GLEC v3.2 in detail — what changed from v3.0 and v3.1, and what changes again in v4

GLEC v3.2 is the current edition of the Smart Freight Centre’s freight emissions methodology as of March 2026. Version 4 is in draft and is expected late 2026. Between v3.0 (2020) and v3.2 there are five changes that actually move numbers in production calculators: a marine container-ship segmentation split, an LNG well-to-wake (WTW) revision driven by methane slip data, a road articulated-truck lane-utilisation adjustment, an air belly-cargo allocation update to the ISO 14083:2023 default, and a separate hub/transshipment line item across every mode. The rest is editorial. The published framework is on the Smart Freight Centre site; the underlying standard is ISO 14083:2023.

I read GLEC change notes and IMO MEPC documents for a living. What I’ll do here is walk through the five v3.2 changes that matter, say what stayed identical so you stop worrying about it, and give my read on the v4 draft. For most operators this is a factor-table swap, not a model rewrite. For the dated record of which revision we pinned into the calculator and when, see our methodology changelog; for the per-shipment migration math I went through with one client, the GLEC v3.0 vs v3.2 post covers the same five changes through a Shanghai–Rotterdam worked example.

The five changes that move numbers

1. Marine: container ship segmentation 3,000–8,000 TEU vs 8,000+ TEU

Through v3.0 and v3.1, “container ship, average fleet” was a single class with one default WTW factor of about 8.4 g CO₂e per tonne-km. The problem with that is the fleet has bifurcated: ultra-large container vessels above 8,000 TEU are meaningfully more efficient per tonne-km than the 3,000–8,000 TEU panamax-to-neo-panamax range that does most of the regional and feeder work. GLEC v3.2 §5.3 splits the class.

The two new defaults, WTW, are roughly:

• Container ship 3,000–8,000 TEU: ~9.1 g CO₂e/tkm
• Container ship 8,000+ TEU: ~7.6 g CO₂e/tkm

A Shanghai–Rotterdam shipment on a 14,000 TEU ULCV now comes in roughly 10% lower per tonne-km than the old undifferentiated factor; a regional shipment on a 6,000 TEU vessel comes in roughly 8% higher. Net impact on a deep-sea portfolio is small (1–3% either way depending on lane mix), but you have to know which class the carrier deployed. If you don’t, v3.2 says fall back to the 3,000–8,000 TEU factor, which is the more conservative choice. My view: the split is the right call, but it leaves the under-3,000 TEU feeder segment without a clean default. v4 will probably fix that.

2. Marine: LNG bunker WTW revised down ~5% on methane slip data

The LNG-fuelled marine factor in v3.0 used an upstream methane assumption that the ICCT’s 2020 and 2023 measurement campaigns showed was too pessimistic for modern dual-fuel engines and too optimistic for older steam-turbine designs. GLEC v3.2 §5.4 replaces the single LNG well-to-tank (WTT) component with engine-type splits. The aggregate WTW factor for LNG, weighted across the operating fleet, drops by about 5% — not because LNG got cleaner, but because the methodology now distinguishes between engines where methane slip is a real problem (low-pressure dual-fuel under low-load conditions) and engines where it isn’t.

A worked comparison: a Qatar–Rotterdam LNG carrier voyage that came out around 24 g CO₂e/tkm under v3.0 is roughly 22.8 g/tkm under v3.2, assuming a high-pressure dual-fuel engine class. For low-pressure dual-fuel, the v3.2 number is slightly higher than v3.0, not lower. The reference data is the ICCT methane slip work; the engine-class breakdown traces to the IMO Fourth GHG Study (2020) and MEPC 80 and 81 supplementary papers.

My take: the revision is overdue. The bigger issue is that operators rarely know which engine class the carrier is actually running on a given voyage, so the Tier 1 default falls back to a fleet-weighted average — fine in aggregate, but loses the point of the split for any individual disclosure.

3. Road: articulated truck (>32 t GCW) lane-utilisation defaults

GLEC v3.0 baked in a sector-average empty-mileage assumption of 25% for European articulated trucks above 32 tonnes gross combined weight. Post-2022 European Commission and Eurostat road-freight statistics show empty mileage in this segment has crept up to about 28%, partly because of regional dedicated-lane retreat after the 2022–2024 energy-price shock and partly because more carriers shifted to backhaul-by-spot rather than backhaul-by-contract. GLEC v3.2 §6.2 adjusts the default lane utilisation accordingly.

Concrete impact: the default WTW factor for an articulated truck >32 t GCW moves from about 62 g CO₂e/tkm (v3.0) to about 64.5 g CO₂e/tkm (v3.2), a ~4% increase. If you used the v3.0 default to model a Munich–Milan corridor in the worked example post, the new number nudges up from 365 kg to roughly 380 kg CO₂e for the same shipment. Operators with carrier-reported empty-mileage data can still claim a Tier 2 adjustment downward; the change just moves the Tier 1 default.

4. Air: belly-cargo allocation updated to the ISO 14083:2023 mass-based default

The belly-cargo allocation between passengers and freight has been the most contested line in air-cargo accounting for a decade. GLEC v3.0 used a hybrid volume-and-mass approach inherited from an older IATA RP 1678 draft. ISO 14083:2023 settled the matter by codifying a mass-based default with a documented option for volume-based where mass data is unavailable. GLEC v3.2 §7.4 adopts the ISO default verbatim.

Practical impact on long-haul belly cargo: the v3.0 hybrid allocation gave roughly 595–620 g CO₂e/tkm depending on aircraft type. Pure mass-based under v3.2 gives roughly 605–610 g/tkm. So the headline factor barely moves; what changes is the consistency — under v3.0 you could get different numbers from different calculators using nominally the same factor because each was making its own hybrid-allocation choice. Under v3.2 the allocation is unambiguous. That’s a real win for audit trail purposes even though the numerical delta is small.

One acknowledged gap: I have not yet seen a v3.2 audit where Tier 3 primary data for belly-cargo allocation was actually required end-to-end. It will happen. CSRD assurance providers are scoping up on this through 2026.

5. All modes: hub and transshipment emissions broken out

Through v3.1, hub operations — container yard handling, air-cargo terminal lift, rail-truck transfer — were bundled into the per-tonne-km factor of the upstream or downstream leg. That hid roughly 1–3% of total freight emissions and made it impossible to attribute hub-side electrification separately. GLEC v3.2 §9.1 breaks hub and transshipment out as its own line, with default per-handling factors that vary by terminal type and electrification grade.

For multi-leg shipments this adds 2–5 kg CO₂e per TEU transshipment hop and 0.6–1.2 kg CO₂e per truck-rail transfer. Small individually, but on a portfolio with heavy transshipment exposure (Singapore, Algeciras, Rotterdam) the cumulative effect matters — one large 3PL portfolio I looked at moved up about 1.8% on this line alone. The Tier 2 expansion in v3.2 is overdue but the data quality criteria are still vague: a terminal claiming “80% electrified yard equipment” means different things depending on whether reefer plug-in or RTG primary movement counts in that percentage. v4 will need to tighten this.

What did not change

This is the part that managing-expectations conversations rest on. None of these changed between v3.0 and v3.2, and none of them are changing in v4 either:

• The fundamental WTW / TTW / WTT split. Well-to-wake is still the headline figure, tank-to-wake plus well-to-tank still adds up to WTW, and CSRD Scope 3 Category 4 disclosure still expects WTW.
• The activity unit. Tonne-kilometre is still the unit of allocation. The push from some quarters to move to a payload-utilisation-adjusted unit didn’t make it into v3.2 and isn’t in the v4 draft.
• The data quality tier definitions. Tier 1 is still defaults, Tier 2 is still modelled or carrier-fleet-average, Tier 3 is still primary data on the specific trip. The wording got tightened in §3.2 but the structure didn’t move.
• The great-circle versus routed distance rule. Air is great-circle plus the GLEC distance-adjustment factor; deep-sea ocean is routed via standard navigation; road and rail are routed. Short-sea ocean is still the messy case where v3.2 punts to “use routed if you have it, great-circle plus correction otherwise.”

If you were comfortable with the v3.0 calculation structure, v3.2 doesn’t require you to rebuild it. It requires a factor-table swap and two or three new line items.

What v4 changes (in draft)

The v4 draft has been circulating in working-group form since late 2025. Three changes are likely to land:

SAF accounting rules. Sustainable aviation fuel is handled inconsistently today — some frameworks apply a 100% reduction on the WTT side, some apply the certified lifecycle reduction, some allow book-and-claim that disconnects the fuel from the flight entirely. The v4 draft moves toward book-and-claim with a hard requirement that the certificate be retired in a named registry. This will land. The contentious piece is what happens to the carrier’s own scope-3 baseline if a shipper claims the SAF certificate; the working group hasn’t reached consensus.

Biofuel sustainability criteria. The draft uses EU Renewable Energy Directive III (RED III) as the baseline test for whether a biofuel counts as emission-reducing at all. Operators on non-RED-III-certified biodiesel blends will see those reductions disallowed. My read: this will survive, though the implementation date may slip to give the supply chain time to catch up.

Rail electrification grid factors. Today electrified rail uses a national-grid average for the country of operation. The v4 draft allows a railway-specific traction-grid factor where the operator can document its own electricity purchasing — which for SNCF, DB, and NS is meaningfully cleaner than the national average. This will survive in some form; the argument is over the evidence the operator has to provide.

What probably won’t survive: a proposal to add a mandatory aviation radiative-forcing multiplier. IPCC AR6 puts the multiplier between 1.7x and 3x, but the working group is conservative about mandating a factor with that much uncertainty in it. I expect v4 to mention radiative forcing in informative annex text and leave mandatory disclosure to a future revision.

What this means for an existing calculator

If your calculator is on v3.0 or v3.1 today, your numbers are within roughly 5–15% of v3.2 depending on which modes dominate your traffic. The biggest single delta is the LNG marine factor and the new container-ship class split, both of which move marine numbers by a few percent in either direction. Road moves up about 4%. Air moves by less than 2%. The hub/transshipment breakout adds 1–3% to multi-leg shipments depending on routing.

Migration is mostly a factor-table swap. The schema changes are small: two new container-ship classes, an engine-type field on LNG marine, a separate hub-handling line per leg. The calculation logic doesn’t change. If you have a versioned factor table (you should), pinning v3.0, v3.1 and v3.2 side-by-side is straightforward and lets you offer customers a switch-over date that doesn’t coincide with their year-end disclosure crunch.

The EcoFreight calculator runs v3.2 as the default and supports v3.0 and v3.1 as named historical methodologies for restating prior-year baselines. The factor table is published on the methodology page with the version stamped on every row.

What this means for an existing report (CSRD, CDP)

Both CSRD ESRS E1 and CDP Climate accept GLEC v3.0 and v3.2, provided you disclose the version and the data quality tier per mode. Restating prior-year numbers under v3.2 isn’t required if you originally reported under v3.0, but it is methodologically defensible and several assurance providers I work with prefer it because year-over-year deltas become comparable. If you restate, declare the restatement explicitly and footnote which factors changed.

One nuance: if you used a hybrid v3.0/v3.1 mixed-version calculator (some legacy tools were, until 2024), an auditor will ask you to pick one. The cleanest answer is to restate on v3.2 for the current year and disclose the prior year on whatever the documented methodology was. Don’t silently swap the factor table mid-year — that’s the single most common audit finding I see. When v4 lands, the same rule applies: disclose the version, disclose the tier, don’t swap mid-period. Auditors care less about which version you used than about whether you said which one it was.

One honest gap I am still working on

Where v3.2 still under-specifies: belly-cargo allocation on ULCV class container vessels carrying mixed-mode shipments that include refrigerated reefer plug-ins from the upstream port move. The standard tells me to default to ISO 14083 mass-based allocation, which I do, but the reefer auxiliary load gets absorbed into the vessel-class average rather than broken out. For a portfolio with heavy reefer exposure that is a real understatement. I expect v4 to address it; today, I flag it on every reefer-heavy disclosure I touch.

Closing

The v3.2 revision is incremental, defensible, and well-evidenced. The five changes above are the ones that move production numbers; the rest is editorial cleanup an end user never sees. The current factor table is on the methodology page. The arithmetic worked end-to-end on three shipments is in the worked-example post. The calculator runs v3.2 by default with v3.0 and v3.1 available as restatement options, and the per-leg version tag ships on every response from the freight emissions API. When the v4 final text drops I’ll do the same write-up for that revision.

Sources

GLEC Framework v3.2, Smart Freight Centre, 2023 — smartfreightcentre.org. ISO 14083:2023 Quantification and reporting of greenhouse gas emissions arising from transport chain operations — iso.org/standard/78864.html. IMO Fourth GHG Study (2020) for marine factor lineage and engine-class breakdown reference. ICCT methane slip studies (2020, 2023) for the LNG WTW revision, available on the International Council on Clean Transportation site. European Commission Eurostat road-freight statistics for the post-2022 empty-mileage data underlying the articulated-truck adjustment. Section references throughout (§5.3, §5.4, §6.2, §7.4, §9.1) refer to GLEC Framework v3.2 as published.