Well-to-Wake (WtW)
A ship switches from heavy fuel oil to LNG and the funnel emissions fall by about a fifth. The press release writes itself. Then the same switch is measured well-to-wake — extraction, liquefaction, shipping the fuel to the bunker, and only then combustion — and roughly half of that gain is already gone before the vessel leaves the berth.
Well-to-wake is the boundary that decides which of those two numbers you are allowed to quote, and the answer now depends on which regulator is asking.
Well-to-wake counts a marine or aviation fuel’s emissions from extraction to combustion — upstream production plus what leaves the funnel. It is the same boundary road transport calls well-to-wheel. The IMO’s CII is tank-to-wake; FuelEU Maritime and the IMO Net-Zero Framework are well-to-wake.
Well-to-wake is the lifecycle emissions boundary for transport fuels burnt by ships and aircraft: everything from producing the fuel to burning it, expressed per tonne of fuel or per unit of transport work. It is the reporting basis that fuel-standard regulation has converged on, and it is deliberately wider than the operational-efficiency measures that sit alongside it.
What “well-to-wake” means
Well-to-wake (WtW) is the sum of two stages. The first is well-to-tank — extraction, processing, refining or liquefaction, and distribution of the fuel up to the point it enters the vessel’s bunker or the aircraft’s tank. The second is tank-to-wake: the emissions released when that fuel is burnt in the engine.
The arithmetic is simply additive:
Well-to-wake = well-to-tank + tank-to-wake. Upstream production plus combustion. It stops at the fuel: building the ship, maintaining it and scrapping it are outside the boundary, which is what separates well-to-wake from a full life-cycle assessment of the vessel.
The term matters because the two stages behave completely differently under a fuel switch. Combustion emissions are governed by the carbon in the fuel and how efficiently the engine burns it. Upstream emissions are governed by how the fuel was made and moved — which, for liquefied and synthetic fuels, can be a far larger share than it is for a conventional distillate. A boundary that stops at the funnel cannot see that difference at all, and every alternative-fuel comparison lives or dies on it.
Note the deliberate asymmetry with corporate GHG accounting. In a company inventory the two stages land in different scopes: tank-to-wake in Scope 1 for a vessel you own or operate, and well-to-tank in Scope 3 Category 3. Well-to-wake is the sum of the two and is therefore not a reporting line in an inventory — it is a comparison and compliance basis. Adding a well-to-wake figure to a scope total double-counts the combustion half.
Wake, wheel and wing — one boundary, three idioms
The most common confusion around this term is not conceptual. It is that the same boundary carries a different name in each transport mode, and the abbreviations collide.
| Mode | Upstream stage | Combustion stage | Full boundary |
|---|---|---|---|
| Road | Well-to-tank | Tank-to-wheel | Well-to-wheel |
| Marine | Well-to-tank | Tank-to-wake | Well-to-wake |
| Aviation | Well-to-tank | Tank-to-wake (or tank-to-wing) | Well-to-wake (or well-to-wing) |
A ship leaves a wake rather than wheel tracks; an aircraft leaves a wing tip vortex. The metaphor changes and the boundary does not. The upstream half is called well-to-tank in every mode, which is why that term has a single page and this one does not restate it.
“WTW” means well-to-wheel on a lorry and well-to-wake on a ship, and “TTW” means tank-to-wheel or tank-to-wake in the same way. In a single-mode document the abbreviation is harmless. In a multimodal freight report — where a container moves by road, rail and sea inside one consignment — the abbreviations are genuinely ambiguous and the boundary they name is the thing being audited. Write the words.
The split that decides which number you owe
Until recently this was a modelling preference. It is now the line between two live obligations that apply to the same vessel at the same time, on different boundaries, and neither substitutes for the other.
| Instrument | Boundary | What it measures |
|---|---|---|
| IMO Carbon Intensity Indicator | Tank-to-wake | Operational CO₂ per capacity-mile, graded A–E annually |
| FuelEU Maritime | Well-to-wake | GHG intensity of the energy used on board, against a declining limit |
| IMO Net-Zero Framework | Well-to-wake | A declining GHG fuel intensity standard, with a pricing mechanism behind it |
| CORSIA | Well-to-wake | Lifecycle values for eligible aviation fuels |
| GLEC Framework / ISO 14083 | Well-to-wake | Freight emissions allocated to cargo across a transport chain |
The practical consequence is that CII and FuelEU point in the same direction and measure different quantities. Slow steaming burns less fuel per capacity-mile and improves the CII directly. Switching to a lower-carbon fuel of similar consumption improves the fuel-intensity number and may barely move the CII. A ship can improve one without improving the other, and a decarbonisation plan built against only one of them will underperform against the other.
If you need to know whether the rating half of that pair reaches your vessel and whether the estimate you steer by is sound, that is what the IMO CII Rating Obligation Checker assesses.
Units, formula and current marine factors
Well-to-wake factors are published per unit of fuel — kilograms of CO₂e per tonne of fuel is the marine convention — or, once combined with transport work, per tonne-kilometre. Fuel-standard regulation increasingly expresses the same quantity per unit of energy, in grams of CO₂e per megajoule, so that fuels of different energy density compare directly.
The table below is live from the corpus: the combustion (tank-to-wake) factor and its upstream (well-to-tank) companion for the three fuels that dominate the world fleet. Well-to-wake is the sum of the pair on each row.
| Fuel | Tank-to-wake combustion | Well-to-tank upstream | Unit |
|---|---|---|---|
| Marine gas oil (MGO) | 3245.3 | 743.8 | kg CO2e per tonne |
| Marine fuel oil (MFO) | 3154.8 | 714.9 | kg CO2e per tonne |
| LNG | 2536.9 | 912.2 | kg CO2e per tonne |
Read the LNG row against the marine fuel oil row and the shape of the problem appears immediately. LNG has the lowest combustion factor of the three and the highest upstream factor — liquefaction is energy-intensive and the fuel is often shipped a long way before it is bunkered. The gap that looks large at the funnel narrows once the upstream stage is added back.
These are generic fuel factors from DEFRA, on the corpus edition stamped in the page header. They are not engine-specific, which matters for LNG in particular — see methane slip below.
Worked example — the LNG switch that shrinks
A bulk carrier consumes 10,000 tonnes of marine fuel oil a year and is considering a dual-fuel LNG conversion on a like-for-like tonnage basis. What does the switch deliver, and does the answer depend on the boundary?
Values are held still here on purpose; the live factors are in the table above and will move with the corpus.
| Marine fuel oil | LNG | Change | |
|---|---|---|---|
| Tank-to-wake, per tonne of fuel | 3,154.8 kg CO₂e | 2,536.9 kg CO₂e | −19.6% |
| Well-to-tank, per tonne of fuel | 714.9 kg CO₂e | 912.2 kg CO₂e | +27.6% |
| Well-to-wake, per tonne of fuel | 3,869.7 kg CO₂e | 3,449.1 kg CO₂e | −10.9% |
| Upstream share of well-to-wake | 18.5% | 26.4% |
At 10,000 tonnes of fuel a year the tank-to-wake saving is about 6,200 tonnes of CO₂e. The well-to-wake saving is about 4,200 tonnes — roughly two-thirds of it. The same physical decision reported on two boundaries produces a 20% headline and an 11% one, and both are arithmetically correct.
Which one is quotable depends on the audience. Against the CII, only the tank-to-wake half is assessed, so the improvement shows in full. Against a well-to-wake fuel standard, the upstream penalty is charged back and the improvement is smaller. Neither figure is wrong; a figure quoted without its boundary is.
Methane slip, the term with no road analogue
There is one tank-to-wake effect in marine engines that has no equivalent in the road idiom, and it falls hardest on the fuel that looks best at the funnel.
Low-pressure dual-fuel engines burning LNG release a fraction of the gas unburnt. Methane is a far more potent greenhouse gas than CO₂ over a century, and much more potent still over twenty years, so a small unburnt fraction carries a disproportionate warming penalty. The rate depends on engine type, load and tuning — not on the fuel alone.
The LNG factors in the table above are generic fuel factors. They are not tied to a particular engine type, and no fuel-indexed factor can be: slip is a property of the machine, not of the molecule. Fuel-standard regulation handles this by applying engine-specific defaults on top of the fuel’s carbon content, which is why a regulatory well-to-wake figure for an LNG vessel can be materially higher than a fuel-table calculation suggests. If an LNG business case rests on a fuel-table number alone, the slip term has not been priced.
This is also why the well-to-wake boundary is the one fuel-standard regulation converged on. A tank-to-wake measure that counts only CO₂ rewards a fuel with a favourable carbon-to-energy ratio regardless of what it took to make it and regardless of what escapes unburnt. Widening the boundary is what closes both gaps at once.
Where well-to-wake is required — and common mistakes
Well-to-wake is the required basis for fuel-standard compliance in shipping and aviation, for freight emissions reported under ISO 14083 and the GLEC Framework, and for any credible comparison between two fuels or two propulsion routes. Whether your own freight reporting meets that standard is what the GLEC & ISO 14083 Conformance Checker tests; the sea freight calculator quantifies a movement on that basis, and the sea freight methodology sets out how the allocation is done.
Four mistakes account for most of the errors we see.
| Mistake | What goes wrong |
|---|---|
| Reporting well-to-wake as Scope 1 | Inflates Scope 1 by the whole upstream share — around a fifth for a distillate and a quarter for LNG. The combustion half is Scope 1; the upstream half is Scope 3 Category 3. See the well-to-tank methodology. |
| Adding well-to-wake to a scope total | Double-counts combustion, because well-to-wake already contains it. Well-to-wake is a comparison basis, not a reporting line. |
| Comparing a tank-to-wake figure with a well-to-wake one | Produces a fuel switch that looks roughly twice as good as it is. This is the single most common alternative-fuel overstatement, and it is usually accidental rather than promotional. |
| Treating a CII improvement as fuel-standard compliance | Different boundary, different instrument. Both apply simultaneously to the same vessel, and neither discharges the other. |
The underlying discipline is small and it is the whole of the term: state the boundary next to the number, every time. A marine emissions figure without its boundary named is not a conservative figure or an approximate one. It is an unreadable one, because the two available readings differ by the upstream share and there is no way to tell from the number which was meant.
Yes — the same boundary in a different idiom. Road transport says well-to-wheel because a vehicle has wheels; shipping says well-to-wake because a vessel leaves a wake; aviation says well-to-wake or well-to-wing. In all three the boundary is identical: fuel production plus fuel combustion, stopping short of building the vehicle. The upstream half is called well-to-tank in every mode.
No. The CII is tank-to-wake: it counts only the CO₂ from fuel combusted on board, per unit of capacity-mile. FuelEU Maritime and the IMO Net-Zero Framework’s fuel standard are the well-to-wake instruments. That split is deliberate — the CII grades operational efficiency, while a fuel standard targets the carbon content of the energy itself — and it means the same vessel is measured on two boundaries at once.
No, and it is an expensive mistake in both directions. Well-to-wake bundles upstream and combustion emissions, so using it as a Scope 1 figure inflates Scope 1 by the entire upstream share. A compliant inventory splits it: tank-to-wake to Scope 1 for vessels you own or operate, well-to-tank to Scope 3 Category 3. Reporting the combined figure in either scope also double-counts against the other.
Because liquefying and transporting LNG is energy-intensive, so its upstream factor is the highest of the common marine fuels while its combustion factor is the lowest. Adding the upstream stage back therefore erodes a large part of the funnel-side advantage. On the figures in the table above, a like-for-like switch from marine fuel oil reads as roughly a fifth better tank-to-wake and about a tenth better well-to-wake — before engine methane slip is considered at all.
Methane slip is unburnt gas passing through a dual-fuel engine and out of the exhaust. It sits in the tank-to-wake stage, and therefore inside well-to-wake. It is engine-specific rather than fuel-specific — the rate depends on engine type, load and tuning — so a generic fuel factor cannot express it, and fuel-standard regulation applies engine-based defaults instead. An LNG case built only on fuel-table factors has left the slip term out.
No. Well-to-wake is a fuel boundary: it stops at extraction on one end and combustion on the other. Construction, maintenance and end-of-life belong to a full life-cycle assessment of the vessel, which is a wider and far less standardised exercise. For most ships the operational fuel dominates the lifetime total, which is why fuel-boundary accounting is the practical basis for regulation.
Three forms are common. Per tonne of fuel is the marine bunkering convention and is what the table above uses. Per megajoule of energy — grams of CO₂e per MJ — is the fuel-standard convention, because it lets fuels of different energy density compare directly. Per tonne-kilometre is the freight convention, once fuel has been allocated to cargo and distance. Converting between them needs the fuel’s energy content and the voyage’s transport work, so a factor quoted without its unit basis cannot be reused.
Use well-to-wake, and spell it out. Multimodal reporting under the GLEC Framework and ISO 14083 puts every leg on the same lifecycle boundary so the legs can be summed, and mixing a well-to-wheel road leg with a tank-to-wake sea leg produces a total that is neither. The abbreviations are the trap rather than the concept: WTW and TTW each mean two different things depending on the mode, and in a multimodal document both readings are present on the same page.
Need someone who does this? 6 carbon accounting & inventory providers in our directory →