A container voyage that emitted 557 tonnes less CO₂ turned out to be the less efficient of two sailings, once you divided emissions by the cargo actually carried. That's the core finding of a new VesselBot case study, and it should make anyone signing an emissions-based carrier contract a little nervous.

The analysis compares two voyages by the same vessel on the Bremerhaven–Norfolk lane in May and July 2026. The May voyage generated 3,104 tonnes of CO₂ equivalent while carrying 2,783 TEUs — 1.12 tonnes of CO₂e per TEU. The July voyage produced only 2,547 tonnes of CO₂e, but it moved just 2,077 TEUs, pushing intensity up to 1.23 tonnes per TEU.

So the dirtier voyage on paper was the cleaner one per box. Nearly 10% more efficient, in fact. Total emissions flatter the light ship.

Why utilisation drives the maths

VesselBot argues total CO₂e measures environmental impact but says nothing about how efficiently those emissions were used. Its preferred lens is Well-to-Wake emissions intensity, expressed in grams of CO₂e per TEU-kilometre — emissions divided by the cargo moved and the distance covered.

The company's Q2 2026 dataset across container voyages shows what that lens reveals:

• Average Well-to-Wake intensity: 231.7 grams of CO₂e per TEU-kilometre.
• Average emissions per voyage: 651.3 tonnes of CO₂e.
• The most efficient category, up to 49 grams per TEU-kilometre, averaged 1,511 tonnes of CO₂e per voyage — with 80% vessel utilisation and 8,946 TEUs carried.
• The least efficient category, 800 grams and above, averaged just 545 tonnes per voyage — on 31% utilisation and a paltry 340 TEUs.

The pattern is blunt. Full ships burn more fuel in absolute terms, but they spread that burn across vastly more transport work. A third-full ship can post a modest total and still be, per box moved, an environmental laggard.

What does this change for shippers?

Plenty, if you're scoring carriers on emissions. A shipper booking 500 TEUs on a well-utilised sailing may carry a lower per-box carbon footprint than one booking on a lightly loaded voyage with a smaller headline number. Judging carriers, services or individual sailings by total emissions rewards the wrong behaviour.

VesselBot's remedy is voyage-level modelling. Fuel consumption depends on vessel characteristics, engine performance, sailing speed, time at sea, port stays and operating conditions, so comparisons built on distance or industry averages can misrepresent an individual vessel's operation. Combining fuel-consumption modelling with shipment-level cargo data lets emissions be set against the transport work each voyage actually delivered.

That approach also separates the two cases shippers most want to distinguish: a voyage producing fewer emissions because it carried less cargo, and one that genuinely burns fuel more efficiently. The first is a scheduling accident. The second is an operator doing its job.

The study's verdict is simple enough to pin to a wall: treat absolute emissions and operational efficiency as separate measures. With EU ETS costs increasingly landing on invoices, expect shippers and forwarders to start asking which number their carrier is quoting — and how full the ship was when it was generated.