Shipping Carbon Emissions: From Ocean Voyage to Cargo Estimate
A wide synthetic port scene shows a container vessel near cranes, water, foreground containers and a distant industrial harbour under low sunlight. No legible carrier or customer mark, vessel name, booking, cargo amount, fuel type, emission figure, certificate or assurance statement appears. Haze and the depicted equipment do not measure local pollution or establish the route, allocation, fuel performance or emissions of a real shipment.
Diagram text description
A shipping carbon emissions figure is useful only when its reporting unit, fuel boundary and service coverage are clear. A sector inventory, a vessel total and an allocated cargo estimate answer different questions. A business moving goods needs the last of these, supported by evidence that connects the shipment to the transport activity.
Begin by identifying the cargo and reporting period. Declare whether the fuel result covers onboard energy use alone or also fuel production and delivery. Map the service legs, choose a supported calculation method, and allocate shared emissions using a stated denominator. Keep the booking, route, factors, assumptions and exclusions in an updateable evidence pack. Those decisions jointly determine whether a voyage figure can become a cargo estimate.
The same discipline applies to comparisons. A smaller number may describe fewer connections, a narrower gas boundary or a different allocation rule rather than a better shipping option. The practical goal is a result that another reviewer can reconstruct, including the parts that remain unknown.
Know whose number you are looking at
Check whose activity the number describes before using it. A sector total aggregates shipping activity over a defined geography and period. A vessel total may cover a year, while a voyage total concerns a particular sailing or service boundary. None automatically belongs to one consignment.
An operator's fuel records can help quantify vessel emissions. A shipper then needs to establish how much of the relevant service is attributable to its cargo. That requires the appropriate cargo quantity, allocation rule and corresponding total activity, not simply the vessel's advertised capacity.
Also distinguish a total from an intensity. Tonnes of CO2e are an emissions quantity; kilograms of CO2e per tonne-kilometre or TEU-kilometre are intensities with different activity denominators. A twenty-foot equivalent unit, or TEU, is a container-equivalent measure, not a tonne of cargo.
Record the original unit beside the number. If a provider supplies an annual vessel indicator but cannot connect it to the sailing and allocation method, describe it as contextual evidence rather than presenting it as the consignment's footprint.
Declare tank-to-wake or well-to-wake before comparing
Tank-to-wake describes emissions associated with using energy onboard. Well-to-tank concerns the fuel's upstream production and delivery. Well-to-wake combines those stages. The diagram shows containment: fuel supply and onboard use are parts of the wider fuel-life-cycle boundary, not three alternative totals to add together.
One enclosing outline labelled Well-to-wake contains two separate boxes. The first is Well-to-tank: fuel production and delivery. The second is Tank-to-wake: onboard energy use. Containment shows parts of a fuel-life-cycle boundary, not a flow between service legs. The components need compatible gases, activity and accounting before combination. Do not add an upstream component again to a total already reported well-to-wake. The diagram does not identify actual factors, allocate vessel emissions to a consignment, verify a fuel saving or show that terminals, feeders and inland legs have been included.
Diagram text description
Ask which gases the result includes. A combustion CO2 figure is not automatically a complete CO2e result covering the relevant greenhouse gases. Likewise, adding a well-to-tank estimate to a number already reported well-to-wake would count the upstream component twice.
The fuel boundary is also different from the service boundary. A well-to-wake mainline sea estimate does not necessarily include the feeder vessel, terminal handling or inland truck. Conversely, a door-to-door offer can include several transport stages while still using a narrow fuel boundary within each.
Keep these two descriptions together: what happened along the logistics chain, and which fuel-life-cycle emissions are covered for each included element. In the worked example, tank-to-wake and well-to-tank are supplied separately and combined once. That arithmetic is transparent, but it does not make the excluded logistics connections disappear.
Map the complete ocean service
Map the physical service rather than assuming the booking describes one uninterrupted sea leg. Follow the cargo from its origin through inland collection, origin terminal, any feeder vessel, mainline sailing, transfer or transshipment, destination terminal and final inland delivery. A transfer may introduce another vessel, handling operation or storage period.
For each element, record whether its emissions are included, separately estimated, excluded or still unknown. A port-to-port mainline result should be labelled that way even if the commercial invoice says “shipping”. If a carrier total already includes a feeder, adding another feeder estimate without checking the overlap can duplicate it.
Distance belongs to a particular leg and method. Record whether it is an observed sailing distance or a modelled distance and retain the routing assumption. The GHG Protocol transport chapter distinguishes actual and estimated distances; a different methodology may prescribe its own distance basis. Do not mix them silently.
The map becomes a reconciliation tool. If the forwarder includes origin handling but the mainline provider's statement also covers it, identify one consistent treatment. If the destination connection is absent, the next task is to obtain or estimate it, not to rename the available subtotal door-to-door.
Choose an evidence route for the sea leg
Choose the best supported method, not simply the most detailed-looking result. The GHG Protocol transport guidance distinguishes fuel-based, distance-based and spend-based approaches. Each depends on different evidence and none removes the need to check boundaries and allocation.
| Evidence route | Inputs and units to retain | What needs checking |
|---|---|---|
| Carrier fuel or service data | Fuel by type and period, compatible gas factors, or a documented service-emissions total. | Link to the sailing, fuel-life-cycle coverage, shared-cargo allocation and any additional energy or relevant leakage. |
| Distance-based model | Shipment mass or container/volume measure, distance by leg, mode and a compatible intensity. | Vessel/service fit, load assumptions, distance convention and whether the factor is per tonne-kilometre or TEU-kilometre. |
| Spend-based screen | Transport expenditure by service type and a compatible economic factor. | Currency, price year, bundled services and uncertainty; use as a screen when physical activity is unavailable. |
Table 1, shown one row at a time. Every field is retained.
Row 1
- Evidence route
- Carrier fuel or service data
- Inputs and units to retain
- Fuel by type and period, compatible gas factors, or a documented service-emissions total.
- What needs checking
- Link to the sailing, fuel-life-cycle coverage, shared-cargo allocation and any additional energy or relevant leakage.
Row 2
- Evidence route
- Distance-based model
- Inputs and units to retain
- Shipment mass or container/volume measure, distance by leg, mode and a compatible intensity.
- What needs checking
- Vessel/service fit, load assumptions, distance convention and whether the factor is per tonne-kilometre or TEU-kilometre.
Row 3
- Evidence route
- Spend-based screen
- Inputs and units to retain
- Transport expenditure by service type and a compatible economic factor.
- What needs checking
- Currency, price year, bundled services and uncertainty; use as a screen when physical activity is unavailable.
Traceable carrier data is valuable when the underlying activity and allocation are usable. If the total is available but the shared-load evidence is not, a qualified distance estimate may be more defensible than an unsupported allocation.
For a tonne-kilometre model, multiply tonnes by kilometres and by a factor per tonne-kilometre. A TEU-kilometre factor instead requires container-equivalent activity. Matching the printed units is necessary before assessing the model's representativeness.
Check the cargo allocation and sailing distance
Fictional worked example: shipment H17 contains 20 tonnes of packaged goods occupying 40 cubic metres. It travels on one hypothetical 8,000 km sea leg from Port A to Port B in June 2026. All example values are invented to demonstrate allocation, not measured carrier data or published emissions factors.
Assume a homogeneous service boundary with all cargo travelling the same leg and an agreed occupied-volume allocation. The assumed actual cargo volume is 20,000 cubic metres, not the vessel's maximum capacity. The illustrative service total is 200 tonnes CO2e tank-to-wake plus 40 tonnes CO2e well-to-tank. Both components cover that same leg and compatible gas accounting; their sum is 240 tonnes CO2e well-to-wake.
| Allocation record | Fictional value or calculation |
|---|---|
| Cargo volume share | 40 cubic metres ÷ 20,000 cubic metres = 0.002, or 0.2%. |
| Tank-to-wake cargo subtotal | 200 tonnes CO2e × 0.002 = 0.40 tonnes CO2e. |
| Well-to-tank cargo subtotal | 40 tonnes CO2e × 0.002 = 0.08 tonnes CO2e. |
| Well-to-wake sea-leg subtotal | 0.40 + 0.08 = 0.48 tonnes CO2e, or 480 kg CO2e. |
| Distance consistency | The same 8,000 km applies to all cargo; volume-distance shares therefore equal volume shares on this single leg. |
| Excluded connections | Inland collection/delivery, terminals, feeders and transfer services are outside this invented subtotal. |
Table 2, shown one row at a time. Every field is retained.
Row 1
- Allocation record
- Cargo volume share
- Fictional value or calculation
- 40 cubic metres ÷ 20,000 cubic metres = 0.002, or 0.2%.
Row 2
- Allocation record
- Tank-to-wake cargo subtotal
- Fictional value or calculation
- 200 tonnes CO2e × 0.002 = 0.40 tonnes CO2e.
Row 3
- Allocation record
- Well-to-tank cargo subtotal
- Fictional value or calculation
- 40 tonnes CO2e × 0.002 = 0.08 tonnes CO2e.
Row 4
- Allocation record
- Well-to-wake sea-leg subtotal
- Fictional value or calculation
- 0.40 + 0.08 = 0.48 tonnes CO2e, or 480 kg CO2e.
Row 5
- Allocation record
- Distance consistency
- Fictional value or calculation
- The same 8,000 km applies to all cargo; volume-distance shares therefore equal volume shares on this single leg.
Row 6
- Allocation record
- Excluded connections
- Fictional value or calculation
- Inland collection/delivery, terminals, feeders and transfer services are outside this invented subtotal.
The general allocation is service emissions multiplied by the shipment's share of the selected service activity. Cubic metres cancel in this example, leaving emissions units. If different cargo travels different distances, use the applicable leg-specific activity rather than extending this simple ratio across the whole network.
Why not divide by vessel capacity? Capacity includes space that may be unused and is not the assumed actual loaded denominator. Why not divide by 20 tonnes? Shipment mass alone does not describe its share of this volume-allocated service. Marine allocation must follow the applicable method and limiting activity, not a convenient available number.
The real load, stowage measure and carrier acceptance of this rule remain unverified. If those records cannot substantiate the denominator, the open decision is whether this allocation is usable; the consequence is to retain it as an illustrative or provisional result, not a final shipment footprint. No missing connection has been assumed to emit zero.
Label spend estimates as screening
A freight invoice can support a spend screen, but its price is not a direct measurement of vessel activity. It may combine freight, insurance, surcharges, terminal services and inland transport. A price increase can change a spend-based estimate even when the physical shipment is unchanged.
Separate transport expenditure consistently with the economic factor's service definition, currency and price year. Do not apply a sea-freight factor to an undifferentiated logistics invoice and assume that every bundled service is represented correctly. Nor should the resulting screen be added to a physical estimate covering the same activity.
For an important lane, request the booking and consignment references, sailing or service identifier, origin and destination, cargo mass and volume or TEUs, included connections and available carrier data. This connects the financial record to a physical activity record.
Until those inputs arrive, label the estimate as spend-based screening and identify the upgrade required. More decimal places would not resolve the missing route, allocation or service boundary.
Compare shipping choices without changing the question
Fictional offer comparison: use the same H17 cargo, 20 tonnes and 40 cubic metres, on the hypothetical Port A–Port B lane in June 2026, with the same delivery deadline. These are invented comparison records, not actual carrier offers. Confirm that both options can meet the required delivery window before treating them as substitutes.
| Comparison field | Offer A | Offer B |
|---|---|---|
| Allocation basis | Assumed actual cargo volume 20,000 cubic metres; H17 share = 40 ÷ 20,000 = 0.002. | Assumed actual cargo volume 18,000 cubic metres; H17 share = 40 ÷ 18,000. |
| Reported sea-service quantity | 240 tonnes CO2e well-to-wake, including 200 tank-to-wake and 40 well-to-tank. | 180 tonnes CO2 tank-to-wake only; other gases and upstream fuel emissions not supplied. |
| Allocated sea-leg result | 240 × 0.002 = 0.48 tonnes CO2e well-to-wake. | 180 × (40 ÷ 18,000) = 0.40 tonnes CO2 tank-to-wake. |
| Connections and route | Single assumed 8,000 km leg; inland and terminal activity excluded. | Lane identified, but any feeder, transfer and connection coverage remain unconfirmed. |
| Evidence still needed | Actual load and occupied-volume rule, plus separately scoped connections. | Actual load/rule, routing and connections, gas coverage and upstream fuel basis. |
Table 3, shown one row at a time. Every field is retained.
Row 1
- Comparison field
- Allocation basis
- Offer A
- Assumed actual cargo volume 20,000 cubic metres; H17 share = 40 ÷ 20,000 = 0.002.
- Offer B
- Assumed actual cargo volume 18,000 cubic metres; H17 share = 40 ÷ 18,000.
Row 2
- Comparison field
- Reported sea-service quantity
- Offer A
- 240 tonnes CO2e well-to-wake, including 200 tank-to-wake and 40 well-to-tank.
- Offer B
- 180 tonnes CO2 tank-to-wake only; other gases and upstream fuel emissions not supplied.
Row 3
- Comparison field
- Allocated sea-leg result
- Offer A
- 240 × 0.002 = 0.48 tonnes CO2e well-to-wake.
- Offer B
- 180 × (40 ÷ 18,000) = 0.40 tonnes CO2 tank-to-wake.
Row 4
- Comparison field
- Connections and route
- Offer A
- Single assumed 8,000 km leg; inland and terminal activity excluded.
- Offer B
- Lane identified, but any feeder, transfer and connection coverage remain unconfirmed.
Row 5
- Comparison field
- Evidence still needed
- Offer A
- Actual load and occupied-volume rule, plus separately scoped connections.
- Offer B
- Actual load/rule, routing and connections, gas coverage and upstream fuel basis.
B's smaller printed result is not proof of lower emissions. Its CO2-only tank-to-wake basis differs from A's CO2e well-to-wake basis, while connection coverage is incomplete. Neither a shared destination nor similar arithmetic makes those quantities comparable.
Request aligned fuel and gas boundaries, confirmed allocation evidence, equivalent service connections, routing and delivery performance. Recalculate the same cargo task once those inputs are available; a different factor denominator or load assumption must also be disclosed.
The open choice is which offer has the lower supported footprint for the required service. Until the missing evidence is resolved, do not rank B as the climate winner or report the difference as an achieved saving. The comparison can still guide targeted questions to each provider.
Test alternative fuels against life-cycle evidence
An alternative fuel label does not establish a shipment reduction. Ask how the fuel was produced, which upstream stages and gases are covered, and what physical use occurred on the relevant vessel or service. A lower combustion CO2 figure can coexist with material upstream emissions or other greenhouse gases.
Keep physical fuel evidence separate from certificate allocation. A book-and-claim arrangement may allocate an environmental attribute without showing that the cargo travelled on the vessel using that fuel. Establish the scheme, accounting treatment, quantity, period, ownership and protection against duplicate claims before describing the attributed benefit.
The baseline must also match the question. Comparing a fuel pathway with a reference fuel is different from comparing two complete logistics services. Changing the sailing, vessel or included connections at the same time can obscure what caused a difference.
Request the method and underlying records rather than treating “biofuel”, “methanol”, “LNG” or “low-carbon” as a guaranteed result. If the evidence supports only an attributed certificate claim, do not rewrite it as proof that the physical H17 voyage emitted less.
Do not confuse climate totals with port air quality
Climate accounting and port air quality require separately labelled indicators. CO2e aggregates greenhouse gases using a stated accounting basis; it is not a universal measure of every pollutant or health effect around a port.
| Question | Appropriate evidence to keep separate |
|---|---|
| What is the climate result? | Gases covered, tonnes or kilograms CO2e, fuel-life-cycle boundary, method and reporting period. |
| What is the local air-quality result? | Pollutant-specific quantities or concentrations, such as nitrogen oxides or particulate matter, with location, operating conditions and exposure context. |
Table 4, shown one row at a time. Every field is retained.
Row 1
- Question
- What is the climate result?
- Appropriate evidence to keep separate
- Gases covered, tonnes or kilograms CO2e, fuel-life-cycle boundary, method and reporting period.
Row 2
- Question
- What is the local air-quality result?
- Appropriate evidence to keep separate
- Pollutant-specific quantities or concentrations, such as nitrogen oxides or particulate matter, with location, operating conditions and exposure context.
A fuel, engine or shore-power change can affect those outcomes differently. Onboard changes may move energy-related emissions elsewhere, and a local pollutant benefit does not automatically establish a full fuel-life-cycle climate saving.
Report the supported result in its own column. If a provider supplies only a CO2e total, it cannot substantiate a claim about the air breathed by nearby communities. Conversely, a port pollutant measurement is not a substitute for a shipment's greenhouse-gas allocation.
Build a carrier evidence pack that can be updated
Build the evidence pack so that a later reviewer can reproduce the estimate and understand a revision. A spreadsheet number without its service, method and denominator cannot explain why the result changed.
| Record group | Information and update trigger |
|---|---|
| Shipment and voyage | Booking/consignment references, cargo description and identifiers, sailing/service, period, origin and destination; revise if routing or booking changes. |
| Cargo and allocation | Mass, volume or TEUs, measurement basis, actual service activity and allocation rule; revise when load or stowage evidence changes. |
| Emissions basis | Fuel quantities or service total, gas coverage, factor source/version, units and tank-to-wake/well-to-wake boundary; revise when better primary data arrives. |
| Connections and omissions | Inland, terminals, feeders and transfers, inclusion status and overlaps; revise when missing elements are supplied. |
| Accountability | Source-provider date, estimate owner, reviewer, unresolved questions and revision history; retain the previous version and reason for replacement. |
Table 5, shown one row at a time. Every field is retained.
Row 1
- Record group
- Shipment and voyage
- Information and update trigger
- Booking/consignment references, cargo description and identifiers, sailing/service, period, origin and destination; revise if routing or booking changes.
Row 2
- Record group
- Cargo and allocation
- Information and update trigger
- Mass, volume or TEUs, measurement basis, actual service activity and allocation rule; revise when load or stowage evidence changes.
Row 3
- Record group
- Emissions basis
- Information and update trigger
- Fuel quantities or service total, gas coverage, factor source/version, units and tank-to-wake/well-to-wake boundary; revise when better primary data arrives.
Row 4
- Record group
- Connections and omissions
- Information and update trigger
- Inland, terminals, feeders and transfers, inclusion status and overlaps; revise when missing elements are supplied.
Row 5
- Record group
- Accountability
- Information and update trigger
- Source-provider date, estimate owner, reviewer, unresolved questions and revision history; retain the previous version and reason for replacement.
For H17, the estimate owner should seek the load denominator and connection coverage before approving a complete shipment result. A reviewer should be able to distinguish a changed assumption from a genuine change in transport activity.
Preserve both versions when actual records replace a model or screen. A falling estimate is not automatically a measured reduction: it may reflect better boundaries, a corrected allocation or different data quality.
Connect product identifiers without overstating the result
A product identifier can connect transport evidence to the goods that moved, but it does not supply the transport calculation. Link the consignment to the relevant product or batch record only after checking that the quantities, period and physical shipment match.
Keep the carrier evidence and its calculation basis available alongside that connection. Product composition or supplier information cannot establish vessel fuel consumption, the shared-load denominator, fuel-life-cycle factors or coverage of logistics connections. Each still needs its own evidence.
The useful handover is therefore bounded: identify the goods, state the supported transport subtotal, preserve exclusions and assign the remaining questions. For the fictional H17 case, that means keeping a transparent sea-leg allocation without presenting it as an assured door-to-door footprint.
A defensible cargo estimate is not just a smaller shipping number. It is a result whose unit, service boundary, allocation and evidence can be explained, updated and used without overstating what the available records establish.
Freight & distribution
