Field notes
Working notes on verifier-grade maritime compliance.
Engineering, methodology, and regulatory perspectives. Written for shipowners, technical managers, DPAs, CFOs, and the verifiers who sit across the table. No marketing copy.
One voyage, five regimes: why your compliance numbers stop agreeing
The same tonne of bunker fuel feeds FuelEU Maritime, EU ETS, IMO DCS and IMO CII — each with a different scope, boundary and emission factor. Reuse one number naively and at least one regime is wrong. Here is why, and the data design that fixes it.
Read noteAll notes
Reconciled to the exact euro: Provenship against the EMSA 2023 public dataset
Provenship's calculation engine reconciles to the exact euro against EMSA's published 2023 emission report across 12,637 vessels. Here is how the ±0.0000% delta was produced, and what was deliberately out of scope.
The FuelEU penalty formula, derived line by line
Annex IV Part B of Reg. (EU) 2023/1805 is short, but the chain from gCO2eq-per-MJ to euros runs through three conversions a sponsor often misses. Here is each step.
Carbon factors, validated against the population: 75,000 vessel-years say 3.13–3.15
Annex II of FuelEU lists 3.114 t CO2 per tonne HFO as the well-to-wake baseline. The median implied factor across 75,000 published vessel-years lands at 3.13–3.15. Here is why we run that check, and what it catches.
Why GLEC and FuelEU disagree on LNG by 30%
Provenship's WtW LNG figure is 3.748 t CO2eq/t. GLEC v3.0 says 2.870. The 30% gap is real, it is documented, and both methodologies are legitimate. The difference is how each treats methane slip.
Refuse rather than approximate: the engine's hardest design choice
When data is insufficient to compute a defensible euro figure, Provenship's engine returns REFUSED_NEEDS_VERIFIER_INPUT instead of an approximation. This is the same posture that makes the rest of the engine trustworthy. Three real scenarios.