Emissions Surveillance - Applied AI Foundation

Emissions Surveillance pipeline reading noon reports and bunker delivery notes into the five emissions regulatory regimes with year-end forecast

CO₂ is no longer a nice-to-track metric. Five regulatory regimes turn it into commercial consequence — directly through carbon allowances and indirectly through ratings that determine charter eligibility. The numbers come from a single source: noon reports plus bunkering records. Everything else is interpretation. The Emissions Surveillance pipeline reads the source data, runs every required calculation, projects the year-end position for each regime, and routes findings to the people who can act before the ratings lock in.

Where the data comes from

Source	What it provides
Noon reports	Daily distance, fuel consumption per grade, cargo on board
Bunker Delivery Notes (BDN)	Bunkered quantity, fuel grade, sulphur content
Voyage plan (ERP)	Port pairs, ECA exposure, scheduled distance
Cargo manifest	Cargo tonnes per voyage — required for EEOI
EU allowance market feed	Current EUA price for EU ETS forecast
IMO required CII tables	Reference required intensity per vessel type and DWT

The numbers come from a single internal source — noon reports plus bunkering records. Everything else (CII bands, EU allowance prices, FuelEU compliance line) is reference data the pipeline applies on top.

The five regimes

Regime	Who enforces	What’s measured	Consequence
CII	IMO	Annual carbon intensity (CO₂ per transport work)	A–E rating; D for 3 years or E for 1 year requires SEEMP III
EEOI	IMO (voluntary historically, increasingly mandatory)	Per-voyage CO₂ per cargo-tonne-mile	Operational benchmark
EU ETS	European Commission	CO₂ on EU-touching voyages	Allowance purchase obligation
FuelEU Maritime	European Commission	GHG intensity well-to-wake	Penalty per gram non-compliance
IMO DCS	IMO	Annual fuel consumption reporting	Compliance reporting only

Each regime uses the same source data but slices it differently — what counts as “emissions in scope” varies, and the metric definition varies. The pipeline computes each regime independently because they don’t reduce to a single number.

CII — the centerpiece

The IMO Carbon Intensity Indicator is the most operationally consequential of the five. The attained CII for a vessel:

\text{CII}_\text{attained} = \frac{\sum_{i} \text{CO}_2_i \text{ (g)}}{\text{Capacity} \cdot \sum_{i} \text{Distance}_i \text{ (nm)}}

where capacity is DWT (or GT for passenger vessels) and the sum is over all voyages in the operating year. CO₂ per fuel:

\text{CO}_2_i = m_\text{fuel}_i \cdot \text{CF}_\text{fuel}

with CF (carbon factor) values:

Fuel	CF (gCO₂/g fuel)
HSFO	3.114
VLSFO	3.151
MGO	3.206
LNG (combustion)	2.750

The required CII for the year,

\text{CII}_\text{required}

, is set by IMO per vessel type and DWT. The rating bands:

Rating	Threshold (relative to required)
A	$\leq 0.86 \cdot \text{CII}_\text{required}$
B	$0.86$ to $0.94$
C	$0.94$ to $1.06$
D	$1.06$ to $1.18$
E	$> 1.18 \cdot \text{CII}_\text{required}$

A vessel rated D for three consecutive years must update its SEEMP III with corrective measures. A vessel rated E for one year must do the same.

CII trajectory and forecast

The pipeline doesn’t just report the latest CII — it forecasts year-end. Two methods compose:

Method 1 — Linear projection

\text{CII}_\text{forecast} = \text{CII}_\text{attained}(t) + \dot{m}_\text{recent} \cdot (T - t)

where

\dot{m}_\text{recent}

is the trend over the last 3 months and

T

is year-end.

Method 2 — Operational profile

If the vessel’s voyage plan for the rest of the year is known, the analyzer projects emissions using planned distance and projected fuel consumption per leg. More accurate when the operational pattern is stable. The forecast is reported as a confidence range — a vessel currently at C-band attaining 1.05 with a forecast of 1.09 is heading into D territory, but the band depends on the residual operational pattern. Surfacing the range early (mid-year) gives time to act.

EEOI

The voyage-level energy-efficiency indicator:

\text{EEOI} = \frac{\sum_i \text{CO}_2_i}{\sum_i (m_\text{cargo}_i \cdot D_i)}

EEOI uses cargo-tonne-miles rather than capacity-miles, so it captures actual cargo carried rather than theoretical capacity. A vessel running below capacity (typical on backhauls) shows high EEOI relative to CII — useful signal that the operational pattern is the issue rather than the engineering. EEOI is computed per voyage and aggregated per quarter and per year for trending.

EU ETS

EU Emissions Trading System obligations cover:

100% of CO₂ on intra-EU voyages
50% of CO₂ on EU-extra-EU voyages (one end EU)
0% on extra-EU voyages

For each scope-relevant voyage:

\text{Allowances}_i = \text{CO}_2_i \cdot \text{Scope factor}_i

Aggregated across the year, multiplied by the EU allowance market price, produces the financial obligation. The pipeline computes the running total and projects year-end allowance demand so the commercial team can plan the EUA purchase or surrender position. A phased compliance applies in early years: 40% of obligation in 2024, 70% in 2025, 100% from 2026 onwards.

FuelEU Maritime

FuelEU is intensity-based and well-to-wake, so it captures the upstream emissions of the fuel as well as the combustion emissions. The regime uses GHG intensity (gCO₂eq/MJ) rather than CO₂ mass:

\text{GHG}_\text{intensity} = \frac{\sum_i \text{GHG}_i^\text{wtw}}{\sum_i E_i}

where

E_i

is energy content of fuel consumed in scope. Compared against a reducing target intensity (the “compliance line”) that drops year-on-year. Non-compliance is penalised per gram of GHG above the target, multiplied by total energy consumed. Renewable fuels (biofuels with ISCC certification) get a multiplier — they reduce GHG intensity proportionally to their share of energy mix, which can move a non-compliant vessel into compliance without operational changes.

IMO DCS

The annual fuel consumption reporting regime. Produces a single annual report per vessel with per-fuel-type consumption, distance, and hours under way. Less commercially consequential than the other regimes but a hard regulatory deadline. The pipeline tracks the data integrity for the DCS report — same data quality requirements that the consumption log validator checks before submission.

Fleet rating dashboard

The fleet view assembles every vessel’s CII rating in one table:

Vessel    Type           DWT      CII attained   CII required   Ratio   Rating
POSUN     Bulker         52,000   5.18           5.34           0.97    C
AQUILA    Tanker         115,000  3.65           3.74           0.98    C
OCEAN     Container      82,000   8.42           7.95           1.06    D
NEXUS     Tanker         58,000   4.21           4.45           0.95    C
…

A reviewer scans for the vessels approaching D / E — those need active management for the rest of the year. The fleet view also surfaces structural patterns: vessels of the same type running consistently above CII required points to a fleet-wide operational issue, not vessel-by-vessel.

Implementation reference

CII attained calculation, condensed:

def cii_attained(voyages, vessel_type, dwt):
    """CII attained = total CO2 / (capacity * total distance)."""
    total_co2 = 0.0
    total_distance_nm = 0.0

    for v in voyages:
        for fuel_type, mass_t in v["fuel_consumed"].items():
            cf = CARBON_FACTORS[fuel_type]            # gCO2 / g fuel
            total_co2 += mass_t * 1_000_000 * cf      # to grams
        total_distance_nm += v["distance_nm"]

    capacity = dwt if vessel_type != "passenger" else gt
    return total_co2 / (capacity * total_distance_nm)

CII rating from attained vs required:

def cii_rating(attained, required):
    ratio = attained / required
    if ratio <= 0.86: return "A"
    if ratio <= 0.94: return "B"
    if ratio <= 1.06: return "C"
    if ratio <= 1.18: return "D"
    return "E"

The CII trajectory analysis tracks attained through the operating year and projects year-end using the linear-projection method above.

Worked example

MV OCEAN, 82,000 DWT container, end-of-April review (4 months into operating year):

Metric	Value
YTD CO₂	14,820 tonnes
YTD distance	38,400 nm
YTD CII attained	8.92 (g CO₂ / DWT-nm × something)
CII required (2026)	7.95
YTD ratio	1.12 → D-band
Linear forecast	1.16 → D-band but close to E (1.18)
Operational forecast (with planned voyages)	1.18 → E-band threshold

EU ETS YTD obligation: 4,180 EUAs (~~€352k at €84/EUA). EU ETS year-end forecast: 12,500 EUAs (~~€1.05M). FuelEU GHG intensity YTD: 86.8 gCO₂eq/MJ vs target 89.3 → currently compliant by margin. Verdict: HIGH-CRITICAL on CII trajectory. The operational forecast crosses the E threshold. Three actions become candidates:

Slow-steaming: reducing average speed by 1 knot reduces CO₂ per nm meaningfully. Cross-references the voyage pipeline for CP impact (some CPs allow it).
Bunker mix shift: switching VLSFO portion to MGO improves combustion efficiency but increases mass slightly — net carbon impact is marginal. Better to look at LNG dual-fuel if available.
Operational efficiency: hull cleaning, propeller polishing — engineering changes that reduce required power for the same speed.

The pipeline:

Routes to the Operations / Commercial team — operational change candidates.
Routes to the ME performance pipeline for the engineering side (hull fouling? Engine SFOC?).
Generates the EU ETS purchase plan at the projected year-end demand.

What the senior review contains

Headline ratings — current and projected CII, EU ETS YTD and forecast, FuelEU compliance margin.
CII trajectory — chart with attained-through-time, projected year-end, rating bands shaded.
Per-voyage EEOI — operational efficiency metric for trending.
EU ETS obligation — running EUA demand, projected year-end, financial impact at current market.
FuelEU intensity — GHG intensity vs target, biofuel share if any.
IMO DCS data quality — completeness check before annual submission.
Operational change candidates — modelled impact of slow-steaming, bunker shift, hull-cleaning.
Recommendations — prioritised by impact and feasibility.
Escalation decision — to whom, and why.

Escalation triggers

Trigger	Severity
CII rating projecting E with $\geq 6$ months remaining	CRITICAL
CII rating projecting D for third consecutive year	CRITICAL
FuelEU intensity above target by mid-year	HIGH
EU ETS obligation forecast variance >25% from plan	HIGH
IMO DCS data integrity gaps in submitted reports	HIGH

Why all five together

A vessel can have A-rated CII and still fail FuelEU if its fuel mix is heavy. A vessel passing FuelEU can be commercially exposed on EU ETS price spikes. A clean IMO DCS report doesn’t help if the underlying CII rating is dropping. The five regimes operate independently and have to be tracked together — the pipeline runs all five computations on every refresh.

The single most consequential improvement most fleets can make on CII is hull cleaning frequency. A fouled hull adds 5–10% to the power required for the same speed; that translates directly to attained CII. The pipeline cross-references hull condition and propeller polishing dates so the operational levers are visible.

References

Source templates

Emission-monitoring suite — per-voyage EEOI, CII data with historical and forecast trajectory, and the fleet-wide CII rating view.

Microapp: Emissions

Live dashboard with all five regimes — the user-facing surface for the same data.

Related: Voyage

Voyage performance and slow-steaming decisions are CII-relevant.

Related: Fuel oil

Fuel mix decisions are emissions decisions — bunker grade affects CII, EU ETS, and FuelEU.

Documentation Index

​Where the data comes from

​The five regimes

​CII — the centerpiece

​CII trajectory and forecast

​Method 1 — Linear projection

​Method 2 — Operational profile

​EEOI

​EU ETS

​FuelEU Maritime

​IMO DCS

​Fleet rating dashboard

​Implementation reference

​Worked example

​What the senior review contains

​Escalation triggers

​Why all five together

​References

Source templates

Microapp: Emissions

Related: Voyage

Related: Fuel oil

Where the data comes from

The five regimes

CII — the centerpiece

CII trajectory and forecast

Method 1 — Linear projection

Method 2 — Operational profile

EEOI

EU ETS

FuelEU Maritime

IMO DCS

Fleet rating dashboard

Implementation reference

Worked example

What the senior review contains

Escalation triggers

Why all five together

References