Four Regimes, One Voyage: What Each One Counts

Rotterdam to New York is 3,300 nautical miles. At 12.5 knots that is 264 hours of steaming, and over those eleven days our reference ship, a 50,000 dwt MR product tanker, burns 28.5 tonnes a day of VLSFO and 3.2 tonnes a day of MGO in her auxiliaries: 313.5 and 35.2 tonnes, and 1,089 tonnes of CO2 out of the funnel. That is the physics, and it is the only unambiguous number in this lesson. Four regimes read that leg and write down four different answers, in four different units, over four different boundaries, and not one of them is 1,089.

Our free course Maritime Regulations, Explained sets out what each regime is; this course does the sums. Every constant below is the position as it stood in August 2026, and this is the fastest-moving material in the Academy: check each one against the current instrument before you quote it. The seven lessons after this one take these four boundaries in turn.

Four questions, three axes

MRV and DCS are meters, not targets. The IMO Data Collection System gathers fuel by type and distance sailed from every ship of 5,000 GT and above worldwide, through the flag State to the IMO; the CII then grades only the cargo, ro-pax and cruise types among them. The EU's MRV regulation, Regulation (EU) 2015/757 as amended by 2023/957, gathers the same measurements, independently verified, from any ship of 5,000 GT and above calling at an EEA port, and since 1 January 2025 from general cargo ships of 400 to 5,000 GT and offshore ships of 400 GT and above as well. DCS feeds the CII, MRV feeds the ETS and FuelEU, and one set of flowmeters serves both; where the two reports disagree the problem is on board, not in Brussels.

The three regimes that bite differ along three axes: mass or intensity, global or geographic, funnel or well to wake.

• The Carbon Intensity Indicator (CII) counts mass of CO2 per unit of transport work over the calendar year, worldwide: grams of CO2 over deadweight times distance sailed.
• The EU Emissions Trading System counts tonnes of CO2 inside a geographic boundary and turns each tonne into one allowance to surrender.
• FuelEU Maritime counts grams of CO2 equivalent per megajoule of energy used on board, well to wake, inside that same geographic boundary, with methane and nitrous oxide converted at GWP100.
• The IMO's draft Net-Zero Framework counts the same kind of intensity globally, on its own default factors, as the greenhouse gas fuel intensity, or GFI.

Two constants per fuel do most of the work. The funnel arithmetic is

CO2 = FC × C_F

where FC is the mass of fuel burned in tonnes and C_F the carbon factor in tonnes of CO2 per tonne, fixed by the IMO's EEDI calculation guidelines, MEPC.364(79), which superseded MEPC.308(73) with the same values, at 3.114 for heavy fuel grades (ISO 8217 RME to RMK, where an RMG-quality VLSFO sits) and 3.206 for marine gas oil. The energy arithmetic adds the lower calorific value, 40,500 MJ per tonne for that RMG VLSFO and 42,700 for MGO; well to wake adds the FuelEU Annex II defaults, 91.74 gCO2eq/MJ and 90.77. Every conventional liquid fossil fuel lies within a gram of every other, which is the first hope this course takes away.

The same twelve months, counted four times

Over 2026 the ship burns 6,100 tonnes of VLSFO and 1,150 tonnes of MGO and sails 62,000 nautical miles, about 207 days of steaming. Funnel CO2 is 3.114 × 6,100 + 3.206 × 1,150 = 22,682 tonnes.

CII. Transport work is 50,000 × 62,000 = 3.10 thousand million tonne-miles, so 22,682 tonnes of CO2 over it is an attained CII of 7.317 grams per tonne-mile. The tanker reference line is CII_ref = a × DWT^(-c) with a = 5,247 and c = 0.61, giving 7.137 at 50,000 dwt; the 2026 reduction factor is 11 per cent, so the required CII is 6.352. The tanker band edges are 0.82, 0.93, 1.08 and 1.28 times that: 5.209, 5.907, 6.860 and 8.131. Attained 7.317 lands in the fourth band, so the rating is D, and reaching a C means cutting 6.2 per cent. Reduction factors run 5, 7, 9 and 11 per cent for 2023 to 2026, and resolution MEPC.400(83) of April 2025 filled in the rest of the ladder: 13.625, 16.25, 18.875 and 21.5 per cent for 2027 to 2030.

The Atlantic leg alone scores 1,089 tonnes over 50,000 × 3,300 tonne-miles, or 6.60, a C with under four per cent to spare. A laden passage is the best a ship ever looks; the year is worse because berths and anchorages burn fuel and make no miles.

EU ETS. Only European fuel counts: 100 per cent of what burns on intra-EEA voyages and in an EEA port of call, 50 per cent on a voyage into or out of the EEA, nothing outside. Here that is 3,250 tonnes of VLSFO and 590 tonnes of MGO, so 3.114 × 3,250 + 3.206 × 590 = 12,012 tonnes of CO2. The phase-in was 40 per cent of 2024 emissions and 70 per cent of 2025, reaching 100 per cent for 2026, so the obligation is 12,012 allowances, EUR 900,900 at an assumed EUA price of EUR 75. Treat that as a floor: methane and nitrous oxide join the surrender obligation for 2026 and are not in it.

FuelEU. In-scope energy is 3,250 × 40,500 + 590 × 42,700 = 156,818,000 MJ, and the attained intensity is the energy-weighted average of the two defaults, taken unrounded: (131,625,000 × 91.74420 + 25,193,000 × 90.76745) / 156,818,000 = 91.59 gCO2eq/MJ. The 2026 limit is the 2025 step, 2 per cent below the 91.16 baseline: 89.34. The balance, (89.34 - 91.59) × 156,818,000, is a deficit of 353 tonnes of CO2 equivalent. The penalty divides it by the attained intensity times 41,000 MJ per tonne of VLSFO equivalent and prices that at EUR 2,400: 353 million grams over 91.59 × 41,000 is 94.0 tonnes, about EUR 225,600.

The draft IMO framework. GFI is global, so all 296,155,000 MJ count, and the same two fuels return 94.74 gCO2eq/MJ, 3.15 grams above the European figure. The scope is not the reason: the IMO's life-cycle guidelines, MEPC.391(81), publish their own default factors, with 16.8 grams upstream for VLSFO and 17.7 for MGO against FuelEU's 13.5 and 14.4, and methane and nitrous oxide on the IPCC's fifth assessment potentials rather than the fourth. For 2028 the draft sets a base target of 89.568 and a direct compliance target of 77.439 against a 93.3 baseline, and a ship above both owes both tiers: Tier 1 of (89.568 - 77.439) × 296,155,000 = 3,592 tonnes at US$100, and Tier 2 of (94.74 - 89.568) × 296,155,000 = 1,532 tonnes at US$380, about US$941,000 together. That keeps the 40,500 MJ a tonne used above for VLSFO; the IMO table's own 40,200 gives 95.23 and about US$990,000. Every figure there is draft: MEPC 83 approved the text in April 2025, October 2025 did not reach consensus, and the adoption vote moved to a special session in late 2026. Check the current position before quoting it.

The bill is a trading pattern, not a ship

12,012 of 22,682 tonnes is 53 per cent, and that fraction, not the ship's design, decides the European bill. The rule that moves the most money is the definition of a port of call: a stop for bunkering or drydocking does not break a voyage, and nor, for a containership alone, does a call at a listed transhipment port, so Rotterdam to New York with bunkers at Las Palmas stays one 50 per cent voyage. Insert a cargo call at a European port and that leg becomes a 100 per cent leg plus a fresh 50 per cent leg.

The levers do not point the same way either. Slow steaming cuts absolute CO2, so it cuts the allowance bill and improves the CII; FuelEU divides by energy, so on an unchanged fuel mix its intensity does not notice; the deficit and its penalty shrink only in step with the energy burned, and never change sign. Fuel switching inside the fossil family is worth almost nothing: per megajoule MGO emits 3.206 / 42,700 = 75.1 grams of CO2 against VLSFO's 76.9, so an all-MGO year, at the same energy, cuts funnel CO2 by 2 per cent and the intensity by 1. Against an 11 per cent CII factor and a ladder heading for 18.23 gCO2eq/MJ by 2050, that is noise.

One comparison is worth carrying away. That deficit cost EUR 639 a tonne of CO2 equivalent against EUR 75 for an allowance, because the penalty is charged on energy rather than carbon: the implied price is 58,537 divided by the attained intensity, so the cleaner the ship the dearer each remaining tonne. On a fossil-fuelled ship it is FuelEU, not the ETS, that binds.

Monday morning

Pull last year's verified MRV report and the DCS submission for one ship and check that the fuel totals agree; if they do not, nothing downstream can be trusted. Then split that year's bunkers three ways, intra-EEA, into or out of the EEA, and wholly outside, because that split drives both European bills and most companies cannot produce it on demand. Next we open up the method behind a well-to-wake number: the ISO 14040 framework, the functional unit, and why changing the unit reverses the ranking of two fuels.