The Routing Decision Hasn't Changed. What It Costs Has.

Weather routing as a discipline has not changed much since the 1990s. The aim is still to get the cargo to the discharge port without taking damage or losing time. What has changed is what a single routing decision is now accountable for, and the master making the call carries more of it than the model does.

Twenty years ago a routing decision answered one question: how do we get this cargo to the discharge port without taking damage or losing time? Good routing saved a day here, a few tonnes of bunker there, and kept the crew off a bad sea state. The number that mattered was fuel, and fuel was visible.

The same decision now carries three more costs that do not show up on the bunker invoice. That is the shift, and it is the reason a route chosen on fuel alone can be the wrong one.

What the same decision is accountable for now

• •Fuel cost: still the biggest absolute number, still the most operationally visible.
• •CII rating contribution: every voyage adds to the annual AER calculation, and the contribution varies by routing choice more than most operators think.
• •FuelEU intensity: for EU voyages, the routing decision interacts with fuel type and pooling economics.
• •ETS allowance cost: for EU voyages, every additional tonne of HFO burned is a priced carbon liability.

Stack those four and the consequence is plain. A route that optimizes only for fuel can cost more than one optimized across all four, because the three carbon-linked costs can move against you while the fuel number looks fine. Most fleets have not made that shift in how they decide.

The mechanics are the same. The inputs decide whether it works.

• •High-frequency weather data with enough resolution to matter. A forecast at 25 nautical miles is not useful when squalls move on 5-mile scales. The gap between the best provider and the average one is wider than most operators realize until they switch.
• •A fuel-consumption model calibrated to the specific vessel, not a sister-class average. A 5% error in the fuel curve means every routing recommendation is built on fiction. This is where most off-the-shelf routing tools quietly fail.
• •Real-time decision updates during the voyage. A plan set at departure is worth less on day four than it was on day one. If the routing system cannot recalculate as conditions change, the voyage is committed to a worse outcome before the weather even arrives.

The part the operators getting this right have worked out

None of this removes the master from the decision, and the better operators do not pretend it does. They build a feedback loop instead. They track the gap between the routed voyage and the sailed one, and when the captain deviates they want to know why, and whether it was a judgment call or a sign the model is wrong. The model that argues with a master who has run the route thirty times is usually the one that needs correcting.

• •Tracking the gap between the routed voyage and the sailed voyage, so a deviation becomes a question about the model rather than a complaint about the master. That loop closes the modeling error over time.
• •Running routing as a forecast exercise, not just an execution exercise. Before the charter is signed, the voyage economics are modeled, including CII contribution and ETS exposure, and the fixture is priced accordingly.
• •Making routing decisions commercially visible. The chartering team sees the intensity and allowance implications of a routing choice, not just the fuel number.

The weather has not become more dangerous or more important. The financial consequence of each weather decision has, and that consequence now sits across four ledgers instead of one. Until a fleet can put all four in front of the person making the call, in time to change it, weather routing is solving the 1990s problem while paying the 2026 bill.