures and pressure-tests the claim; you own the reference.
Anchor the whole comparison on a single, clearly-stated functional unit — the equivalence anchor both scenarios must deliver (WRI comparative-impact method). “One tonne of cement”, “one megawatt-hour delivered”, “the system over its 25-year life”. If the reference and solution do not deliver the same functional unit, the comparison is invalid before any factor is applied. The tool captures the functional unit as a required field for exactly this reason.
Common reference-scenario types (educational — not a tool setting)
The four reference types below are how practitioners think about constructing a counterfactual. They are not options the calculator enforces or labels — the engine takes whatever reference figures you enter and scores whether you have justified them. Use this as a framework for building a defensible reference, then enter the resulting numbers.
| Reference type | What the counterfactual assumes | Defensibility notes |
|---|---|---|
| Market-average | The average product or service the customer would otherwise have bought. | Defensible for mass-market substitutes; requires a credible market-average dataset and a stated geography and year. |
| Technology / like-for-like | The specific incumbent technology the solution replaces (standard motor, gas boiler, incandescent bulb). | Strong when the substitution is genuine and one-for-one; weak if the customer would not otherwise have bought the incumbent at all. |
| Regulatory / minimum-standard | The least-efficient option still legally permitted — the regulatory floor. | Conservative and audit-friendly, because it avoids over-crediting against an already-improving market. Often the most defensible choice. |
| Performance-standard | A benchmark level of performance (e.g. sector-average intensity) rather than a specific product. | Useful where no single incumbent exists; the benchmark’s source and vintage must be disclosed. |
Whichever construction you use, the reference is a claim you must be able to defend to an assurer. The calculator captures your free-text reference justification into the audit trail and export, and factors its presence into the robustness score — the mechanism the next section describes. For a fully project-level treatment of counterfactual construction, monitoring, and verification, the ISO 14064-2 Project Calculator is the sibling tool built around the same reference-and-solution logic applied to GHG projects.
How the tool pressure-tests your claim
Most avoided-emissions tools produce a number and stop. This one produces a number and then interrogates it, because an unexamined handprint claim is a liability, not an asset. Two mechanisms do the work: an anti-cherry-pick guard and a four-check claim-robustness meter.
The anti-cherry-pick guard
The classic avoided-emissions cherry-pick is to claim a large use-phase saving while quietly omitting the solution’s production burden. A solar panel that displaces grid electricity for 25 years looks spectacular if you count only the use phase and ignore the 8 tonnes of embodied carbon in manufacturing it. The tool watches for exactly this shape: a use-phase saving entered with no solution-side embodied burden anywhere in the register triggers a guard warning, prompting you to account for production before the claim is treated as complete. A symmetric, whole-life comparison is the only kind the tool will score as robust.
The guard enforces life-cycle symmetry: if you claim a benefit at one stage, you must account for the costs at the others. This is not a stylistic preference — the WBCSD guidance and the WRI comparative-impact method both require the reference and solution to be compared on the same, complete life-cycle boundary. The guard turns that requirement into an on-screen prompt rather than a footnote an assurer discovers later.
The claim-robustness meter (four checks)
The tool grades every claim on four checks and reports one of four tiers — Weakly, Partially, Well-supported, or Robust. The tiers are a self-assessment of methodological completeness, not a verification, but they map directly onto what an assurer will ask for:
Reference justified
A free-text reference (counterfactual) justification is present. An unjustified reference is the fastest way to lose an avoided-emissions claim under review — this check makes its absence visible immediately.
Life-cycle symmetry
The solution’s production or embodied burden is accounted for, not just its use-phase benefit. This is the anti-cherry-pick guard expressed as a robustness dimension.
Adjustments considered
Rebound, attribution, and additionality have been engaged with rather than left at silent defaults where they plainly matter. Conservative discounting strengthens the claim.
Boundary declared
The functional unit and scenario boundary are specified, so the comparison is reproducible. A claim without a declared boundary cannot be independently checked.
A claim that passes all four renders as “Robust” (4/4); one that passes none renders as “Weakly supported”. The meter, the guard, the full audit trail, and a calculation hash are all carried into the JSON/CSV export — so the rigour is portable, and a reviewer can reconstruct exactly how the number was produced.
Worked example — rooftop solar displacing grid
This example reproduces exactly against the live calculator (MasterBrain v2026.38). It shows both input modes on one claim: a derived reference stage and a direct solution stage, on a symmetric whole-life boundary. Figures below are the tool’s rendered strings, which apply house rounding — the underlying precise values differ slightly and are noted where relevant.
Solution: a 10 kW rooftop solar PV system. Functional unit: “the system over its 25-year life”. Scenario basis: absolute totals. GWP basis: AR6 GWP-100.
| Stage | Type | Input mode | Reference (counterfactual) | Solution |
|---|---|---|---|---|
| 1 — Production (embodied) | Production A1–A3 | Direct | 0 tCO2e | 8 tCO2e |
| 2 — Use-phase ⚡ | Use-phase | Derive from energy | Electricity, UK grid, 250,000 kWh → 32.7 tCO2e | None (0 tCO2e) |
The reference use-phase stage derives from 250,000 kWh of UK grid electricity — the electricity the system’s output displaces over its life — at the live grid factor grid.gbr.electricity.location_based = 0.13096 kg CO2e/kWh (DEFRA 2026, MB v2026.38). That yields a reference use-phase figure the tool renders as 32.7 tCO2e (precise 32.74). The solution emits nothing in the use phase — a running PV system has no fuel-side emissions — but carries 8 tCO2e of embodied production carbon, entered directly. Adjustments are left at their conservative defaults: rebound 0%, attribution 100%, additionality 100%.
The handprint waterfall
| Waterfall step | Value |
|---|---|
| Reference total | 32.7 tCO2e |
| − Solution total | 8 tCO2e |
| = Gross avoided | +24.7 tCO2e |
| − Rebound (0%) | no change |
| − Attribution (100%) | no change |
| − Additionality (100%) | no change |
| = Net avoided | +24.7 tCO2e |
Because all three adjustments sit at their defaults on this claim, gross and net avoided coincide at +24.7 tCO2e (precise 24.74). The tool reports the reduction versus the reference as 75.6% (precise 75.565%) — the solution delivers the same functional unit while emitting roughly a quarter of the reference’s life-cycle carbon.
This claim scores 4/4 precisely because it is symmetric: the use-phase benefit is stated alongside the production burden, so the anti-cherry-pick guard is satisfied. Had the 8 tCO2e production stage been omitted, gross avoided would have read a flattering +32.7 tCO2e — but the guard would have fired and the robustness meter would have dropped, flagging the claim as incomplete rather than letting the inflated number stand.
One further discipline is worth naming: the +24.7 tCO2e is a handprint. It does not reduce the installer’s or manufacturer’s own Scope 1–3 footprint, and it is not subtracted from anything. The 8 tCO2e of embodied carbon is a real footprint entry for whoever manufactured the system; the 24.7 tCO2e avoided is a separate, parallel disclosure. The next section explains why the tool enforces that separation rather than merely recommending it.
The firewall — why avoided emissions never enter your inventory
The most important rule in avoided-emissions accounting is also the most frequently broken: avoided emissions are reported beside your footprint and are never netted into it. The calculator does not merely advise this — it enforces it structurally, which is a deliberate integrity feature rather than a disclaimer.
What “firewalled” means in the tool
- A “do not net against your inventory” notice fires as an always-on insight on every result, regardless of inputs.
- The scope/boundary panel states in plain terms what the number is (a counterfactual handprint) and what it is not (a Scope 1, 2, or 3 figure, an offset, or a reduction).
- The JSON export carries the machine-readable flag
nettable_against_inventory: false, so any downstream system that ingests the result is told explicitly that the figure must not be summed into a footprint total.
In effect, the tool refuses to let the number enter an inventory. You cannot export a result that presents itself as nettable, because the export asserts the opposite.
Why netting is prohibited
The prohibition is not bureaucratic caution — it follows from what the two numbers are:
- Avoided emissions are counterfactual. They describe emissions that would have occurred in a scenario that did not happen. Nothing was removed from the atmosphere; a higher-carbon path was simply not taken. The atmospheric stock is unchanged by the claim itself.
- Footprint emissions are factual. Scope 1–3 tonnes were actually emitted. Subtracting a counterfactual handprint from a factual footprint produces a number that means nothing — it mixes “what we emitted” with “what someone else didn’t”.
- Netting manufactures false net-negativity. A company with a large handprint could show a negative “net” footprint and claim to be removing carbon from the atmosphere. It is not. The handprint is a contribution claim about products sold; the footprint is an accountability claim about emissions caused. Collapsing them destroys both.
In the worked example, the solution carries 8 tCO2e of embodied carbon and avoids 24.7 tCO2e. It is tempting to report “net −16.7 tCO2e” as though the product were carbon-negative. That figure is meaningless: the 8 tonnes are a footprint entry for the manufacturer, and the 24.7 tonnes are a counterfactual saving accruing to the electricity system. They live in different ledgers. Report them separately — 8 tCO2e footprint, 24.7 tCO2e avoided — never as a single netted number.
Common mistakes and how assurance flags them
The findings below are the ones reviewers raise most often on avoided-emissions claims. Each maps to a specific guard or check in the tool — the calculator is built to surface them before an assurer does.
01 — Netting the handprint into the footprint
Subtracting avoided emissions from Scope 1–3 to claim net-negativity. Prohibited because avoided emissions are counterfactual and footprint emissions are factual. The tool’s always-on firewall notice and nettable_against_inventory: false export flag exist to stop this at source.
02 — Use-phase saving with no production burden
Counting a solution’s operational benefit while omitting its embodied carbon. The anti-cherry-pick guard fires when a use-phase saving appears with no solution-side embodied burden anywhere in the register.
03 — An unjustified reference scenario
Asserting a counterfactual without stating what it is or why it is credible. The robustness meter’s “reference justified” check drops immediately when the free-text justification is absent.
04 — Ignoring rebound
Claiming the full gross saving when the solution makes a service cheaper and consumption rises. Leaving rebound at 0% where it plainly applies overstates the benefit; a rebound above 100% is a backfire the tool reports as a net burden.
05 — Over-attribution
Claiming 100% of a saving that several value-chain actors legitimately contributed to. The attribution input exists to apportion the tonne fairly; leaving it at 100% by default where multiple parties contribute inflates the claim.
06 — Mismatched functional units
Comparing a reference and solution that do not deliver the same service. The required functional-unit field forces the equivalence anchor to be declared; a mismatch invalidates the comparison before any factor is applied.
07 — Blended GWP bases
Combining reference and solution figures computed on AR5 and AR6 without reconciliation. The GWP-basis selector is declarative and cites the basis, but it does not re-weight your tonnes — reconcile before entry.
08 — Treating the claim as verified
Presenting a 4/4 robustness score as third-party assurance. The meter is a self-assessment of methodological completeness, not a verification. Export the audit trail and calculation hash and have a qualified practitioner review it before external use.
How avoided emissions fit standards and disclosure
Avoided emissions sit outside the scopes, which means they are governed by a different set of references than a footprint. The frameworks below cover the methodology and disclosure surface a company making a handprint claim will navigate.
| Framework | Role for avoided emissions | Status |
|---|---|---|
| WBCSD Avoided Emissions Guidance (2023) | The primary methodology reference. Defines avoided emissions as a distinct category reported outside the scopes, sets the reference-and-solution comparison, and requires the no-netting firewall. | Voluntary guidance; the de facto standard for corporate handprint claims |
| WRI — Comparative Emissions Impacts of Products (2019) | The comparative-impact method underlying the functional-unit and reference-scenario discipline. Establishes the equivalence-anchor requirement the tool enforces. | Voluntary method; widely cited alongside the WBCSD guidance |
| ISO 14064-2 (project accounting) | The project-level GHG accounting standard for quantifying reductions or removals against a reference. The formal, verifiable counterpart when the claim is a project rather than a product. | International standard; verifiable under ISO 14064-3 |
| SBTi Corporate Net-Zero Standard | Explicitly excludes avoided emissions from target accounting. Targets are met by reductions in your own footprint; the handprint is reported separately and never counts toward the target. | Voluntary target framework; avoided emissions are out of scope for target-setting by design |
| EU Green Claims Directive / CSRD ESRS | Governs how a handprint claim may be communicated. Avoided-emissions marketing claims must be substantiated, must not imply a footprint reduction, and must disclose the reference scenario. | EU regulatory regime; substantiation and anti-greenwashing requirements apply to public claims |
The through-line across all five is that avoided emissions are a contribution claim, held to a substantiation standard, and rigorously separated from footprint accounting and target-setting. Where the claim is a defined project rather than a product line, the ISO 14064-2 Project Calculator applies the same reference-and-solution logic in a verifiable project frame. For the downstream reduction and abatement work that avoided emissions never substitute for, the Net-Zero Pathway Calculator and Marginal Abatement Cost Calculator operate on the footprint side of the ledger.
Data sources, factor versioning, and update transparency
What is live, and what is yours
Only two kinds of value on this page are live MasterBrain reads: the grid and fuel factors used on derived stages, and the GWP constants that the declarative basis selector cites. Everything on the reference side of a comparison — the counterfactual itself — is user-asserted, because the tool deliberately maintains no avoided-emissions or reference factor family. A counterfactual is a claim, not a lookup, and the calculator will not manufacture one.
| Value | Source | How it enters |
|---|---|---|
| Electricity grid factors (derived stages) | DEFRA 2026 (GB), EPA eGRID (US subregions), plus other countries as published | Live: grid.<iso3>.electricity.location_based, location-based |
| Fuel factors (derived stages) | DEFRA 2026 (GB) and regional fuel datasets | Live: fuels.<iso3>.<fuel>.<unit> |
| GWP constants | IPCC AR6 (default) / AR5 GWP-100 | Live: gwp.<gas>.<basis>, declarative — cited, not re-weighted |
| Reference (counterfactual) figures | You — from LCA, EPD, market data, or a stated assumption | User-asserted; no factor family exists by design |
The grid and fuel pickers auto-discover whatever countries and fuels are present in the MasterBrain, so newly-published geographies appear without any change to this page. Each derived stage stamps its factor, source, and basis into the audit trail via the citation block, and the result footer records the MasterBrain version the computation ran against — so a figure produced today and the same inputs re-run against a later factor vintage remain distinguishable in any restatement.
GWP-basis discipline
UK DEFRA factors carry AR5 GWP-100 internally by DEFRA’s own convention, while the tool’s default reporting basis is AR6. The selector cites which basis a result rests on but never blends bases inside one total. If your reference and solution figures were computed on different bases, reconcile them to a single basis before entry — the global warming potential and CO2e glossary entries cover the reconciliation in detail.
Frequently asked questions
No. The GHG Protocol defines exactly three scopes — Scope 1 (direct), Scope 2 (purchased energy), and Scope 3 (value chain), all of which count emissions a company causes. “Scope 4” is an informal nickname for avoided emissions, also called the carbon handprint. The WBCSD Avoided Emissions Guidance treats them as a distinct accounting category reported separately from and outside the scopes, not as a fourth scope. Adding avoided emissions to a scopes total, or netting them against Scope 1–3, is the central error the discipline exists to prevent.
No. The SBTi Corporate Net-Zero Standard explicitly excludes avoided emissions from target accounting. Science-based targets are met by reducing your own Scope 1, 2, and 3 emissions — actual decreases in your footprint. Avoided emissions are a counterfactual contribution claim about what your products help others not emit; they do not reduce your footprint, so they cannot count toward a reduction target. Report the handprint separately, as a contribution disclosure, and pursue your target through actual reductions.
An offset is a tradeable credit representing a tonne reduced or removed elsewhere, purchased to compensate for your own emissions and counted against your footprint under integrity criteria like ICVCM and VCMI. Avoided emissions are neither tradeable nor a credit — they are a counterfactual measure of emissions a third party did not produce because your solution displaced a higher-carbon reference. You cannot buy, sell, or retire avoided emissions, and they are reported outside the scopes rather than against a footprint. The Carbon Offset Cost Calculator covers the offset workflow separately.
No, and it structurally refuses to. A “do not net against your inventory” notice fires as an always-on insight on every result, the scope panel states what the number is and is not, and the JSON export carries nettable_against_inventory: false so no downstream system can treat the figure as a footprint deduction. Avoided emissions are counterfactual and footprint emissions are factual; subtracting one from the other produces a meaningless number and a false claim of net-negativity. The handprint is reported beside the footprint, never inside it.
The reference scenario is the world without your solution — what would have been emitted otherwise. It is the anchor of every avoided-emissions claim. The tool does not choose or supply it: there is deliberately no reference or baseline factor family in the MasterBrain, because a counterfactual is a judgement, not a lookup, and fabricating a “standard” one would create false objectivity. You construct and justify the reference; the tool captures your justification, scores the claim’s robustness on whether it is present, and pressure-tests the comparison. Choosing a defensible reference is the most consequential input and rests entirely with you.
They mean the same counterfactual scenario, but this tool and the WBCSD guidance use “reference” to avoid confusion with the emissions baseline year used in target-setting, which is a different concept. When you see “baseline” in avoided-emissions material, read it as the reference (counterfactual) scenario — the world without your solution. The calculator labels the two compared worlds “reference” and “solution” consistently for this reason.
Gross avoided emissions overstate the real benefit unless three discounts are applied. Rebound captures the saving eroded when a solution makes a service cheaper and consumption rises — above 100% it is a “backfire” that increases emissions. Attribution apportions the saving fairly when several value-chain actors contribute to the same avoided tonne. Additionality removes the share of the saving that would have happened anyway. Each defaults conservatively (rebound 0%, attribution and additionality 100%), but leaving them at defaults where they plainly apply inflates the claim and weakens it under review.
It catches the most common avoided-emissions inflation: claiming a use-phase saving while omitting the solution’s production burden. If you enter a use-phase saving with no solution-side embodied carbon anywhere in the stage register, the guard warns you to account for production before treating the claim as complete. The WBCSD guidance and the WRI comparative method both require the reference and solution to be compared on the same, complete life-cycle boundary — the guard turns that requirement into an on-screen prompt rather than a finding an assurer discovers later.
No. The robustness meter is a self-assessment of methodological completeness across four checks — reference justified, life-cycle symmetric, adjustments considered, and boundary declared. It maps onto what an assurer will ask for, but it is not third-party verification. A 4/4 claim is well-structured; it still needs review by a qualified practitioner before external use. Export the audit trail and calculation hash from the tool so a reviewer can reconstruct exactly how the number was produced.
Only the grid and fuel factors on derived stages, and the GWP constants, are live MasterBrain reads — grid factors from DEFRA 2026 and EPA eGRID, fuels from DEFRA and regional datasets, GWP from IPCC AR6 or AR5. Everything on the reference (counterfactual) side is user-asserted, because the tool maintains no avoided-emissions or reference factor family by design. When you derive a stage from energy, the factor, its source, and its GWP basis are stamped into the audit trail; when you assert a reference figure, its justification is captured instead. The result footer records the MasterBrain version the computation ran against.
Both use the same reference-and-solution logic, but at different granularity. This calculator is built for product- and solution-level handprint claims reported under the WBCSD guidance — a product line, a service, a technology substitution. The ISO 14064-2 Project Calculator applies the same comparison to a defined GHG project with formal quantification, monitoring, and verification requirements under an international standard verifiable via ISO 14064-3. Use this tool for a product or solution contribution claim; use the ISO 14064-2 calculator when the claim is a discrete, verifiable project.
Methodology notes and limitations
Methodology basis. The calculator implements the reference-and-solution comparison from the WBCSD Avoided Emissions Guidance (2023) and the WRI “Comparative Emissions Impacts of Products” method (2019), with rebound, attribution, and additionality adjustments and an in-tool robustness assessment. Avoided emissions are a distinct accounting category reported outside the GHG Protocol scopes.
The reference scenario is user-asserted. The tool does not choose, supply, or classify a reference scenario, and maintains no avoided-emissions or baseline factor family. It captures a free-text reference justification, scores the claim on whether one is present, and warns on the classic use-phase-without-production cherry-pick. Defensibility of the counterfactual rests entirely with the user.
Live values are limited to derived stages and GWP constants. Only grid factors (grid.<iso3>.electricity.location_based), fuel factors (fuels.<iso3>.<fuel>.<unit>), and GWP constants (gwp.<gas>.<basis>) are read live. The GWP-basis selector is declarative — it records and cites the basis but does not re-weight entered tonnes. Reconcile mixed AR5/AR6 figures before entry.
The firewall is enforced, not advisory. Avoided emissions are never netted against a Scope 1, 2, or 3 inventory. The tool fires an always-on no-netting insight, states the boundary in the scope panel, and carries nettable_against_inventory: false in the JSON export. Results are a handprint reported beside the footprint.
Rounding. Displayed figures use the house rounding convention (values ≥1000 to 0dp, ≥10 to 1dp, ≥1 to 2dp, smaller values to more places; percentages to 1dp). The worked example shows rendered strings; underlying precise values differ slightly (e.g. 32.7 renders 32.74; 24.7 renders 24.74; 75.6% renders 75.565%).
No FX and no unit blending. Where a scenario basis multiplies a per-unit figure by a number of units, the per-unit figures must share a consistent basis and boundary. The calculator does not convert currencies or reconcile inconsistent functional units — the required functional-unit field exists to force a single equivalence anchor across both scenarios.
No assurance opinion. Results are estimates and a self-assessment of methodological completeness, not a verification. The robustness meter does not constitute assurance. Export the per-claim audit trail (stage register, factor sources, GWP basis, adjustments, reference justification, calculation hash) and have a qualified practitioner review it before use in regulated disclosures or public claims. For the project-level, verifiable counterpart, see the ISO 14064-2 Project Calculator; for real reference-versus-solution comparisons already modelled on-site, see the Renewable vs Grid Carbon Payback Calculator and the Reusable vs Single-Use Calculator.