493 berths: the closest thing to a saturation diver headcount

There is no register of saturation divers

Ask how many saturation divers there are in the world and you will get answers that vary by an order of magnitude, none of them sourced. There is a structural reason for that. No government publishes a count. The certifying bodies audit companies and systems rather than keeping a public roll of individuals. Training schools know how many people they have put through a closed-bell course, but not how many of those are still working a decade later, and the crewing agencies that actually place divers treat their books as commercial property. There is no equivalent of an airline pilot licence register for this trade.

What does exist, published and checkable, is the accommodation. A saturation diver at work does not commute. He lives inside a pressurised system bolted into a ship, for the length of the job, and that system has a fixed number of living chambers with a fixed number of bunks in them. Operators advertise that number, because a charterer needs to know it before hiring the vessel. It is a physical constraint, expensive to change, and public.

So we counted the bunks. Every vessel in satdiver's dive support vessel directory that publishes a saturation diver capacity was added up, and the total is 493 berths across 28 ships. That is not a headcount of divers, and the rest of this post is largely about why not. But it is, as far as we can establish, the closest thing to one that the industry has produced.

What the 28 ships add up to

The directory holds 46 dive support vessels. Of those, 28 publish a diver capacity and the remaining 18 leave the field blank — blank in the directory means the operator does not publish it, not that the ship carries nobody. The 28 that do publish come to 493 berths between them.

The average of that group is 17.6 berths per vessel and the median is 18. Those two figures sitting almost on top of each other is the first useful thing the data says. When a mean and a median converge like that, there is no long tail pulling the average around: you are not looking at a handful of enormous ships and a crowd of small ones, you are looking at a fleet built repeatedly to roughly the same specification. The full spread runs from 6 berths at the bottom end to 24 at the top, a factor of four, but very little of the fleet actually lives at either extreme.

Eighteen is the number

Plot how many vessels publish each capacity and one bar makes the rest look incidental.

Fourteen vessels publish a capacity of exactly 18 divers — more than all the other published figures put together. When half a fleet independently arrives at the same number, that number is usually a consequence of how the work is organised rather than a coincidence of shipyard design.

The likeliest reading is straightforward, and it comes from how a saturation dive actually runs. A bell run puts two divers in the water with a third, the bellman, staying in the bell to tend their umbilicals and haul an injured man back inside. Three divers per run, then. A proper dive support vessel carries two bells, and offshore saturation work is round-the-clock — the whole economic argument for keeping people at pressure is that the system never has to stop. Stack enough three-man teams to cover a 24-hour cycle on two bells, with bodies left over for the days when someone is unfit to dive, and you land somewhere near eighteen. We cannot verify that arithmetic against the directory, because the directory records the capacity and not how operators divide it into watches. But eighteen is what the fleet keeps choosing, and a shift pattern is the explanation that fits.

If you want the underlying mechanics — the chambers, the bell, the decompression at the end — we set them out separately in what saturation diving is.

Twenty-four at the top, six at the bottom

Above the standard complement there is a second cluster, smaller but distinct, and it marks the practical ceiling of the published fleet.

Six vessels publish a capacity of 24 divers, and nothing in the directory publishes more. That is a meaningful boundary. It suggests the industry has a view about how large a single saturation spread usefully gets: past a certain point you are better off sending a second ship than building a bigger system into one hull, because the bells, the launch systems and the deck space become the constraint rather than the bunks.

Below the standard, the pattern thins out fast. Five vessels publish 12 berths, which reads like a single-system, single-bell configuration rather than a smaller version of the same thing. One publishes 16, one publishes 15, and one publishes 6 — the lowest figure in the directory and a genuine outlier, more likely a specialised or shallow-water system than a conventional offshore spread. Six berths cannot sustain a round-the-clock two-bell operation, so whatever that ship does, it is not doing it the way the eighteen-berth vessels do.

Source: satdiver.com dive support vessel directory, read 28 September 2026. 18 further vessels in the directory publish no diver capacity.

A berth is not a diver

Here is where the number has to be handled carefully, because it is easy to misread 493 as a population.

A berth is a place, and places are reused. Saturation crews work in rotations — a man joins the ship, does his trip, decompresses, goes home and is replaced by someone else while the vessel carries on working. Over a year, one berth is filled by several different individuals. That alone means the number of people qualified and available to occupy these 493 berths is larger than 493, and probably by a multiple rather than a margin. We cannot say what that multiple is. The directory publishes capacity, not rotation lengths, crew lists or how many trips a year a given diver takes, and no source we have would let us derive it honestly.

The error also runs the other way. A published capacity is a maximum, not an occupancy. A vessel in transit, in dry dock, between charters, or working a project that does not need saturation at all is carrying fewer divers than its brochure figure, sometimes none. On any given day, some fraction of the 493 is empty steel. The directory cannot tell us what fraction, and anyone who quotes you a utilisation rate for the saturation fleet is guessing.

What the figure is genuinely good for is a ceiling. Whatever the rotations, however many names sit on the agencies' books, no more than 493 saturation divers can be at pressure at sea at once on the vessels that publish a number — and the real simultaneous figure is lower. That is a harder constraint than any estimate of the total workforce, and it is the reason the trade behaves like a small one: a few hundred working positions worldwide, concentrated in the hands of a small group of operators. We looked at that concentration separately in the seventeen companies that actually run saturation, and it explains a good deal about why the money in the trade holds up the way commercial diver pay suggests it does.

The 18 ships that publish nothing

The largest single source of error in the 493 figure is not interpretation, it is silence. Eighteen of the 46 vessels in the directory publish no diver capacity — well over a third of the fleet. Those ships exist, they are in the directory because they are identifiable saturation vessels, and they have bunks. We just cannot count them.

The obvious thing to do with the gap is to assume the silent ships resemble the ones that talk. Apply the published average across all 46 vessels and the fleet comes to roughly 810 berths rather than 493. That is a projection and we are labelling it as one: it is the number you get from an assumption, not from counting anything. It could be wrong in either direction. Operators who publish a capacity may be the ones with a capacity worth advertising, which would make the missing ships smaller than average and 810 too high. Or the silence may reflect nothing more than a thin website, in which case 810 is about right.

We report both numbers deliberately. 493 is what the industry publishes. 810 is what the industry plausibly has. The honest version of the headline is that the working saturation population sits somewhere between those two figures multiplied by an unknown rotation factor, and anyone who gives you a single confident number for the size of this workforce has not explained where it came from. The age profile of these same hulls, which is a better-documented question, is in our audit of how old the saturation fleet is.

How we counted this

The figures come from satdiver's own dive support vessel directory, read on 28 September 2026. The directory holds 46 vessels. For each one we record a saturation diver capacity where the operator or the vessel's own specification sheet publishes one; 28 vessels have that field filled and 18 do not. We summed the 28 published values to get 493 berths, and took the mean (17.6), median (18), maximum (24) and minimum (6) of the same 28 values. The 810 figure is not a count: it is the published mean applied to all 46 vessels in the directory, and it should be read as an illustration of the size of the gap rather than a measurement.

The limits are worth stating in order of how much they matter. First, this is a directory of what operators publish, not a census of the world's saturation fleet, and we make no claim to have found every saturation-capable ship afloat. Second, the 18 blanks are blanks, not zeros, and they are a large enough share of the fleet that any total derived from the other 28 is a floor. Third, a published capacity is a design figure and we have no way to confirm it against a crew list, a manifest or a class certificate — if an operator's website is out of date, so are we. Fourth, as set out above, the boundary between divers and diving-support personnel inside a published figure is not visible to us. Fifth, capacity says nothing about whether a ship is working, and a directory entry is not evidence that a vessel is in service today.

What would improve this count is not more cleverness with the same data. It is the 18 operators who currently publish nothing deciding to publish something.

Frequently asked questions