Employee Commuting and Homeworking Emissions

This page covers Scope 3 Category 7 — employee commuting and the homeworking emissions that the GHG Protocol Corporate Value Chain (Scope 3) Standard folds into the same category. It covers why the category splits into two structurally different calculations, the distance-based, fuel-based and average-data methods the standard recognises, the DEFRA 2026 per-kilometre commuting factors and the per-FTE-working-hour homeworking factor, the data-collection and survey design that determines the data-quality grade, and four worked examples that reconcile against their stated inputs. Commuting shares its factor basis with the business travel methodology (DEFRA Category 6), and both ultimately roll into the company’s Scope 3 total under the GHG Protocol Corporate Standard.

Why Category 7 Is Two Methodologies, Not One

The GHG Protocol places employee commuting and teleworking in a single category — Category 7 — but the two activities have nothing in common at the calculation layer. Commuting is a transport problem: a distance, a mode, and a per-kilometre emission factor. Homeworking is an energy problem: hours worked at home, multiplied by a factor expressing the heating and equipment energy that working from home draws. The activity data is different (kilometres versus working hours), the unit is different (kg CO₂e per km versus kg CO₂e per FTE working hour), and the data source is different (a commuting survey versus a headcount-and-hours model). They share a category number and an inventory line, and nothing else.

Reporting them as one calculation — most often by trying to force homeworking into a distance frame, or by quietly dropping it because it has no distance — is the single most common Category 7 error. The two halves are computed independently, each in its own unit, and summed only at the end into one Category 7 total.

Where the boundary sits

Commuting is travel between an employee’s home and their normal place of work, in vehicles not owned or controlled by the company — if the company owned the vehicle, the journey would sit in Scope 1. Homeworking covers the energy consumed when an employee works from home instead of commuting to an office: heating, cooling, and office-equipment electricity for the hours worked at home. Business travel — flights, hotels, hire cars for work trips — is Category 6, not Category 7, and is handled in the business travel methodology. The well-to-tank upstream fuel emissions of commuting fuel do not belong in Category 3, which covers only fuel and energy the reporting company itself buys and consumes. DEFRA’s commuting factors are tank-to-wheel by default, so add the WTT variant only if you choose to report those upstream emissions within Category 7.

The remote-work shift

Before 2020, homeworking was a rounding error in most inventories and many companies omitted it. The shift to hybrid and remote work made it material: a workforce that works two or three days a week from home moves a meaningful share of its energy footprint out of the commute and into domestic heating. The GHG Protocol does not mandate homeworking accounting, but CSRD ESRS E1 and investor-grade disclosure now expect a credible treatment of it where the workforce is materially remote. A homeworking figure of zero for a hybrid workforce is now a flag, not a simplification.

The Calculation Methods

The GHG Protocol recognises three methods for the commuting half, in descending order of data quality. Homeworking uses its own hours-based method, treated separately in §4. The commuting methods are not interchangeable in accuracy — the standard’s data hierarchy ranks reported, distance-and-mode-specific data above modelled averages.

Distance-based — the default

The distance-based method multiplies the distance each employee travels by mode by that mode’s per-passenger-kilometre factor, then sums across employees and modes. It is the GHG Protocol’s preferred commuting method because it uses real, employee-specific activity data: who travels how far, by what mode, how many days. The accuracy of the result is governed almost entirely by the quality of the commuting survey behind it, which is why §5 treats survey design as a first-class part of the methodology rather than an operational afterthought.

Fuel-based — accurate but rarely feasible

If an employer somehow knows the fuel each employee burned commuting — for example through a fuel-card scheme tied to private vehicles — the fuel-based method applies a per-litre fuel factor directly and is highly accurate. In practice almost no employer holds employee commuting fuel data, so this method is a theoretical best case rather than a working default. Where partial fuel data exists for a subset of employees, it can anchor the distance-based estimate for the rest.

Average-data — the fallback

When no survey data is available, the average-data method models commuting from headcount, an assumed average commute distance, an assumed number of working days, and a modal split drawn from national transport statistics. It is the lowest-quality method and should be a documented fallback, not a convenience. A company that uses average-data for a workforce it could survey understates its achievable data quality and produces a figure no assurance reviewer can trace to source.

Method-selection logic

Commuting Emission Factors (DEFRA 2026)

GreenCalculus applies the DEFRA 2026 conversion factors for commuting, surfaced live from the MasterBrain. Car factors are per vehicle-kilometre; bus and rail factors are per passenger-kilometre — a distinction that matters when an employee shares a car versus rides a train, and one the calculation must respect. All values below render live and update with each DEFRA release.

The per-vehicle versus per-passenger distinction

The single most consequential factor-selection decision in commuting is the difference between a car factor (per vehicle-km) and a public-transport factor (per passenger-km). A car factor already represents the whole vehicle: two employees car-pooling 20 km generate one vehicle-journey of emissions between them, not two. A rail factor is already divided down to one passenger, so two employees on the same train each carry their own passenger-km. Applying a per-vehicle car factor as though it were per-passenger double-counts car-poolers; applying a per-passenger rail factor per vehicle would be meaningless. The orders of magnitude make the point: a diesel car sits near a fifth of a kilogram per kilometre, while rail modes run roughly five to eleven times lower.

The DEFRA car-size factors (small, medium, large, average) carry the same values as the Scope 1 passenger-vehicle factors used in the business travel methodology; the difference is the category tag, not the number. Where survey responses give vehicle size, use the size-specific factor; where they give only “I drive”, use the average. Default to the average row only when size is genuinely unknown.

Homeworking — The Hours-Based Method

Homeworking does not have a distance, so it cannot use any of the commuting methods. DEFRA publishes a per-FTE-working-hour factor that expresses the additional energy — heating and office equipment — drawn when an employee works from home. The calculation is headcount-and-hours based: full-time-equivalent hours worked at home, multiplied by the per-hour factor.

What the factor covers — and what it does not

The combined factor bundles two components: heating, which dominates the figure, and office-equipment electricity, which is comparatively small. The heating component is the reason homeworking is seasonal and climate-dependent: it represents the marginal domestic heating attributable to a person being home and working rather than out. The factor does not cover cooling (DEFRA’s basis is a temperate-climate heating assumption), nor does it cover the employee’s full domestic energy — only the portion attributable to the working hours. Lighting is treated within the office-equipment component rather than separately.

The UK-only limitation

The DEFRA homeworking factor is derived from UK housing stock, UK heating patterns, and the UK grid, and the MasterBrain holds it under the gbr key only. There is no equivalent published factor for other countries in the current dataset. For a UK workforce it is directly applicable. For a non-UK or mixed-geography workforce, applying the UK factor unadjusted is a known approximation: a workforce in a hot climate that air-conditions rather than heats, or one on a much cleaner or dirtier grid, will be mis-stated. The defensible treatments are to apply the UK factor and disclose it as a UK-derived proxy, or to build a country-specific estimate from local grid factors and heating assumptions and document the derivation.

Data Collection and Survey Design

Because Category 7 activity data lives with employees rather than in company systems, the survey is the methodology. A distance-based commuting figure is only as good as the survey that produced its distances, modes, occupancies and working patterns. The same is true of homeworking, where the binding inputs are days worked from home and contracted hours.

Survey versus modelling

A direct survey asks employees for their commute distance, primary mode, car occupancy where relevant, and the number of days per week they work in the office versus at home. Modelling substitutes assumptions — average distance, national modal split — for some or all of those answers. The survey supports a distance-based (highest-quality) calculation; pure modelling forces the average-data method. Most real inventories sit between the two: a survey of a representative sample, extrapolated to the full headcount.

Sampling and extrapolation

Where surveying every employee is impractical, a representative sample is surveyed and the result scaled to total headcount. The extrapolation is defensible only if the sample is representative across the dimensions that drive emissions — site location, role pattern, and home/office split. Extrapolating a head-office sample to a distributed field workforce, or a five-days-in-office sample to a hybrid population, imports a bias the headline number then carries silently. Document the sample size, response rate, and the basis for believing the sample represents the whole.

The data-quality ladder

This ladder mirrors the GHG Protocol’s data-quality guidance: reported, specific data outranks modelled averages. Track the basis explicitly so the figure’s quality can be disclosed and improved over reporting cycles — moving a workforce from average-data to a surveyed distance-based figure is the highest-value data-improvement step in Category 7, and usually changes the number, not just its confidence.

Worked Examples — Audit Records

Four audit-record snapshots. The values are fixed: they reconcile to their stated inputs and must continue to reconcile regardless of future DEFRA releases. The DEFRA factors used are the 2026-06-09 review-date values (matching the live MasterBrain on that date); a future factor revision will change the live tables above but must not change these worked examples. Figures are illustrative and do not represent any specific named employer.

Example A — one employee, distance-based car commute

The car factor is per vehicle-km, so this figure is the whole vehicle’s emissions. If a colleague car-pools the same route, the 1,305.23 kg is split between the two employees (652.62 kg each), not counted twice. Switching the same commute to national rail at 0.03092 kg CO₂e/passenger-km would yield 7,560 × 0.03092 = 233.76 kg — roughly a fifth of the car figure, and counted per passenger.

Example B — average-data fallback for an unsurveyed team

Every input above is an assumption, which is exactly why this is the lowest-quality method. The modal split, the average distance, and the office-day count are all modelled, and a verifier cannot trace any of them to an employee. The figure is reportable only with those assumptions disclosed and a stated intent to replace it with a survey.

Example C — homeworking for one hybrid employee

The same employee no longer commutes on those 104 days, so their commuting figure must be calculated on office days only — the two halves of Category 7 partition the working year between them. Using the combined factor captures both the heating (0.30234) and office-equipment (0.02159) components; an employer that could separately evidence a heat-free summer period could split them, but the combined factor is the standard treatment.

Example D — blended Category 7 total for the hybrid employee

The two halves are computed in their own units — vehicle-kilometres and FTE working hours — and added only at the end. The 156 office days and 104 home days sum to 260 working days, partitioning one employee’s working year with no overlap and no gap. This partition discipline is what keeps a hybrid workforce’s Category 7 figure free of the double-count that arises when commuting and homeworking are estimated independently against the same calendar.

Edge Cases and Known Pitfalls

Governance, Assurance, and Disclosure

Category 7 is increasingly assured rather than estimated, particularly for companies in scope of CSRD ESRS E1 or reporting under IFRS S2. An assurance reviewer reads the data basis before the number: a surveyed distance-based figure with documented occupancy and a reconciled hybrid split is defensible; a single modelled average for a workforce that could have been surveyed invites challenge. The ISO 14064-1 inventory boundary requires Category 7 to be either quantified or explicitly justified as excluded — silent omission is not a neutral choice.

Run the following pre-flight before submitting a Category 7 figure for assurance:

1. Two halves computed separately. Commuting (per-km) and homeworking (per-FTE-hour) are calculated independently in their own units and summed only at the end.
2. Method documented per population. Each employee group is tagged distance-based, fuel-based, or average-data, with the data basis recorded behind it.
3. Correct factor basis. Car commuting uses per-vehicle-km factors with occupancy captured; bus and rail use per-passenger-km factors.
4. Hybrid split reconciled. Office days plus home days equal contracted working days for every employee group — no overlap, no gap.
5. Homeworking geography disclosed. The UK-derived DEFRA factor is either applied to a UK workforce or disclosed as a proxy / replaced with a country-specific derivation for non-UK staff.
6. Survey representativeness documented. Sample size, response rate, and the basis for extrapolation to full headcount are recorded.
7. Boundary on WTT stated. Whether well-to-tank fuel emissions are included in Category 7 is stated, and they are counted only once.
8. DEFRA factor version recorded. The DEFRA release used is documented so the figure is reproducible against the right edition.
9. Exclusions justified. Any omitted population (e.g. homeworking set to zero) carries an explicit materiality justification, not silence.

What the Calculator Handles vs What You Decide

A calculator automates the factor lookups and the arithmetic across thousands of employees. It cannot design the survey or judge the representativeness of the data behind the inputs.

Standards Alignment

Four standards govern Category 7 and its disclosure consequences.

Factor reference: the per-mode commuting factors, homeworking included, are published with their sources as employee commuting emission factors.

Frequently Asked Questions