Original illustration of employees commuting to a worksite by train, bicycle, walking and shared car, without real emissions data.

An employee commuting inventory should answer a practical question: how much travel did the workforce make between home and work during the reporting year, and how was that travel converted into emissions? Scope 3 Category 7 covers this activity in vehicles outside the reporting company's ownership or operation. Start with journeys, attendance and transport modes. A headcount multiplied by an unexplained annual emissions average hides the decisions that most need checking.

The difficult cases usually come before the calculation. A train to a customer meeting differs from a train to the office. A company-controlled vehicle needs a Scope 1 or Scope 2 boundary check. Hybrid attendance changes the number of commuting trips. Homeworking energy is an optional, separately estimated component; it is not the employee's entire household energy bill.

The approach below turns those distinctions into a usable record: define the population, choose a method for each relevant group, calculate activity, match the factor's units and test the result. Worked examples use invented activity values to explain the arithmetic. They are not workforce benchmarks or recommended emissions factors.

What belongs in Scope 3 Category 7?

Define the reporting population and period before sending a survey. Include the relevant home-to-work journeys of employees within the organisational boundary. If the inventory also includes contractors, consultants or other workers, state that choice and identify their population separately. The GHG Protocol allows these wider inclusions; a supplier's headcount should not quietly become part of the employee count used to scale a survey.

Map the journey as it actually happens. Someone who drives to a station and then takes a train has two motorised legs, with different distances and potentially different factor units. Ask for each leg rather than forcing the person to choose a single “main mode”. Walking between the station and office can remain visible in the travel profile without being assigned an invented motorised-travel factor.

Attendance determines how often the journey occurs. Use the reporting year's pattern, allowing for remote days, leave, part-time schedules, site changes and employees joining or leaving. If the survey asks about a typical week, also ask whether that week represents the whole year. A move halfway through the year may require two commute records.

For example, one employee can have a 12-kilometre journey for six months and a 5-kilometre journey for the remainder. Combining the periods after calculating their actual trip frequencies is clearer than applying the latest address to the entire year. Preserve the period boundaries so another reviewer can reproduce the annual activity.

Route map from home to station by car and station to work by rail, with the return direction shown separately.

One commute, separate legs

Illustrative route: car travel is 4 kilometres one way; rail travel is 18 kilometres one way. Annualise each leg separately using actual attendance, and use the appropriate mode and factor unit. The return trip is counted once.

Separate commuting from business travel and company vehicles

Classify a journey using its purpose and the company's ownership or control of the transport. Ordinary home-to-work travel in a privately owned car belongs in the commuting assessment. Travel undertaken for business, such as visiting a customer, needs a business-travel assessment. Company-owned or controlled transport needs the organisation's Scope 1 or Scope 2 treatment checked before anything is added to Category 7.

Table 1: scroll horizontally to view all columns. Keyboard: focus the table region and use the left and right arrow keys.

Journey Starting classification check Record to retain
Private car from home to regular office Employee commuting Distance, attendance and occupancy
Train from office to customer meeting Business travel Trip purpose and business-travel reference
Company-controlled vehicle Scope 1 or Scope 2 boundary Ownership/control decision and energy record
Mixed commute and business journey Separate the identifiable activities Leg endpoints and allocation rationale

The table is a triage aid, not a substitute for resolving unusual arrangements. A leased shuttle, temporary worksite or journey starting at home and continuing to a customer can require a documented policy decision. Expense reimbursement alone does not determine the emissions category.

Keep disputed journeys in an exception register with an owner and a resolution. Once a decision is made, cross-reference the other inventory category where necessary. This makes double counting detectable: a journey should not be included in both the commuting total and the business-travel total simply because two teams supplied records. When policy changes, explain the classification change separately from any real change in travel.

Decision path: check company control first; then distinguish home-to-work commuting from business travel.

Classify the journey

Company-owned or controlled vehicles require a Scope 1 or Scope 2 boundary decision. For other transport, ordinary home-to-work journeys go to the Category 7 assessment; business trips go to the Category 6 assessment. Unusual arrangements need a documented decision.

Decide whether optional teleworking belongs in this inventory

Decide explicitly whether to include teleworking energy. The Category 7 guidance permits it but does not require a company to treat all household consumption as work-related. The relevant question is the additional energy caused by working at home compared with a baseline that would have occurred anyway.

Consider a home where a refrigerator runs throughout the day whether the employee is present or absent. Its ordinary consumption does not become a work-at-home increment. Equipment used specifically for work may create an increment; heating or cooling needs a more careful baseline because other occupants, weather and operating schedules affect what would have happened without homeworking.

Keep a separate calculation for each included energy source. Record the assumed working hours or days, the incremental consumption method, the factor and the uncertainty. If consumption is estimated from equipment power and operating hours, distinguish that estimate from metered incremental energy. Do not label an assumption as a measurement merely because the spreadsheet produces decimal places.

Report commuting and optional teleworking as separate components before combining them. Fewer journeys do not automatically mean a known net emissions saving: home energy may change, transport factors may differ and the comparison period may be inconsistent. If sufficient home-energy evidence is unavailable, record the inclusion policy and limitation. A clear exclusion or qualified estimate is more useful than a fixed remote-work number whose baseline cannot be explained.

Two schematic bars share the same baseline, with only the extra homeworking segment highlighted.

Count the increment

The common grey baseline represents energy that would have occurred without homeworking. Only the additional green segment is considered for optional teleworking energy. Bar lengths are schematic, not measured consumption.

Collect commuting patterns without pretending every employee replied

Build the survey around the calculation you intend to perform. For each material journey leg, collect mode, one-way or return distance, site attendance, weeks or days covered and region. Include car-sharing information where a vehicle-based factor will need allocation. Make distance conventions explicit in the question: “one-way kilometres from home to work” is less ambiguous than “commute distance”.

Retain both the eligible population and the number of usable responses. Suppose an illustrative site has 200 eligible employees and 80 usable replies. Those replies cover 40% of the population; they do not prove that the remaining 120 employees made no journeys. Nor does the response rate alone establish that the sample represents the whole site.

Compare the respondents with the workforce by characteristics that affect travel, such as site, shift and attendance pattern. Where office staff and shift workers travel differently, assess those groups separately. Scaling one group's average across the other can introduce bias even when the spreadsheet formula is correct. A group with no usable responses needs additional evidence or an explicitly qualified estimate.

Document incomplete answers, exclusions and imputation. For example, retain whether distance came from an employee's estimate, a route calculation or a substitute value. Use only the personal information needed for the calculation, with an appropriate access and retention approach. The final review pack can use aggregated activity and coverage counts; it need not expose individual home addresses to every reviewer.

A grid of 100 dots highlights 40 dots, representing 80 replies out of 200 eligible employees.

Replies are not the population

Illustrative survey coverage: 80 usable responses divided by 200 eligible employees equals 40 percent. Each of the 100 dots represents two people. This response rate does not by itself establish representativeness; compare respondents with workforce groups before extrapolating.

Method 1: use commuting fuel activity when it is known

Use a fuel-based calculation where the relevant commuting fuel can be identified. The basic operation is fuel quantity multiplied by a compatible fuel factor, aggregated across the included activity. Before calculating, check that the record covers employee commuting rather than a mixture of business journeys, household purchases and company-controlled transport.

For an illustrative unit check, 100 litres multiplied by a factor expressed in kilograms of CO2e per litre produces kilograms of CO2e. A payment of 100 currency units is a different input. Fuel expenditure requires a justified conversion to fuel quantity, with an appropriate fuel price and period; a fare or general travel allowance is not itself a fuel quantity. Keep the conversion and its assumptions with the calculation.

Public transport needs particular care. A bus operator's fuel use relates to a service carrying many passengers, not solely to the employee whose commute is being assessed. If operator fuel data are used, retain the allocation method. Do not assign the whole vehicle's fuel to each surveyed passenger.

Prefer the method that the evidence can support. A fuel card with reliable trip-purpose records may provide useful activity. A reimbursement record containing no fuel quantity, price basis or commuting allocation may not. In that case, a distance-based estimate can be more transparent. Record why each population uses its selected method, rather than describing the entire inventory as “fuel-based” when most of its activity is estimated another way.

Method 2: calculate annual distance by transport mode

For a stable commute, calculate annual activity from one-way distance, the return journey and actual attendance. An illustrative employee travels 12 kilometres each way, attends the site three days per week and follows that pattern for 40 weeks:

12 km × 2 journeys/day × 3 days/week × 40 weeks = 2,880 km.

The result is distance, not emissions. Multiply it by a factor with the correct mode and denominator only after checking those units. If the distance supplied was already a daily return distance, omit the extra multiplication by two. If attendance varies materially, calculate the periods separately instead of assuming one schedule throughout the year.

For a mixed-mode commute, keep each leg separate:

Table 2: scroll horizontally to view all columns. Keyboard: focus the table region and use the left and right arrow keys.

Illustrative one-way leg Annualisation Annual activity before factor
Car to station: 4 km 4 × 2 × 3 × 40 960 km
Train to work: 18 km 18 × 2 × 3 × 40 4,320 passenger-km

For the car leg, establish whether the activity and factor are vehicle-based or passenger-based, and how any shared occupancy is handled. Apply the car factor to the car activity and the rail factor to rail passenger activity; do not add the distances and apply one convenient factor to the combined number.

This split also improves review. If a new rail factor changes the result next year, the reviewer can isolate its effect without reconstructing the car journey. Store the activity totals alongside the emissions totals so changes in travel and changes in factors remain distinguishable.

Calculation ladder: 12 kilometres times two journeys, three site days and forty weeks gives 2,880 annual kilometres.

Distance before emissions

Illustrative calculation: 12 kilometres one way multiplied by two for a return journey is 24 kilometres per site day. Three site days make 72 kilometres per week. Forty such weeks make 2,880 kilometres. This is activity, not emissions; apply a compatible mode-specific factor next.

Method 3: estimate from representative average data

When individual fuel or distance data are unavailable, use documented average activity that reasonably represents the workforce. The estimate still needs an employee count, transport-mode distribution, journey distance, attendance and compatible factors. Choose a relevant geography and period. A city-centre office profile is not automatically suitable for a rural plant operating night shifts.

An illustrative calculation makes the assumptions visible. For 100 employees, assume 60% use a particular mode, travel 10 kilometres one way and attend for 200 days. Before applying that mode's factor, the activity is 100 × 0.60 × 10 × 2 × 200 = 240,000 kilometres. Specify whether those kilometres are passenger or vehicle activity; the example alone does not resolve car occupancy.

Repeat for each mode, using mode-specific distances where supported. Check that the population and shares describe the same thing. A share of employees using rail at least once a week is not necessarily the share of all commuting trips made by rail. Mixed-mode journeys and alternating daily choices need a consistent treatment before aggregation.

Keep the proxy source and its limitations. If later survey data replace a national average, explain the methodological change before attributing the resulting emissions difference to employee behaviour. Where uncertainty matters to a decision, compare reasonable alternative activity assumptions and show how sensitive the estimate is. This is a sensitivity check, not evidence that either alternative is the true result.

Check the factor, occupancy and teleworking assumptions

Treat the factor denominator as part of the calculation. A vehicle-kilometre factor and a passenger-kilometre factor are not interchangeable labels. For a shared car using a vehicle-based factor, calculate the vehicle activity and allocate it appropriately. Where a factor is already expressed per passenger-kilometre, do not divide by occupancy a second time.

Consider an illustrative 20-kilometre shared-car journey with two occupants. At vehicle level it is one 20-kilometre journey, not two separate 20-kilometre vehicle journeys. If equal allocation is appropriate, each occupant receives half the vehicle emissions. If both occupants belong to the reporting workforce, their allocated shares add back to the one vehicle total. If only one belongs to the population, the included share differs.

Check more than the denominator. Retain the factor publisher, version, geography, mode, fuel or energy carrier and included emissions stages. Confirm whether the selected factor represents vehicle operation, upstream energy or a wider boundary. A change in factor scope can move the reported result without any change in attendance or transport choice.

Finally, reconcile units and optional components. Convert miles and kilometres deliberately; identify whether reported distance is one-way or return; and use actual commuting days. Keep any included teleworking estimate separate until its baseline and energy units have passed review. Do not net a hypothetical avoided commute against home-energy estimates and describe the difference as measured savings.

One 20 vehicle-kilometre journey splits into two equal occupant shares, which sum to one vehicle total.

One vehicle, two shares

For the illustrative shared-car journey, calculate vehicle emissions once and allocate equally if that allocation is appropriate. Two included occupants each receive half, adding to one vehicle total. Do not divide a passenger-kilometre factor by occupancy again.

A review checklist for a Category 7 inventory

Review a small set of journeys from original input through annual activity to emissions. Include a private-car commute, a shared car, a mixed-mode journey and an employee whose attendance changed. These cases expose different errors; repeating the same simple calculation several times provides less useful assurance.

Table 3: scroll horizontally to view all columns. Keyboard: focus the table region and use the left and right arrow keys.

Check Evidence and decision required
Population Included sites, workers, reporting period and exclusions
Classification Commuting/business-travel distinction and company-vehicle decisions
Coverage Eligible count, usable responses, segments and non-response treatment
Activity Leg distances, direction convention, attendance and units
Factors Publisher, version, geography, denominator and emissions stages
Allocation Shared-vehicle assumptions and avoidance of double allocation
Teleworking Inclusion choice, incremental baseline and uncertainty
Changes Method, source or factor changes separated from activity changes

Reproduce any survey expansion using the retained counts. Confirm that the same employees or trips have not been included in multiple segments. Investigate unusually large changes before rounding them away. A missing factor, unrepresented site or unresolved company-vehicle boundary should remain a visible exception with an owner.

Keep the final package understandable without specialist knowledge of its spreadsheet. It should identify what was reported, what was estimated and what remains uncertain, together with the calculation version and review outcome. Passing this checklist shows that the controls were performed; it does not prove every survey answer is accurate or confer independent assurance on the inventory.

A loop connects journey, activity, factor and result, with review returning from the result to the inputs.

Trace a result back to inputs

Trace a reported result to its factor, annual activity and original journey. When a discrepancy changes an input, resolve it and recalculate while retaining the earlier version. The loop describes a review process, not automatic approval.

Where Carbon Loop fits

If you want to discuss this inventory with Carbon Loop, start with the reporting problem rather than a request for an unexplained total. Bring the organisational boundary, workforce groups, available survey or attendance data and the decisions that remain unresolved. A short description of those inputs makes it easier to establish what support is needed and whether the proposed work is within the available service scope.

For example, distinguish a request to organise evidence from a request to select emissions factors, calculate the inventory or provide independent assurance. Those are different deliverables. Ask who will own the method, what information will be retained and how changes will be documented. Agree the output and its limitations before relying on it for a reporting decision.

Do not assume that a product-carbon workflow automatically includes an operational employee-commuting tool. Confirm the current availability of any Category 7 service or company-emissions application directly. The sustainability team still needs a clear owner for boundary decisions, personal-data handling, unresolved assumptions and final approval. A useful first conversation ends with a defined scope and evidence request, not a promise that software can remove every uncertainty.

Conclusion: keep the journey and method visible

A useful Scope 3 Category 7 inventory connects real or estimated commuting activity to a defensible calculation. Define the population and journey boundary first. Separate business travel and company-controlled transport, then decide whether optional incremental teleworking energy belongs in the assessment.

Choose fuel, distance or average data according to the evidence available for each group. Retain attendance, mixed-mode legs, sample coverage and the factor denominator. Expand the calculation where the workforce differs materially; do not make a single unsupported average carry the whole inventory.

Before approving the result, trace representative journeys through to the total and resolve the exceptions that could change its meaning. Keep activity changes separate from changes in methods or factors. The essential safeguard is reproducibility: another person should be able to follow the population, journey, factor and allocation decisions without guessing what an unexplained cell or missing response meant.