Spare Parts Stock Risk Worksheet

Use this worksheet for consumables, repairable assemblies, electronics, valves, seals, bearings and other components whose absence can delay restoration. The output is a documented stocking decision and an action register. It may justify local stock, shared stock, a repair loop, an agreed supply arrangement, redesign or no dedicated stock. The calculation does not establish supplier inventory, a universal stock quantity or guaranteed availability.

1 Start with the function that the spare restores

A part fitted to a large compressor is not automatically more important than a small component in a protected control path. The relevant question is which function becomes unavailable and how long the process can tolerate that condition.

Write the failure mode and restoration action. Replacement of an oil filter during planned service differs from replacement after contamination or mechanical damage. A controller fault may require configuration, firmware and commissioning as well as hardware. A valve kit may be useful only when the body and mating surfaces remain serviceable. Stocking decisions should correspond to a credible repair action rather than an unqualified parts list.

Document any alternate service that protects the process while the component is unavailable. A functioning standby unit can reduce immediate restoration urgency but may leave the plant exposed to another failure. The period of reduced resilience still has a consequence. Record both the current function and the additional risk created while the repair is outstanding.

2 Separate a spare part from a maintenance consumable

Scheduled consumables are driven by the maintenance plan, operating hours, condition and the correct equipment configuration. Breakdown spares are driven by uncertain demand and restoration consequences. Repairable assemblies circulate through service, removal, repair and return. These populations need different inventory rules even when the procurement system stores them in the same catalogue.

For scheduled work, the principal question is whether the correct quantity is available by the job date with suitable remaining life. For an insurance spare, long periods without issue can be normal because the purpose is to protect against a rare high-consequence event. For repairables, stock availability depends on repair turnaround and the condition of items in the loop, not just purchasing lead time.

Classify each item before calculating stock. A filter replacement planned for next month should not be treated as random breakdown demand. A planned overhaul can consume several components simultaneously, invalidating a model that assumes independent single-item requests. Keep known demand separate from uncertain demand so the worksheet does not hide an upcoming commitment.

3 Define the part identity precisely

Capture manufacturer, equipment model, serial number, installed configuration, part number, revision and quantity per assembly. Where available, link a controlled bill of materials, drawing, identification photograph and approved supplier reference. A description such as separator element or solenoid valve is insufficient to establish compatibility.

Record electrical ratings, connection details, materials, dimensions, pressure and temperature suitability, signal type and firmware requirements where they affect fit or function. These are item-specific fields; do not infer them from a visually similar component. A nominally matching thread or footprint does not prove that a replacement meets the duty.

Treat cross-references and supersessions as evidence requiring review. A supplier's proposed equivalent may need an engineering assessment, while an OEM supersession may include additional parts or configuration changes. Preserve the original identifier and the approved replacement relationship so technicians can understand why the stock item differs from the removed component.

5 Establish the actual installed population

Count the units that can demand the part during the planning horizon. Distinguish equipment in service, standby, storage, retirement and temporary deployment. A purchasing record showing ten compressors does not prove that ten identical component configurations remain installed.

Review serial ranges and modifications. A fleet described by one model name may contain different revisions, drive arrangements, control systems or treatment components. Pooling demand across incompatible variants can produce an apparently adequate stock quantity that is unusable for the machine that fails.

Record the population date and the source of the count. When an asset is replaced or modified, update both the equipment record and the spare applicability. Retired equipment can leave valuable-looking but obsolete stock. New equipment can create a coverage gap before the maintenance catalogue is updated.

6 Obtain demand history with its context

Collect issue dates, quantities, equipment identifiers, work-order reasons, operating exposure and the fate of the removed item. Separate planned service, breakdown replacement, trial substitution, installation damage and stock correction. These events represent different demand mechanisms.

A part issued from stores is not always consumed. It may be returned unused, transferred to another site or held in a maintenance cabinet. Conversely, emergency purchases may bypass the normal issue system. Reconcile these pathways before estimating demand, otherwise the model can overstate or understate actual usage.

Check whether past demand represents the future operating regime. Increased hours, harsher ambient conditions, changed fluids, a new process or an ageing population can invalidate a historical average. Record the reason for using a history window and the circumstances that would trigger a revised estimate.

7 Treat zero demand history carefully

No recorded issue can mean a reliable component, missing records, a new installation, unused standby equipment or a failure mode that has not yet occurred. It is not proof of zero future demand. For a rare high-consequence item, the stocking decision may depend more on recovery feasibility and obsolescence than on a short consumption history.

Where data are sparse, use scenarios rather than a falsely precise failure rate. Compare the consequences of no local stock, one verified item and an alternative service arrangement. State the evidence supporting the scenario and the uncertainty that remains. A sensitivity range should come from a defensible engineering judgment or data source, not a universal default hidden in the calculator.

Identify the information worth collecting. Operating exposure, failed-part examination and repair records can improve future decisions. The worksheet should preserve why an insurance spare was selected so that several years without demand do not automatically lead to its disposal.

8 Define replenishment lead time from need to usable stock

Lead time includes more than a supplier's quoted dispatch interval. It may include identification, technical approval, purchase authorization, manufacturing, transport, customs where relevant, receiving inspection, configuration and release to usable stock. Record the stages that apply to the actual procurement route.

Distinguish quoted, observed and contractually committed durations. An old quotation is not evidence of present stock or current delivery. An emergency purchase from a different source may have different inspection and compatibility requirements. Capture the quotation date, scope, validity and assumptions when using a commercial input.

For repairables, record the return loop separately: removal, shipment, diagnosis, quotation approval, repair, testing, return and receiving acceptance. A nominal workshop repair time can represent only one stage. The plant's exposure persists until an acceptable unit is available for installation or service.

9 Build a restoration timeline

Restoration with local stock still requires diagnosis, isolation, removal, installation, configuration, testing and process release. The benefit of stock is the part of restoration delay that the stock actually removes.

Some activities can overlap. Procurement may begin while diagnosis continues, but uncertainty about the failed component can create the risk of ordering the wrong item. Configuration preparation may run while a controller is transported, provided the correct hardware and software basis is known. Record dependencies rather than assuming every stage is sequential or simultaneous.

If the installation task itself exceeds the process tolerance, stocking alone cannot meet the requirement. Review alternate service, modular replacement, prepared connections or another recovery strategy. The worksheet should identify the dominant delay rather than recommending additional inventory for a problem inventory cannot solve.

10 Distinguish physical stock from usable stock

Physical quantity includes everything found in stores. Usable quantity includes only items with confirmed identity, compatibility, condition, preservation and release status. Reserved items are not freely available for a different job. Quarantined or expired items must not be counted as ready coverage.

Record location and accessibility. A verified spare at another site may require transport, release authority and receiving checks before use. A spare in a locked store may be unavailable outside staffed hours unless an authorized access arrangement exists. These practical details determine whether the stock supports the stated recovery objective.

Keep the inventory record tied to physical verification. Cycle counts should check condition and identification as well as quantity. An unopened box with an unclear label is not equivalent to a verified part, and an inventory system entry is not proof that the item remains on the shelf.

11 Protect parts during storage

Preservation requirements depend on the component, packaging, materials and manufacturer instructions. Bearings, elastomers, electronics, lubricants and assembled rotating equipment can have different storage conditions and inspection needs. Do not apply one universal shelf life or inspection interval to all compressor spares.

Record the manufacturer's storage instructions, receipt date, preservation actions and next review where applicable. Protect identification and packaging from damage. If packaging is opened for inspection, record the action and any required re-preservation. A component that appears unused can still be unsuitable after poor storage.

SKF's bearing guidance identifies the value of intact original packaging and appropriate stock rotation for its products. That does not establish a universal lifetime for every bearing or other component. Use the applicable manufacturer guidance for the specific stored item and keep any date-sensitive limits in the inventory record.

12 Address electronics and configuration dependencies

A controller, drive or sensor can require configuration before it restores the machine. Record approved hardware revision, firmware, parameter backup, licenses where relevant, communication settings and the method for restoring the configuration. The spare-parts strategy should include the information and tools needed for the authorized task.

Maintain controlled backups and confirm that they correspond to the actual installed configuration. A backup from a superficially similar machine can contain unsuitable settings. Configuration changes should follow the site's control and cybersecurity arrangements. Do not treat a spare electronic module as permission to bypass protection or install unverified software.

Test the recovery procedure in a suitable approved environment. A stored module may have battery, firmware or obsolescence issues that are not visible from the box. Record the scope and date of verification rather than assuming that a purchase record proves operational readiness.

13 Consider repairables as a circulating population

For a repairable assembly, distinguish installed units, ready spares, units awaiting shipment, units under repair, units awaiting acceptance and units rejected beyond economic repair. These states determine the stock available during the next demand.

The repair loop can contain multiple delays and uncertain outcomes. A unit sent for repair may be condemned or require unavailable components. A repair quotation does not establish a return date or accepted performance. Include repair yield and turnaround evidence where available, and use explicit scenarios when evidence is weak.

Decide how a removed unit becomes a ready spare again. Acceptance should match its function and the scope of repair. For an airend or other major assembly, link to the relevant controlled inspection and post-rebuild validation pages.

14 Evaluate shared stock honestly

Shared stock can reduce duplicate inventory where compatible demand is sufficiently independent and transport fits the recovery objective. Confirm the equipment applicability, ownership, release authority, replenishment responsibility and priority rule when two sites request the same item.

Pooling is less effective when common events create simultaneous demand. A shared maintenance shutdown, contaminated lubricant batch, environmental event or common design issue can defeat an independence assumption. Record these scenarios separately instead of letting a simple average justify one spare for every site.

Compare actual time to usable stock, not map distance alone. Packaging, collection, transport, site entry and inspection all matter. A shared spare may be a sound second layer of protection while local stock covers the shortest recovery requirement. The worksheet should state which role the arrangement fulfils.

15 Review supplier and repair options

Compare approved supply routes by compatibility evidence, lead-time evidence, acceptance requirements, commercial conditions and continuity risk. Lowest unit price can be outweighed by a long or uncertain restoration delay. Equally, a high-consequence label does not justify accepting unverified stock or bypassing technical identification.

Record whether a supply arrangement is purchase-to-order, consignment, reserved inventory, exchange, repair or ordinary catalogue availability. These terms can imply different rights and obligations. Only an actual agreement establishes whether a specific item is reserved or available to the user.

The public worksheet can help collect a precise requirement while the commercial response establishes actual scope and availability.

16 Treat obsolescence as a separate exposure

A component can have low demand and high obsolescence risk. Manufacturer discontinuation, unavailable firmware, loss of a repair capability or withdrawal of compatible materials can change the restoration strategy before a failure occurs. Record the evidence and date of any obsolescence notice.

Possible responses include a justified lifetime purchase, approved substitution, planned retrofit, exchange arrangement or equipment replacement. Each has costs and technical dependencies. A lifetime purchase can create preservation and unused-stock risk; a retrofit may require shutdown, integration and validation.

The worksheet should compare the remaining equipment service horizon with supply continuity. Avoid buying a large quantity solely because a supplier reports limited stock. Confirm applicability and the organization's asset plan before converting an obsolescence concern into an irreversible inventory commitment.

17 Use a decision matrix before a stock formula

A formula can support quantity selection after the service and demand basis are established. It cannot determine compatibility, acceptable consequence or a credible supply route. First classify the decision using the evidence below.

Record the chosen strategy in words before reporting a quantity. A decision such as retain one locally verified exchange assembly and replenish immediately after issue is more informative than a bare minimum-stock number. It identifies the operating rule as well as the stock level.

18 Calculate inventory position consistently

For a simple purchasing policy, inventory position can be defined as usable on-hand stock plus confirmed usable inbound supply minus committed demand and backorders. The exact definition must match the organization's system. Do not include uncertain inbound supply as though it were available today.

Separate inventory position from immediate coverage. A purchase order due next month can improve the position used for replenishment decisions while offering no protection against today's failure. Show available now, reserved, inbound and unavailable quantities separately in the worksheet.

For a planned maintenance job, time-phase the commitments. Stock adequate at the start of a month may be insufficient after several scheduled services consume it. The decision should consider demand and arrivals in date order rather than relying on one static total.

19 Calculate deterministic demand during lead time

Where consumption is sufficiently regular and the approximation is justified, expected demand during lead time equals average consumption rate multiplied by lead time expressed in consistent units. Add known scheduled demand separately if it is not already included in the rate. Do not double count the same maintenance activity.

The buffer should reflect demand and lead-time variability, service consequences and the selected policy. If those inputs are unknown, show that uncertainty rather than filling it with a default.

Deterministic arithmetic is useful for checking a policy, not proving its performance under irregular breakdown demand. Compare it with the actual sequence of historical issues and replenishments where data permit. An average that looks adequate can conceal a cluster of demand during a long replenishment interval.

20 Use a Poisson model only when its assumptions fit

For an illustrative demand model, let the mean number of independent demands during replenishment exposure be mu. Under a Poisson count model, the probability of k demands is exp(-mu) times mu to the power k divided by k factorial. The probability of no more than s demands is the sum of those probabilities from zero through s.

This model assumes a suitable stationary independent count process for the exposure being analysed. Planned campaigns, common-cause failures, ageing trends and batch replacements may violate the assumptions. A failure-rate estimate derived from a small or mixed data set can make the result misleading even when the formula is correct.

If s usable items are available at the start of a fixed exposure with no replenishment during that interval, the model probability of demand exceeding stock is one minus that cumulative probability. This is a scenario probability for the defined interval, not an annual availability guarantee or a universal customer service level. Repeated replenishment policies and repair loops require their own models.

21 Keep service probability and fill rate distinct

The probability of no stockout during an interval asks whether all interval demand can be met from available stock. Fill rate asks what fraction of demanded units is supplied under a defined policy. These are different quantities and should not share an ambiguous label such as service percentage.

A high probability of at least one demand does not by itself establish a need for many spares. The count distribution, consequence of shortage, replenishment arrangement and recovery alternatives all matter. Similarly, a high fraction of units supplied can coexist with a severe consequence from the one unavailable item.

Name the metric in the result and explain the exposure period and assumptions. If the user has not selected a target based on the business requirement, the worksheet should compare scenarios rather than choosing a hidden universal target. The decision remains with the accountable owner.

22 Compare lifecycle costs without hiding consequences

Inventory costs can include purchase, capital tied up, storage, preservation, inspection, insurance where applicable, obsolescence and disposal. Recovery costs can include expedited supply, repair, overtime, alternative service and production consequence. Include only relevant documented categories and avoid treating every accounting allocation as an incremental cash cost.

Use a consistent time horizon and currency basis. A one-year carrying cost should not be compared directly with a multiyear avoided-loss estimate. Distinguish costs paid once from recurring costs and explain any residual value. Where discounted cash flow is appropriate, use the organization's approved basis and link the detailed economic method to its controlled page.

Safety and other unacceptable consequences should not be traded away through a simplistic expected-cost average. The worksheet can identify economically attractive choices among acceptable alternatives. An option that fails a mandatory process requirement remains unsuitable even if its average modeled cost is low.

23 Test sensitivity before making a purchase decision

Vary inputs that materially influence the result: lead time, demand exposure, repair success, process loss and remaining equipment life. Use values supported by evidence or clearly identified planning scenarios. Report which assumption changes the preferred strategy.

Sensitivity analysis is particularly useful when one quotation or one failure drives the apparent answer. If a small change reverses the recommendation, the decision needs better evidence or a robust contingency. Avoid presenting the midpoint as certain simply because a calculator returns a precise number.

Record what new information would change the decision. A verified shorter lead time, approved interchangeable component or planned equipment replacement can justify reducing stock. Increased exposure, longer repair queues or a new protected process can justify increasing it. This makes future review purposeful.

24 Worked case A with planned demand and inventory position

This case uses illustrative values, not supplier stock or a recommended policy. A site has eight verified filters on hand. Three are reserved for a scheduled service. Four confirmed units are due before the next planning review, and one unit is backordered to an outstanding job. Under the stated definition, inventory position is 8 plus 4 minus 3 minus 1, giving eight units.

Immediate uncommitted coverage is different: eight on hand minus three reserved minus one outstanding commitment gives four units, assuming the backorder represents demand not otherwise included in the reservation. The inbound four do not protect an immediate breakdown before arrival. The worksheet must prevent the two outputs from being confused.

If the scheduled job actually requires five filters rather than three, the position and immediate coverage both fall by two. The cause is a changed commitment, not a changed random failure rate. Correct the maintenance bill of materials before applying a statistical buffer.

25 Worked case B with lead time exposure

Assume an illustrative regular demand of two units per month and a three-month replenishment interval. Mean demand during that interval is six units. A separate planned shutdown requires four additional units and was not included in the monthly rate. The total expected requirement over that defined horizon is ten units before any justified uncertainty allowance.

If the four-unit shutdown was already represented in the historical rate used to estimate two units per month, adding it again would double count part of demand. The analyst should instead construct the actual forward schedule or adjust the history basis. The correct approach depends on what the rate represents.

Now suppose observed end-to-end lead time is five months rather than the quoted three-month manufacturing period. Rate-based demand becomes ten units, and the separate shutdown brings the deterministic total to fourteen. This example shows why receipt and release lead time matters more than one quoted stage.

26 Worked case C with an explicit count model

Assume a hypothetical compatible fleet generates a mean of 0.4 independent demands during a fixed replenishment exposure. Under the stated Poisson model, the probability of zero demands is exp(-0.4), approximately 0.67032. The probability of one demand is 0.26813. With one usable item initially available and no replenishment during the interval, the probability of more than one demand is approximately 0.06155, or 6.16 percent.

With two items, include the probability of two demands, approximately 0.05363. The probability of demand exceeding two falls to approximately 0.00793, or 0.79 percent. The reduction is a result of this example's inputs and assumptions, not a recommended stock level or guaranteed service performance.

The decision must still consider consequence, model fit, repair options and preservation cost. If several machines share a failure cause, independent demand is not a sound assumption. Keep a separate common-event scenario rather than using the reassuring probability to dismiss it.

27 Worked case D with stock that is not usable

A hypothetical store lists three identical-looking controllers. One is verified for the installed hardware and firmware. One has an unverified revision. One is reserved for a scheduled retrofit. Physical quantity is three, but immediate verified uncommitted quantity is one.

An inventory report that counts all three can conceal a recovery gap. The unverified unit needs a compatibility decision and any required configuration work. The reserved unit needs an explicit release decision before it can support an unrelated breakdown, and that release may create a new project delay.

No statistical calculation corrects this identity problem. First establish usable stock and commitments. Then model the exposure. The worksheet should show why each excluded item does not contribute to current coverage and who owns the action required to change its status.

28 Worked case E with a repair loop

Assume a hypothetical site has one ready exchange assembly and one failed assembly awaiting repair. The repair shop quotes ten working days for repair after approval, but shipment, inspection, quotation approval, return transport and site acceptance add further stages. The ten-day quote is not the total interval during which the site has no second ready spare.

A timeline review finds that approval is the largest avoidable delay. Preparing the commercial and technical decision process may reduce exposure more effectively than buying a second assembly immediately. The alternatives still need comparison against the consequence of a second failure while the first unit is away.

If the failed unit is condemned, the repair-loop assumption changes to replacement procurement. Record that branch in the scenario. A model that assumes every removed assembly returns serviceable will understate exposure where repair yield is uncertain.

29 Worked case F with a limited economic comparison

For illustration, a candidate spare costs 80,000 currency units. An approved annual carrying and preservation estimate is 12 percent of purchase cost, or 9,600 per year. A planning scenario estimates that having the spare could avoid 30,000 of incremental disruption cost in a year. On that restricted annual comparison, the modeled benefit exceeds the carrying estimate by 20,400.

That is not a payback calculation and does not by itself recover the initial purchase price. It also does not establish the probability or value of the avoided event. A full investment comparison needs the purchase outlay, horizon, residual value, alternative arrangements and the evidence behind disruption cost.

If the equipment will retire shortly, obsolescence may dominate. If the spare cannot be installed within the process tolerance, the assumed benefit may be overstated. Keep the arithmetic visible and explain what it does not establish rather than turning a favourable difference into a universal stocking recommendation.

30 Diagnose recurring stock problems

When stockouts persist despite a nominal minimum quantity, compare physical issues with the replenishment trigger and actual lead time. Delayed reorder, hidden reservations, inaccurate demand classification or unusable items can explain the result. Increasing the minimum without finding the mechanism can leave the same failure in place.

When excess stock accumulates, examine retired equipment, duplicated part numbers, unapproved equivalents and maintenance schedules that changed. A slow-moving item may be justified insurance stock, but that justification should be documented. Consumption alone cannot distinguish prudent insurance from obsolete inventory.

A corrective action should identify its owner and completion evidence. A catalogue edit, purchase order or supplier email is not automatically proof that the recovery risk has been resolved. Recheck the service consequence after the action is implemented.

31 Required spares prioritisation worksheet

The required asset is a Spares prioritisation worksheet. Use one record for each part and materially different compatibility group.

Add a decision rationale that explains why the strategy fits the consequence and why alternatives were rejected. A high-priority label without that explanation is insufficient. Keep unverified information visible on the export rather than removing it for presentation.

32 Define a stock policy that people can operate

Specify the replenishment trigger, who reviews it, the ordering quantity rule and the treatment of reservations. For repairables, define the action immediately after an exchange and the escalation if repair cannot restore coverage. For scheduled consumables, connect the policy to the maintenance plan.

Identify the evidence needed before receiving stock is released. This may include identity, quantity, documentation, condition, compatibility and configuration. The exact acceptance depends on the part. Do not invent a universal inspection procedure that could conflict with manufacturer instructions or specialist requirements.

State how exceptions are handled. Emergency substitution, borrowing from a reserved job and using shared stock should create a traceable decision and replenishment action. Otherwise a successful urgent repair can silently transfer risk to the next job or site.

33 Commission the inventory strategy

Commissioning the strategy means verifying that the intended recovery arrangement actually exists. Confirm the physical item, correct records, storage condition, access, companion materials and installation information. For a shared or external arrangement, verify the applicable agreement and the route to release and transport.

Use a suitable tabletop or approved practical exercise to test the sequence from fault identification to accepted restoration. Record delays, missing information and unclear authority. The exercise should not create an uncontrolled interruption or require unnecessary disassembly of healthy equipment.

After the exercise, update the timeline and stock status. A part that can be found but cannot be configured or installed within the objective does not close the service gap. Link remaining actions back to the criticality record so the plant does not assume that purchasing the spare completed the continuity project.

34 Review after an issue or a change

Every significant issue provides evidence: was the right item available, was the diagnosis correct, did installation need additional parts, and how long did restoration take? Record whether the removed part was repaired, returned, scrapped or retained for analysis. This improves both demand history and future repair preparation.

Review after equipment modification, fleet replacement, revised operating exposure, supply disruption or a new obsolescence notice. A previously valid compatibility group can become invalid when a machine is upgraded. A service agreement may change the preferred balance between local and pooled stock.

Preserve prior decisions and the evidence that supported them. Reducing stock after improved supply or retiring equipment can be appropriate, but it should be a deliberate decision. A low issue count alone is not enough to discard a spare that protects a rare severe event.

35 Frequently asked questions

Should the most expensive part receive the highest priority

Priority follows the consequence of unavailable service, restoration time and practical alternatives. A small component can stop an important function. An expensive component may have a verified exchange or standby route. Purchase price belongs in the economic comparison after the service requirement is understood.

Is one spare enough for several identical compressors

It may be, but only after compatibility, demand exposure, replenishment time and common-event risk are considered. Model assumptions must fit the installed population. A shared design or maintenance event can create simultaneous demand that a simple independent model misses.

Does an approved cross reference remove compatibility risk

Approval has a scope. Check that the exact equipment, revision, materials, duty and configuration fall within it. Some substitutions require companion changes. Preserve the approval evidence and do not extend it silently to other equipment.

Can a quoted lead time be used in the worksheet

Use it as a dated commercial input with its scope and conditions. Distinguish dispatch, manufacture and delivery from time to accepted usable stock. Compare the quotation with observed end-to-end experience where available and retain uncertainty.

Should a part with no issues in several years be removed from stock

First determine why it was stocked. It may protect a rare high-consequence failure or an obsolete component. Review remaining equipment life, preservation, alternative recovery and current supply. Lack of consumption is evidence to investigate, not an automatic disposal instruction.

Can the worksheet guarantee a service level

The worksheet records a strategy and its assumptions. Statistical outputs describe a selected model, and commercial commitments come from actual agreements. Installed recovery, usable stock and the relevant service conditions need their own evidence. No calculation can guarantee all future failures.