A fleet that is stranding vessels on missing parts and tying up capital in slow-moving stock does not have one inventory problem that has been got wrong in two directions. It has two different problems being managed with a single policy. Consumables are a consumption problem and respond to levels and reorder points. Critical spares are a risk problem and respond to consequence and lead time, with consumption barely entering into it. Applying one model to both produces exactly the symptoms in the brief: empty where it hurts, full where it does not.
Two inventories, two opposite questions
Why one policy produces both failures at once
Consumables are forgiving. Usage is reasonably steady, the item is cheap, and a reorder point calculated from consumption and lead time will be roughly right. Run that same logic over a critical spare and it fails in a specific way: the item has almost no consumption history, because it is the part that has not failed yet. A model driven by usage concludes, correctly on its own terms, that the stock is not needed.
The reverse error is just as common. A fleet frightened by one stranding starts holding everything on board, and capital goes into items with redundancy behind them, short lead times, or shelf lives that will expire before use. The result is a store that is expensive and still missing the part that matters.
What makes a spare critical
The critical spares list is not a procurement document and should not be written from a maker’s recommended list. It is the output of a failure analysis you are already expected to carry out, because the identification of equipment whose sudden failure may produce a hazardous situation is a safety-management obligation in its own right. The spares follow from that list rather than the other way round.
Find the single point failures
Work from system single-line diagrams and record the possible outcome of each individual component failure. A single point failure is any one failure that can lead on its own to a hazardous situation.Run the hazard study with more than one discipline in the room
What hazards exist, what could cause or escalate them, how likely they are, what the consequences would be, and whether those consequences are tolerable. An engineer alone will under-rate the operational consequences and an operator alone will under-rate the technical ones.Walk the failure path to the end
Is there redundancy, is it segregated, is there another way of operating, can the repair be done in the conditions the vessel will be in, can the crew actually do it, and are the tools, drawings, spares and remote support there? If the failure still reaches a hazard, the equipment is critical.Judge likelihood from your own fleet first
Fleet experience, then wider industry experience, then credible events with no recorded history. Your own failures are the best evidence you have and the most frequently ignored.Count standby units in
A standby unit that has to take over immediately is critical. It is not a spare sitting on a shelf; it is equipment on which something depends the moment it is called for.Exclude the large items deliberately, not by default
Nobody carries a spare crankshaft. Leaving a large component out is defensible when the risk assessment says so and indefensible when it simply never came up.Deciding where a spare is held
Once an item is on the list, the question is not whether to hold it but where. Three factors settle it, and the balance between them is what keeps capital proportionate.
| Consequence of failure | Lead time | Real redundancy | Where it belongs |
|---|---|---|---|
| Hazard to people or environment | Any | None, or shares a space with the duty unit | On board, in a known location, with its documents |
| Hazard to people or environment | Short and reliable to the trade | Segregated and proven | Fleet store, with a named dispatch route |
| Vessel stopped, no hazard | Long, or a minimum order quantity applies | None | Fleet store — one holding covering several vessels |
| Vessel stopped, no hazard | Short | Another way of operating exists | Supplier, ordered on demand |
| Performance or comfort only | Any | Any | Consumable policy — reorder point, no special treatment |
| Any, with a shelf life shorter than the holding period | Any | Any | Supplier or fleet store on rotation — never aboard to expire |
The pattern worth noticing is that a long lead time pushes an item toward a central holding rather than toward every vessel. One part in a fleet store, with a dispatch route that has been used and timed, protects more hulls per dollar than the same part sitting in eight engine rooms. The machinery that generates most of these decisions is covered in our guides to main engine overhaul scheduling and auxiliary engine and generator maintenance.
The lead time that matters is not the one you were quoted
Every holding decision above rests on a lead time, and the figure most fleets use is the one the supplier gives for despatch from their works. The number that actually governs is door to deck: from the moment somebody on board decides the part is needed to the moment it is in the engineer’s hands on a vessel that has been moving the whole time.
Two of those five steps have nothing to do with the supplier, and they are where most of the time goes. A part quoted at two weeks and delivered in nine is not a supplier problem; it is an approval step, a trade pattern and a customs regime that nobody measured. The practical fix is to record the real elapsed time for every critical part movement and use those figures rather than the quoted ones — and you can see in a short demo how that history is kept against each item.
Measuring it also changes which items need holding at all. Where the real door-to-deck time is short and dependable on a fixed trade, a central holding will do. Where the vessel is tramping and clearance cannot be relied on, the same part has to be on board, whatever the catalogue says about availability.
Redundancy that is not redundancy
This is the assumption that quietly removes items from a critical list and should not. Two units are only redundant if a single event cannot take both. Where redundant equipment shares a compartment that could be lost to one fire, one explosion or one flood, the redundancy exists on the drawing and not in the casualty.
Why the stock figure is wrong
Most fleets discover their inventory is inaccurate at the worst possible moment. The causes are mundane and all four are recording failures rather than theft or carelessness.
The first is the one that compounds. If consumption is not recorded where the work happens, every figure downstream drifts, and the drift is always in the same direction — the system believes you have more than you do. That is the same mechanism that undermines a maintenance record, and the fix is the same: the issue of a part is logged as part of doing the job, not reconciled afterwards. Our guide to planned maintenance systems covers the job side of that record.
What the records have to show
A critical spares regime is open to audit, and the documents are specific. A fleet manager who can produce these has a defensible position; one who has the stock but not the reasoning does not.
The risk assessment behind the list
The critical equipment list or matrix
Time-limited assessments for deferred work
Firmware revision records
Stock checks and failure reports
Freeing capital without creating exposure
The reduction exercise is worth doing and is usually run the wrong way, by value. Starting with the most expensive items guarantees an argument with the technical department, because expensive and critical overlap heavily. Starting from the other end is quieter and releases more.
What is left after those four passes is close to the inventory the fleet actually needs, and the argument about it becomes much shorter. You can book a demo to see the critical list and the stock position held against each other.
