The fundamental problem with comparing ships by raw emissions is that bigger ships emit more — but they also do more work. A fair comparison asks a narrower question: among ships of the same type, size, and age, which emit the least for the work they do.

The methodology uses k-means clustering within each ship type to identify natural peer groups based on size and age. A bulk carrier of 32,000 deadweight tonnes is not compared with a 180,000-tonne vessel; a five-year-old container vessel is not compared with a twenty-year-old one.

Within each peer group, the median CO₂ per tonne per nautical mile is calculated, and each ship’s percentile position against that median produces its rank.

A company’s emissions footprint is determined by the ships it operates. The methodology converts each ship’s emission pattern into a percentile score from 0 to 100, then combines all the scores belonging to one operator using a weighted average.

To make the operator comparison fair, the methodology then finds each operator’s genuine peers by building a fleet fingerprint. Cosine similarity identifies which operators run the same kind of fleet.

The final comparison is each operator against the median score of its closest peer operators.

The methodology tracks each ship’s carbon intensity year by year and calculates how quickly it is improving. This is called the Carbon Annual Intensity Reduction, or CAIR.

That improvement rate is then compared against its peer group’s average. The comparison is what matters — a ship in a slow-moving group can still be a leader if it outruns its peers. A ship in a fast-moving group can still fall behind if it does not keep up.

The MRV data reports each ship’s annual CO₂ in four mutually exclusive segments: intra-EU voyages, outbound voyages from EU to non-EU ports, inbound voyages from non-EU to EU ports, and emissions while at berth in port.

The composition matters because where carbon occurs determines what can be done about it.

The EU MRV data provides the inputs for a simplified version of CII: total CO₂ divided by capacity multiplied by distance sailed. The result is compared to a per-type required threshold to produce an A to E rating.

The required threshold drops a few percent each year. A ship that holds its emissions flat moves from C to D to E as the target descends beneath it. The projection traces each ship forward to 2030.

The projection shows when each ship will tip from C to D/E if no improvements are made. It also calculates how much it needs to improve each year to stay compliant.

The fleet overview brings all the per-ship insights together at the operator level. It shows each company’s total emissions and what share of that comes from D and E rated ships.

This is the at-risk percentage CO₂. Weighting matters — a company may have only ten ships rated D or E, but if those ten are its biggest emitters, the percentage will be high. That is where decarbonisation effort really makes a difference.