Which Spare Parts Deserve a Place on the Shelf

The storeroom has plenty of low-cost sensors. The conveyor that several packing stations depend on has no confirmed replacement drive. Which gap deserves attention first?

The answer starts with the function that would be lost, the ways it could be restored and the evidence behind those options. Purchase price and past issue counts help, but neither tells the whole story. A rarely used part may protect a difficult recovery; a frequently used part may be easy to obtain.

Inspect a critical-spares decision record

Illustrative decision artifact. The packaging operation, asset identifiers, findings and approval below are fictional. They show how to make a stocking decision inspectable; they do not describe a client deployment, tested drive or measured saving.

In this example, drive failure would stop the shared conveyor feeding the packing stations. Purchasing has a replacement quotation. Maintenance knows a repair vendor and a possible spare at another site, but the remote unit has not been checked. Those are three leads to investigate, not three proven recovery routes.

The approval is deliberately narrow. It does not declare every conveyor drive interchangeable, establish a failure rate or claim that holding one spare prevents all interruption. If the assumed evidence cannot be supplied at a real site, that site's decision remains open.

Start with what stops when the part fails

Connect the part to the installed equipment and the work that depends on it. Identify the failure being considered, the affected production, possible workarounds and who can accept the resulting interruption. Separate a production inconvenience from a safety, environmental or product-quality requirement; a simple cost comparison cannot waive those requirements.

NASA's 2008 facilities RCM construction checklist connects spare-parts and material requirements with failure analysis and logistical issues. Its broader analysis starts with functions and failure consequences. Those are useful framing questions here, rather than a universal inventory formula or a classification of this example conveyor.

For C-01, the question is whether the packing operation can recover within the interruption its owner accepts. The sensors already on the shelf may still be justified; their presence does not answer the drive question. Do not remove other stock merely to make this example's purchase look affordable.

Record the installed population: the specific assets an approved spare can support, their configurations and where they operate. More installed units can create more calls on a shared spare, but common appearance is not evidence of interchangeability. A shared spare also cannot cover simultaneous demands without a plan for that situation.

If the maintenance approach itself is undecided, begin with How to Choose the Right Maintenance Strategy. This article addresses the narrower stocking choice once the failure and recovery needs are understood.

Check whether the spare can actually be used

Match the actual installed configuration to the approved replacement evidence. For a drive, the relevant checks may include its exact type and ratings, interfaces, configured options and applicable firmware or parameter requirements. The equipment manufacturer and qualified technical reviewer determine what matters for that installation; a matching catalog description alone is insufficient.

Storage requirements are product-specific. For example, the ABB ACS880-11 hardware manual, revision H describes capacitor preparation after an unpowered period and separately specifies storage conditions. That illustrates why an unused drive may need attention before service. It is not a procedure, interval or compatibility rule for the unnamed drive in CR-01; use the applicable current documentation for the exact item.

The remote unit in the example therefore stays unverified until its identity, condition and availability are established. Do not count the same unit as protection for two sites when its allocation is unresolved. A supplier's stock listing needs similar scrutiny: it is not necessarily stock reserved for you.

Replace the quoted lead time with a recovery path

A lead-time quotation is useful evidence, with a date and stated conditions. It is not the same as elapsed time until the required function is restored. Trace the route from recognizing the failure through authorization, obtaining the item, transport, preparation, installation and verification. Include waiting for people and access as well as the physical task.

For the repair option, ask when the clock starts, whether the quotation includes transport and diagnosis, and what happens if the unit is not repairable. An exchange arrangement may be a different offer from repairing the failed unit. Retain the actual supplier or repair terms and relevant order history; label estimates as estimates.

For a remote spare, include confirmation that it can be released, transported and received when needed. For a local spare, keep retrieval and authorized installation in view. Removing procurement delay does not remove every source of downtime.

Useful warning may change which route is practical, but do not assume a future failure will be predicted in time to buy a part. How Early Must a Maintenance Warning Arrive? examines the separate question of whether a finding leaves enough time for a useful response.

Compare holding, repair and supplier reservation

Put the alternatives against the same required function, accepted interruption and review period. Keep the finance-approved consequence basis and current purchasing evidence with the comparison. Where a failure likelihood or recovery duration cannot be supported, show the uncertainty rather than inventing an expected saving.

GSA's life-cycle perspective includes financial costs beyond initial purchase, such as installation, operation, maintenance and disposal. Applying that broad perspective does not supply a stocking quantity or prove which option is cheapest.

Ask finance to distinguish costs that would actually change between the options from costs that remain. Treat delayed production, lost output, recovery work and purchasing costs consistently so the same consequence is not counted twice. Show the evidence date, assumptions and unresolved ranges. Do not use an arbitrary revenue-per-hour figure as if every stopped hour creates that amount of permanent loss.

CR-01 selects one local replacement within its fictional scope. Choosing a quantity for a wider pool needs a further demand and replenishment review, including competing uses and the time before a used spare can be replaced. A simple count of historical failures does not establish the next failure date.

Keep the stocking rule current

Link the decision to the asset register, approved part reference and physical stock record. Maintenance or engineering owns technical applicability; stores owns the physical item and its condition record; purchasing owns supplier commitments; operations and the authorized budget owner approve the business choice. Name the actual people and review date when applying the record.

• When a spare is issued, show that the protection has been used and assign replenishment.
• When equipment, options or firmware change, recheck the approved compatibility scope.
• When a supplier's lead time, reservation or repair terms change, reopen the affected recovery route.
• When condition evidence expires or storage history is missing, withdraw the unsupported availability claim pending review.
• After a failure, compare actual waiting and recovery with the decision's assumptions.

A spreadsheet or an existing maintenance system may be enough for the first equipment family. Connecting asset, stores and purchasing references can make disagreements visible without replacing every system. AI can summarize referenced records, but it cannot establish compatibility, certify a stored item or approve purchase and physical work from a plausible description.

If the broader proposal is to accept failure of a low-consequence function, review When Is Run-to-Failure a Sensible Maintenance Choice? separately. Holding a spare does not itself justify that operating policy or establish uninterrupted redundancy.

Bring one failure that waited on a part

A useful starting point is a single equipment family and one stocking decision that people currently resolve from separate asset, stores and supplier records. Bring the failure history, installed configuration, physical stock findings and the terms behind each recovery option. Leave unsupported fields open.

A focused engagement with Adaptiv Stratum could connect those references into an asset-to-spares register and a reviewable decision record. The first scope, customer participation, technical reviewers and any later implementation are agreed for the actual situation. The purpose is a decision the team can explain and maintain, not a promised inventory reduction.

Discuss one critical-spares decision →