Fleet Intelligence

Fleet Decarbonization Tracker

Is the world fleet's carbon output bending down? The multi-year trajectory of total reported CO₂ — per segment and rolled up to the whole fleet — alongside the share of ships in the cleaner A/B/C efficiency bands and an EEXI-disclosure proxy, from verified emissions reports. The intensity-focused carbon page, extended into a trend over time.

94.0

Mt CO₂ reported · 2025

Fleet AnalyticsReporting years · 2023–2025

Reported CO₂

94.0

Mt · -4.9% YoY · 2025 provisional

Segments decarbonising

3/5

lower CO₂ than 2024 · 2025 provisional

EEXI disclosed

67%

report an attained EEXI

Ships reported

10,448

verified · 2025

On a like-for-like basis — ships reporting in both years — the reported fleet’s carbon fell 4.9% into 2025, with 3 of 5 segments cutting output; container shipping led at -10.0%, while gas carriers (+6.2%) and tankers (+0.8%) rose. (2025 still provisional.)

CO₂ trajectory

Total reported CO₂ (megatonnes) over the reporting years, per segment and rolled up to the whole reported fleet. A sum of mass — naturally robust to barely-sailed ships, so no clipping is needed. The headline decarbonisation question: does the line bend down? Note: 2025 is still accreting — EU-MRV emissions reports for a given year are not all filed until 30 April of the following year — so the rightmost point and its year-on-year figure are provisional and may revise upward. The final-year dip is not yet a confirmed bend-down.

Megatonnes of reported CO₂ per reporting year. The reporting scheme covers ships above 5,000 GT calling at EU ports, so this is the reported cohort’s output, not the entire world fleet’s.

Like-for-like carbon change · 2025

Year-on-year change in reported CO₂ measured over the ships that reported in BOTH years — the honest, cohort-fixed counterpart to the raw totals above, which also move as more ships come into the reporting scope. Below zero (green) = the segment’s carbon fell; above (red) = it rose. The whole-fleet bar is emphasised.

Bars are like-for-like — the year-on-year change over ships reporting in both years, so a widening reporting scope cannot masquerade as more carbon. The raw totals in the trajectory above also move with how many ships report each year.

For 2 of 5 segments, holding the cohort fixed reverses the direction the trajectory above shows. Other cargo’s raw total rose 11.2% (36.6 → 40.7 Mt) but its like-for-like output fell 2.0% — the rise is the reporting scope widening, not more carbon. The reverse also shows in tankers (raw -3.3% vs matched +0.8%).

Per-segment detail · 2025

Reported CO₂ and its year-on-year change (negative = the segment decarbonised), the A/B/C compliant share, the EEXI-disclosure proxy, reported sample size and coverage, and the count of non-reporting ships whose band was modelled.

Per-segment fleet decarbonization detail for 2025: reported CO₂, year-on-year change, A/B/C compliant share, EEXI-disclosure proxy, reported sample size, coverage and modelled ship count.
Segment	CO₂ (Mt)	YoY	A/B/C	EEXI disc.	Reported	Coverage	Modelled
Bulkers	14.20	-4.8%	65%	65%	3,650	27.5%	6,372
Tankers	6.69	+0.8%	65%	66%	769	15.8%	1,974
Container ships	30.12	-10.0%	65%	82%	1,374	40.8%	1,893
Gas carriers	2.31	+6.2%	65%	53%	116	37.8%	189
Other cargo	40.72	-2.0%	65%	64%	4,539	8.8%	0
Whole fleet	94.03	-4.9%	65%	67%	10,448	—	10,428

Efficiency-band mix by segment · 2025

The A–E carbon-intensity band split of reported ships, normalised to 100% within each segment so the mix is comparable across segments of very different size. A is the cleanest fifth of peers, E the least efficient. These are our own peer-relative ranking of the reported cohort — NOT an official carbon-intensity rating.

Bulkers3,650 reported · 28% of segment fleet

65% A/B/C

Tankers769 reported · 16% of segment fleet

65% A/B/C

Container ships1,374 reported · 41% of segment fleet

65% A/B/C

Gas carriers116 reported · 38% of segment fleet

65% A/B/C

Other cargo4,539 reported · 9% of segment fleet

65% A/B/C

Whole fleet8,648 ships

A (cleanest)BCDE (least efficient)

Modelled band mix (non-reporting fleet)

For ships that do not report emissions, the A–E band is estimated from segment, size and age. Shown separately from the reported mix above — never blended with it — to extend coverage beyond the reporting cohort. Point-in-time (attached to the latest reported year).

Bulkers6,372 modelled · 73% A/B/C

Tankers1,974 modelled · 33% A/B/C

Container ships1,893 modelled · 46% A/B/C

Gas carriers189 modelled · 33% A/B/C

Method & caveats

CO₂ trajectory is the sum of verified reported CO₂ mass per (segment, year), in megatonnes. Being a sum of mass it is naturally robust to barely-sailed ships, so no clipping is applied. Year-on-year change is computed against the same segment’s prior year; negative = the segment’s absolute carbon output fell.
The latest reporting year is a partial period — EU-MRV reports for a year are not all filed until 30 April of the following year, so the most recent year typically has fewer filed ships and a lower per-ship CO₂ than the complete prior years. Read its total and year-on-year figure as preliminary: the terminal-point dip may revise upward as more reports land and is not yet a confirmed bend-down.
The A/B/C top-3-band share and A–E bands use our own peer-relative carbon-intensity ranking, assigned on a fixed percentile split (A = best 15%, B 20%, C 30%, D 20%, E 15%). Because that split is fixed, A/B/C is ~65% by construction — a definitional artifact of the banding, not a measured compliance rate; it barely differentiates segments and does not move over time. This follows the A–E compliant-or-better idea (A/B/C compliant-or-better, D/E action plan) but is not an official carbon-intensity rating and should not be read as one. The denominator is banded ships only, so unrated ships never count as compliant.
EEXI disclosed is a disclosure proxy — the share of the cohort that publishes an attained EEXI value — not an EEXI-compliance figure. The required-vs-attained pass/fail is not published, so we can only honestly report how many ships have a published attained EEXI, not how many clear their reference line. This share is a snapshot of the currently-loaded EU-MRV report versions and is not comparable across reporting years: different years were imported from different THETIS report versions, so any apparent year-over-year rise in EEXI disclosure may reflect report-version coverage rather than a genuine change in how many ships disclose.
Coverage is reported ships as a share of the segment’s catalogued fleet. The reporting scheme only covers ships above 5,000 GT calling at EU ports, so coverage is thin for smaller-ship segments — this makes that explicit rather than passing the sample off as the whole fleet.
Modelled bands are estimated for the non-reporting in-distribution fleet and are kept in a separate panel so reported and modelled coverage are never silently merged. They are point-in-time (attached to the latest reported year).
Segments use the platform’s single shared vessel-class scheme (gas, tanker, container, bulker, other), so this can never drift from the valuation, fleet-supply, carbon-intensity or owner-stats views.

Snapshot computed 2026-08-27 16:27:38.046329+00.

Fleet Decarbonization Tracker

Segment

CO₂ (Mt)

YoY

A/B/C

EEXI disc.

Reported

Coverage

Modelled

Bulkers

14.20

-4.8%

65%

3,650

27.5%

6,372

Tankers

6.69

+0.8%

65%

66%

769

15.8%

1,974

Container ships

30.12

-10.0%

65%

82%

1,374

40.8%

1,893

Gas carriers

2.31

+6.2%

65%

53%

116

37.8%

189

Other cargo

40.72

-2.0%

65%

64%

4,539

8.8%

Whole fleet

94.03

-4.9%

65%

67%

10,448

—

10,428