Recommended Spares Lists, and why the supplier and the customer never quite agree
Every piece of capital equipment arrives with one. A list of parts the OEM says you should hold, with quantities, sometimes with prices, occasionally with a lead time. It lands alongside the manuals and the commissioning paperwork, and more often than not it starts an argument.
The maintenance manager thinks it is padded. The OEM thinks the customer is being reckless with a machine they have just spent two million pounds on. A number gets negotiated somewhere in the middle, both parties walk away mildly irritated, and nobody ever goes back to find out who was right.
That last part matters more than the argument itself.
What the supplier is actually doing
It is easy to assume the list is inflated on purpose. Mostly it isn't, and the reasons are more interesting than that.
The penalty for under-specifying is far worse than the penalty for over-specifying. If an OEM recommends a part that never gets used, nothing happens to them. If they omit a part and a line stops for six weeks waiting on a casting, that is a warranty escalation, a damaged reference site, an unhappy customer, and in some contracts liquidated damages. Faced with that asymmetry, any rational engineer rounds up.
The list is written before anyone knows how the machine will be used. Most Recommended Spares Lists are produced at design or quotation stage. At that point the OEM does not know whether you run one shift or three, whether the environment is washdown or clean and dry, how skilled your technicians are, or whether this line is the bottleneck or the backup. In the absence of that information, they assume the harshest plausible operating case. That is not padding, it is the only defensible position available to someone with no operating data.
Much of the list is derived from the Bill of Materials, not from failure evidence. Every wear item, every consumable, every long lead item, anything with a seal in it. Rule based, or inherited from the last similar build. Very few OEMs have a working feedback loop from field failures back into the recommendation, and that is not usually a lack of will. The failure data sits in the customer's CMMS, and it almost never travels back.
They know things the customer does not. Lead times inside their own supply chain, which sub-supplier is unreliable, which component is heading for end of life in three years, which item they will struggle to obtain at all once the current production run finishes. That knowledge is genuinely valuable. It also, almost without exception, fails to make it onto the list in any form the customer can see.
What the customer is actually seeing
Cash on a shelf. Spares are working capital. A hundred and fifty thousand pounds of initial spares against a two million pound machine is real money, competing directly against projects that would have moved the plant forward.
A storeroom full of evidence. Most MRO stores carry a substantial proportion of lines that have not moved in three years or more, and a recognisable share of that traces straight back to original spares packages. The customer has bought the recommended list before, several times, and can walk you down an aisle and show you the result. That history is the strongest argument they have, and it is difficult to counter with a spreadsheet.
A list with no reasoning attached to it. This is the real problem. A typical RSL is part number, description, quantity, price. No failure rate, no criticality assessment, no lead time, no explanation of why the answer is two rather than one. A recommendation with no rationale is indistinguishable from a sales order, so it gets treated like one.
Duplication the OEM cannot see. The customer may already hold the item for a sister machine. A sister site may hold it. The same commercial bearing may appear on three different OEMs' lists under three different part numbers, and get stocked three times.
Commodity pricing. Standard bearings, standard sensors, standard contactors, all at OEM part numbers and OEM prices. Customers increasingly know exactly what those components cost at source. Even where they represent a small share of the list value, they undermine the credibility of the whole document.
Parts that expire before the risk does. Elastomers, hoses, belts, batteries, electronic boards, greases. Some insurance spares degrade quietly on the shelf and are unusable by the time they are finally needed.
Both sides are right, and that is precisely the problem
Strip the commercial noise out and the position is this. The supplier is making a risk judgement without operating data. The customer is making a cash judgement without failure data. Each one holds exactly the half of the evidence the other one needs, and neither has any established mechanism for sharing it.
So the conversation collapses into the only dimension both parties fully understand, which is price. Quantities get halved, a discount gets applied, everyone signs. The actual question, which parts genuinely need to be on a shelf and why, never gets answered by either side.
What a defensible list looks like
It stops being a list and becomes a risk position, argued line by line:
consequence of failure, expressed as what stops, for how long, and at what cost per hour
likelihood, from a wear interval, an MTBF, or observed failures across the installed base
replenishment lead time, and whether a credible alternative source exists
whether the customer, or another site in the group, already holds the item
the cost of holding it, set against the cost of not holding it
Once each line carries that, the conversation changes shape entirely. Hold these. Accept the risk on these. Share or consign these across sites. Review the remainder in twelve months against what actually failed. Both parties can put their name to that. Neither party can reasonably put their name to a column of quantities with nothing behind it.
The other change worth making is to stop treating the list as a one time document. Twelve to twenty four months after commissioning, real consumption and real failure history exist for the first time. That is the moment the recommendation finally becomes worth something, and it is almost always the moment it gets filed and forgotten.
Where we come in
At Parts Control we sit on both sides of this. For OEMs, we build recommended spares lists from installed base and consumption evidence rather than engineering judgement alone, so the recommendation arrives with its reasoning attached and survives contact with a sceptical customer. For asset owners, we test those same lists against criticality, actual usage, lead times and what is already sitting on the shelf across the estate.
Same data, same model, both sides of the fence.