The world fleet is 22.2 years old and 12.6 years old, and the gap between those two numbers is the whole problem

MUSCAT — The commonly quoted figures for the age of the world fleet appear to contradict each other, and the apparent contradiction is the most useful fact in the subject. UNCTAD's figures, measured as at 1 January 2025, put the average age of the world fleet at 22.2 years by vessel count, up 1.8 per cent year on year, and 12.6 years by gross tonnage in 2024, up 3.2 per cent on 2023. The fleet was more than three years older in 2024 than a decade earlier. Both numbers are correct. They describe different fleets. By count, ships over twenty years old account for 41.5 per cent of developed-economy vessels and 39.9 per cent of developing-economy vessels. By deadweight tonnage, ships over twenty account for 9.3 per cent of developed-economy capacity and 21.1 per cent of developing-economy capacity — a developing-economy share more than double the developed one. Average age by economy group tells the same story: developed economies at 22.9 years by count and 11.8 by tonnage in 2025, developing economies at 21.3 and 13.9. ANALYSIS. The gap of 9.6 years between the count-weighted and tonnage-weighted averages is not a statistical artefact. It is the shape of the fleet. Carrying capacity sits in a relatively young population of large ships; vessel numbers sit in an old population of small ones. That matters because the three institutions bearing the consequences are each sized against a different one of those measures. Freight capacity, carbon exposure and newbuilding economics scale with tonnage. Survey attendance, certification workload, port State inspection and registry administration scale with hulls. The fleet is getting older far faster by the measure that drives institutional workload than by the measure that drives market commentary. What it means for owners: a base rate of nearly twelve to one ClassNK's port State control annual report, published in June 2026 on 2025 data, recorded 425 detentions and 1,752 detainable deficiencies across its fleet, up from 381 detentions in 2024 and 419 in 2023. The age gradient is severe. The detention ratio was 1.4 per cent for ships up to five years old, 11.1 per cent for ships over twenty and up to twenty-five, and 16.2 per cent for ships over twenty-five, the last accounting for 96 detentions. That is a ratio 11.6 times higher for the oldest band than the youngest, and 7.9 times higher for the twenty-to-twenty-five band — my arithmetic from the published figures. By ship type, reefer carriers carried the highest ratio at 10.9 per cent, while bulk carriers recorded the most detentions in absolute terms at 268. The detainable deficiency categories were fire safety at 426, life-saving appliances at 172, emergency systems at 169, ISM at 159, safety of navigation at 136 and water and weathertight conditions at 119, with ISM the most frequent single detainable item at 105 cases. The society's recommendations for older ships are tighter condition monitoring, faster defect close-out, more frequent checks of critical systems and realistic maintenance planning. One figure cuts against the age narrative and deserves noting: detentions among ships up to five years old rose from 10 in 2024 to 28 in 2025. The ratio remains low, but the direction is wrong, and it suggests the problem is not purely an artefact of age. ANALYSIS. The 11.6 multiple is not a prediction about any particular vessel. It is the base rate the system will apply to it, and the system applies it mechanically: through the Paris MoU risk profile, which takes age as an explicit parameter and sets inspection frequency accordingly; through the US targeting matrix, which scores flag and recognised organisation detention ratios; and through commercial vetting, which reads the resulting record. An owner of older tonnage is not fighting a perception. They are fighting a statistic, and the only instrument that moves it is evidence of a differential — which is precisely why the inspection-readiness differential RightShip reports for its own inspected vessels, at 2.0 deficiencies per inspection against 2.4 and a 2.4 per cent detention ratio against 3.6, is commercially interesting despite being vendor data describing a correlation rather than a proven cause. What it means for class: the work moves from documents to tanks The classification consequence of an ageing fleet is not more surveys. It is a change in the kind of survey. The IACS procedural requirement governing transfer of class sets entry survey scope by age band, and the escalation is steep. Under five years, the class entry survey is an annual survey. From five to ten years it is an annual survey plus representative ballast spaces. From ten to twenty years it is an annual survey plus representative ballast and cargo spaces, with exceptions for gas carriers and certain tankers. From fifteen to twenty years, for vessels falling under the relevant unified requirement, it is the special or intermediate survey, whichever is due. At twenty years and over it is a special survey. Thickness measurements accepted from a previous society must be within fifteen months for a special survey or eighteen for an intermediate where the entry survey is not credited as periodical, with confirmatory gauging required. Every one of those escalations converts documentation review into physical attendance: internal examination of spaces, thickness measurement witnessing, tank testing. RightShip's inspection data points the same way, with vessels aged 14 to 24 years accounting for the largest share of port State control inspections globally and bulk carriers alone for about 38 per cent. ANALYSIS. This is why the IMO sub-committee's work on coding remote survey eligibility item by item, endorsed in July, matters more than it appears to. If the fleet's age profile is pushing survey content towards work that requires a surveyor inside a space, then the only way to protect attendance capacity is to strip out of the programme everything that does not. The coding exercise is not a digitisation initiative; it is a triage mechanism for a scarce human resource, arriving at the moment the age profile makes that resource scarcest. Owners should expect the physical portion of survey cost and scheduling difficulty to rise even as the documentary portion falls in price — the two halves of a class fee are diverging. What it means for registries: an age cap is an act of self-defence Panama has closed its register to oil tankers and bulk carriers over fifteen years old, communicated through an update to the requirements of its pre-check process under marine notice MMN 11/2024, with Resolution 106-003-DGMM of 3 January 2025 adding quarterly inspections by the recognised organisation issuing statutory certificates and further verification of safety management certificates. The stated basis is the register's own data: its Directorate found that 71 per cent of detentions in its fleet from 2023 through the first half of 2025 involved bulk carriers, general cargo vessels and oil tankers older than fifteen years. ANALYSIS. An age cap looks like a quality measure and is actually a tax calculation. A register's detention ratio determines its standing on the Paris MoU performance lists and its detention ratio relative to the all-flag average, and that ratio feeds directly into the inspection exposure of every other ship on the register — seven points on the US targeting matrix where a flag's ratio is at least twice the all-flag average, which is by itself enough to reach Priority II. So admitting old tonnage imposes a cost on the register's existing clients, who did nothing to earn it. Once a register works that out, the cap follows, and the arithmetic is the same for every register with a reputation to protect. That makes the trend self-reinforcing rather than a matter of policy fashion. The consequence is segmentation. Older tonnage that cannot obtain a quality flag does not stop trading; it concentrates in registers that will take it, which is the mechanism by which the shadow fleet acquired its flag profile. The ageing of the fleet by vessel count therefore does not merely raise average risk — it sorts risk into particular registers, and the gap between the best and worst performing flags widens even if no individual ship changes condition. The replacement arithmetic does not currently work Recycling is running well below the rate the age profile implies, on any measure. UNCTAD recorded about 6.3 million gross tons sold for scrap in 2024, on a par with 2023 and about 0.25 per cent of the active global fleet by gross tonnage. Bulk carriers made up 32.4 per cent of it, container ships 15.2 per cent and offshore supply vessels 12.8 per cent, with Bangladesh, India, Pakistan and Türkiye together accounting for 91.7 per cent of the market. The fleet comprised about 112,500 vessels of at least 100 gross tons, with deadweight capacity growing 3.4 per cent in 2024 to around 2.44 billion, below the 5.1 per cent average for 2005 to 2024. BIMCO's projection, published in May 2023 and therefore now dated, was that more than 15,000 vessels totalling 600 million deadweight tons would be recycled between 2023 and 2032, against 7,780 ships and 285 million deadweight tons in the preceding decade. Its method applied the historical pattern that roughly half of bulk carriers, tankers and container ships are recycled by age 25 and about 90 per cent by 30 to 35 to the current fleet, with the 65 per cent growth in deadweight capacity during the 2010s implying rising recycling for ten to twenty years. ANALYSIS. Fifteen thousand ships across ten years is roughly 1,500 a year, about 1.33 per cent of a fleet of 112,500 by vessel count, against the 778 a year and 0.69 per cent implied by the previous decade. UNCTAD's 0.25 per cent figure is measured by gross tonnage and is not directly comparable to either. But on both bases the conclusion is the same and does not depend on reconciling them: actual recycling is running at a fraction of the rate the age distribution implies, which is the arithmetic reason the average age keeps rising. The age profile is not the product of a decision to keep old ships. It is the product of a decade in which freight markets rewarded keeping them and recycling prices did not compete. That calculus has, if anything, tightened this year. With second-hand values where they are — Korea Ocean Business Corporation assessments on 1 October put a five-year-old VLCC at $178.09 million, 37.6 per cent above the newbuilding price — and with a twenty-year-old VLCC reported sold at $117 million in October, the opportunity cost of recycling a trading vessel has rarely been higher. What to monitor UNCTAD's next Review of Maritime Transport for the 2026 age figures, which will show whether the count-weighted average has passed 23 years. ClassNK's and other societies' annual port State control reports for whether the under-five-year detention trend continues upward, which would complicate the age narrative. Further registry age caps, as the clearest sign the self-reinforcing dynamic above is spreading, and whether any register publishes its own detention data as Panama has. The remote survey eligibility coding work between IMO sub-committee sessions, as the measure of how class intends to protect attendance capacity. And recycling volumes against BIMCO's trajectory — the forecast is three years old and the market has moved decisively against it. The bottom line A fleet of 22.2 years by count and 12.6 by tonnage places the burden of ageing almost entirely on the institutions that work ship by ship rather than ton by ton. Port State control officers, class surveyors and registry administrators each deal in hulls. Freight analysts and carbon accountants deal in tonnes. The first group has been absorbing a problem that the second group's numbers barely register. For an owner, the practical consequence is that the cost of age now arrives through those institutions rather than through the market. The market is currently paying more for old ships than it has in years. The inspection regime, the class programme and the registries are simultaneously making them harder to run. Both of those things are true at once, and the second will outlast the first. ARK Management SPC — Marine Consultancy & Brokerage, Muscat, Sultanate of Oman.