Spare Parts, Support Functions and Asset Supply Networks

Introduction

Total Productive Maintenance (TPM) is often presented as a shop-floor system, but equipment reliability depends equally on materials, information, services, utilities, finance, procurement and external partners. A machine cannot be restored to operating condition merely because a failure has been diagnosed; the correct part must be available, technically suitable, traceable and economically justified. Consequently, spare-parts management and maintenance logistics are central to availability, cost, quality and safety.

This chapter extends TPM beyond production equipment. It examines how organizations classify and stock spares, determine inventory policies, control obsolescence, decide whether to repair or replace, manage vendors and develop reliable asset supply networks. It also applies TPM principles to offices, warehouses, utilities and supply partners.

Learning Objectives

After studying this chapter, the learner should be able to:

1. Classify spare parts according to criticality, demand and replenishment characteristics.
2. Set inventory policies using service requirements, lead time, uncertainty and total cost.
3. Analyse obsolescence, duplication and inactive stock.
4. Make repair-versus-replace decisions using life-cycle economics and risk.
5. Evaluate vendor and maintenance-logistics performance.
6. Extend TPM to support functions and inter-organizational supply networks.
7. Design governance mechanisms that integrate maintenance, operations, procurement, finance and suppliers.

1. Spare Parts as a Reliability Resource

A spare part is an item held or obtainable to restore an asset, subsystem or support process. It may be a standard bearing, an engineered electronic module, a seal kit, a software licence, a filter or an entire replacement assembly. Spare-parts management differs from ordinary production inventory in several important ways:

• Demand is often intermittent and difficult to forecast.
• A stockout may stop a high-value production line.
• Parts may have long or uncertain replenishment lead times.
• Technical interchangeability is frequently limited.
• Some items have safety, regulatory or warranty implications.
• Demand may be generated by preventive, predictive or corrective maintenance rather than customer orders.

TPM views the spare as part of the equipment-maintenance system, not merely as a warehouse item. The relevant question is not “How cheaply can the part be purchased?” but “What combination of stock, repair capability, supplier responsiveness and equipment design provides the required asset availability at acceptable life-cycle cost?”

Availability is commonly expressed as the proportion of scheduled time during which an asset is capable of performing its intended function. For a repairable asset under steady operating conditions, an operational approximation is:

Here, A is inherent or operational availability under the stated definition, MTBF is mean time between failures, and MTTR is mean time to repair. Times must use the same units. The relationship assumes that failure and restoration behaviour can reasonably be represented by average times and does not include every factor affecting operational availability, such as waiting for parts, administrative delay or planned maintenance. In practice, spare-parts availability influences MTTR through waiting time, while poor part quality can reduce MTBF.

2. Classification and Criticality

No single classification is sufficient. A professional system uses several dimensions simultaneously.

2.1 Functional and technical classification

A catalogue should identify the asset hierarchy, manufacturer, part number, revision, specification, unit of measure, interchangeability, storage conditions and approved substitutes. Items may be classified as consumables, repairables, rotables, insurance spares, kits, tools, lubricants, safety-critical components or information assets.

A clean equipment bill of materials (BOM) links each part to an asset and its position. This prevents two common failures: purchasing an unsuitable item because descriptions are ambiguous, and holding duplicate items under different catalogue numbers.

2.2 Criticality classification

Criticality should reflect consequences, not only purchase price. A practical assessment considers:

• Safety and environmental consequence of failure.
• Production loss and customer impact.
• Quality and regulatory consequence.
• Availability of temporary bypasses or redundancy.
• Lead time and supplier vulnerability.
• Probability and frequency of failure.
• Technical substitutability.

A critical item may have low annual demand but require immediate availability. Conversely, a high-value component may be non-critical if a redundant unit is available and replenishment is rapid.

2.3 Inventory analysis

ABC analysis ranks items by annual usage value. For item i, annual usage value is:

where d_i is annual demand in units and c_i is unit cost. ABC categories are then assigned by cumulative value, with organizational thresholds selected from the data. ABC does not measure criticality; therefore an inexpensive safety-critical fuse may be classified as A, B or C by value but must receive a criticality override.

XYZ analysis classifies demand variability: X items have relatively stable demand, Y items have moderate variation or seasonality, and Z items are highly intermittent. Other useful dimensions include lead-time class, repairability, obsolescence risk and supplier dependence.

A criticality matrix should be reviewed with maintenance, operations, engineering, procurement, finance, safety and suppliers. It should not be created solely by the stores department.

3. Inventory Policy and Service Decisions

The fundamental trade-off is between the cost of holding stock and the cost of stockout. Holding costs include capital, space, insurance, deterioration, handling and obsolescence. Stockout costs include downtime, emergency freight, overtime, quality losses, contractual penalties and safety exposure.

3.1 Reorder point and safety stock

For a continuously monitored item with approximately independent demand and lead-time uncertainty, a common policy is:

where ROP is reorder point, μ_LT is expected demand during lead time, and SS is safety stock. If daily demand has mean μ_d, lead time has mean L, and demand and lead time are treated as independent, expected lead-time demand is:

A frequently used approximation for the standard deviation of demand during lead time is:

Here, σ_d is the standard deviation of daily demand and σ_L is the standard deviation of lead time; units must be consistent. If the desired cycle service level corresponds to a standard-normal factor z, then:

These equations assume a suitable normal approximation, stable parameter estimates and no significant demand–lead-time correlation. Intermittent spare demand may violate these assumptions; Croston-type methods, empirical quantiles, simulation or scenario analysis may be more appropriate.

Worked example

A filter has average daily demand of 0.20 units, daily standard deviation of 0.30 units, mean lead time of 10 days and lead-time standard deviation of 2 days. For a service target represented by z = 1.65:

μ_LT = 0.20 × 10 = 2.0 units.

σ_LT = √(10 × 0.30² + 0.20² × 2²) = √(0.9 + 0.16) ≈ 1.03 units.

SS = 1.65 × 1.03 ≈ 1.70 units.

Thus, ROP ≈ 2.0 + 1.70 = 3.70 units. Since physical stock is discrete, the organization would normally set the reorder point at 4 units, subject to minimum order quantities and criticality review. This is a planning approximation, not a guarantee of service.

3.2 Order quantity and intermittent demand

The economic order quantity (EOQ) is useful when demand is reasonably stable, replenishment is instantaneous or suitably modelled, shortages are excluded and ordering and holding costs are known:

D is annual demand, S is ordering or setup cost per order, and H is annual holding cost per unit. Units must be consistent; if H is expressed in currency per unit per year, D must be units per year and S currency per order. EOQ is not automatically appropriate for expensive, obsolete, intermittent or highly critical spares.

For low-frequency failures, a base-stock or one-for-one replenishment policy may be preferable. For a long-lead insurance spare, the policy may be “one unit held, replace immediately after use.” Pooling between plants can reduce total stock, but it introduces transport time, ownership disputes and priority rules.

3.3 Service-level segmentation

Different parts should receive different control intensity:

• Insurance spares: high consequence, low demand; hold or secure guaranteed access.
• Routine consumables: regular demand; use min–max, two-bin or kanban controls.
• Repairables: issue a serviceable unit and return the failed unit to a controlled repair loop.
• Emergency items: define approved suppliers, substitutes and expedited transport.
• Obsolescent items: restrict purchasing and establish redesign or last-time-buy decisions.

A high target service level should be justified by consequence and not used indiscriminately, because increasing service targets can require disproportionately more stock.

4. Data, Warehousing and Maintenance Logistics

The maintenance store is a technical control point. Essential practices include barcode or RFID identification, location discipline, cycle counting, preservation packaging, shelf-life control, calibration status and quarantine of nonconforming material. Parts should be issued against a work order so that actual consumption, failure mode and asset history are connected.

Master-data governance should control duplicate descriptions, units of measure, supersession, manufacturer changes and approved equivalents. A part number without a functional location or technical specification is a weak data object. Digital systems should expose reservations, stock in transit, repair status, purchase orders and supplier commitments.

A closed-loop repair process normally includes: failure identification; removal and tagging; technical inspection; quotation or internal diagnosis; repair authorization; repair and test; receipt inspection; return to serviceable stock; and failure-history update. Without this loop, repairables disappear into uncontrolled queues and the organization purchases replacements unnecessarily.

Warehouse TPM applies the same principles used on equipment: eliminate abnormalities, standardize work, maintain visual controls, measure losses and involve operators. Typical warehouse losses include searching, picking errors, damaged packaging, expired materials, duplicate records and delayed kitting.

5. Obsolescence and Excess Inventory

Obsolescence occurs when a part is no longer required, supported, compatible or economically sensible. Causes include equipment replacement, engineering changes, discontinued electronics, software dependence, regulatory change, supplier exit and inaccurate bills of material.

An obsolescence programme should:

1. Identify inactive and slow-moving stock.
2. Check open work orders, future shutdowns and installed equipment.
3. Validate quantities against current BOMs and asset populations.
4. Determine whether the item is interchangeable or repairable.
5. Consult engineering and original equipment manufacturers.
6. Decide to retain, transfer, return, sell, cannibalize, redesign or dispose.
7. Record the decision, owner and review date.

Disposal must consider hazardous substances, data-bearing devices, counterfeit risk and environmental requirements. Financial write-offs should not be used to conceal weak planning; conversely, retaining unusable stock merely to avoid a write-off can distort inventory and delay better engineering decisions.

A useful governance metric is the proportion of inventory value without a validated current requirement. It should be interpreted alongside criticality: a dormant emergency spare may be intentional, while a dormant standard motor may indicate excess.

6. Repair-versus-Replace Decisions

The decision must combine economics, reliability, time, risk and capability. A repair is attractive when the unit is technically recoverable, repair quality is demonstrable, turnaround meets the need and life-cycle cost is lower. Replacement is attractive when repair has uncertain reliability, parts are obsolete, the asset is unsafe or a new unit provides materially better performance.

A simple comparison is:

These are decision models rather than universal accounting identities. Each term must be defined consistently, and the time horizon and probability of repeated failure must be stated. For alternatives with different lives, net present value is more appropriate:

C₀ is the initial cost at time zero, C_t is the net cash outflow in period t, r is the discount rate per period and n is the analysis horizon. Costs and benefits must be expressed in the same currency and period basis. Risk-adjusted costs should use credible probabilities, not arbitrary safety factors.

Technical criteria should include dimensional conformity, material compatibility, firmware or software compatibility, certification, warranty, maintainability and cybersecurity where connected components are involved. A cheap non-approved substitute can create a larger quality or safety loss.

7. Vendor Performance and Supply-Network Design

Maintenance supply networks include OEMs, distributors, repair specialists, logistics providers, laboratories, contractors and internal stores. Supplier selection should assess total cost and resilience, not unit price alone.

Important supplier measures include:

• On-time-in-full delivery.
• Confirmed lead-time adherence.
• Correct part and revision delivered.
• Incoming acceptance rate.
• Warranty return rate.
• Repair turnaround time.
• Emergency response performance.
• Documentation and traceability.
• Corrective-action effectiveness.
• Total cost, including freight and downtime exposure.

On-time-in-full (OTIF) may be calculated as:

The organization must define “on time,” “in full,” order lines versus orders, partial deliveries and approved exceptions before comparing suppliers. Otherwise, the metric creates misleading rankings.

Supplier development follows the TPM logic of joint problem solving. Teams can conduct value-stream mapping, failure analysis, process audits, capability reviews and shared improvement projects. Contracts should specify technical configuration, change notification, counterfeit prevention, warranty, service-level commitments, emergency support, intellectual property, data access and exit arrangements.

Network resilience requires alternatives for vulnerable items, but dual sourcing is not automatically optimal. Qualification of a second source may increase cost and technical variation. Decisions should consider concentration risk, geopolitical exposure, transport modes, cyber risk, financial health and recovery time after disruption.

8. TPM in Support Functions

TPM’s pillars can be translated beyond the production cell:

For utilities, reliability often affects the whole factory. Compressed air, steam, chilled water, power quality, water treatment and network services should have asset registers, criticality ratings, condition monitoring and contingency plans. For offices, support-process availability may concern payroll, maintenance-management systems, engineering data or permits. A system outage can become a maintenance and production event.

Warehouses should be treated as process systems with standard work, abnormality tags, housekeeping, ergonomic design, rack inspection and visual performance boards. TPM does not mean asking every employee to repair complex equipment; it means assigning appropriate ownership, early detection and escalation.

9. Digital Technologies and Artificial Intelligence

Computerized maintenance-management systems (CMMS), enterprise-resource-planning systems, warehouse-management systems and supplier portals can connect demand, asset history and procurement. RFID, condition sensors and electronic work instructions improve traceability. Digital twins may link component configuration to predicted maintenance demand.

Artificial intelligence can support failure-demand forecasting, duplicate-part detection, invoice anomaly detection, supplier-risk screening and recommended substitutes. However, predictions are only as sound as the failure coding, BOM accuracy and historical data. Rare-event data are especially vulnerable to sampling bias. Human approval remains essential for safety-critical substitutions, write-offs and changes to approved sources.

Governance should address access control, cybersecurity, explainability, data ownership, model drift and audit trails. A predictive model that recommends a part unavailable in the actual supply network is operationally useless. Technology should therefore be evaluated by improved decisions and reduced losses, not by the number of dashboards installed.

10. Illustrative Case: Apex Components

Background

Apex Components operates a continuous process line with three production cells, a central maintenance store and a regional repair vendor. The line’s annual output is high, and an electronic drive module has a long lead time. The module is repairable, but historical records contain three catalogue numbers for apparently identical revisions.

Problem

The plant held six modules, yet experienced two stoppages while waiting for a technically compatible unit. Investigation found that four modules were obsolete revisions, one was awaiting an undocumented repair, and one was serviceable. Procurement measured supplier performance by purchase price and invoice date rather than usable delivery time. The CMMS did not record failure mode or repair turnaround.

Analysis

A cross-functional team mapped the asset BOM, verified revision compatibility and calculated the downtime exposure. Criticality was rated high because there was no bypass and regional replenishment took between 8 and 14 weeks. The team separated serviceable, unserviceable, obsolete and quarantined stock. It also found that planned replacement demand had been mixed with random failure demand.

Approach

Apex implemented the following controls:

• One governed catalogue record per approved revision.
• A criticality-based insurance-spare policy.
• A repairable loop with tagged ownership and a 15-day turnaround target.
• Incoming functional testing and serial-number traceability.
• A qualified second repair source.
• Work-order consumption and failure-mode coding.
• Supplier OTIF and repair-turnaround measures.
• Engineering review before any substitute or last-time buy.
• Quarterly obsolescence review involving engineering and finance.

The organization also improved access to the module, introduced a standard removal procedure and trained technicians in electrostatic-discharge controls.

Outcome

After two review cycles, the plant had fewer nominal stockouts, reduced duplicate records and a visible repair queue. The number of physical modules was not simply increased; stock was reconfigured toward current, tested revisions. Management reported improved decision quality because downtime, repair and supplier data were now connected. The case illustrates that availability improvement may come from data integrity, repair-loop control and supplier governance rather than additional inventory alone.

Learning Points

1. Physical quantity is not equivalent to usable availability.
2. Criticality must override purely financial classifications.
3. Repairables require ownership, status visibility and turnaround control.
4. Engineering configuration management is essential for spare-parts accuracy.
5. Supplier performance should be measured against maintenance outcomes.

11. Challenges, Ethics and Governance

Spare-parts decisions create ethical and regulatory responsibilities. Counterfeit or undocumented components may compromise safety. Disposal of electronic, chemical or contaminated materials must follow applicable environmental requirements. Procurement personnel must manage conflicts of interest, gifts, bid confidentiality and segregation of duties. Emergency purchasing should be controlled, not used as a permanent route around competitive governance.

A robust governance model assigns clear accountability:

• Operations: define production consequence and availability need.
• Maintenance: define technical function, failure modes and restoration method.
• Engineering: approve configuration, substitutions and obsolescence decisions.
• Stores: maintain custody, preservation and transaction accuracy.
• Procurement: manage commercial terms, competition and supplier development.
• Finance: evaluate working capital and life-cycle economics.
• Safety and environment: approve hazardous, regulated and safety-critical decisions.
• Suppliers: provide conformity, traceability, capability and corrective action.

Common limitations include poor failure data, unstable demand, hidden downtime costs, biased supplier metrics and overreliance on average values. Managers should use sensitivity analysis: test how conclusions change if lead time, failure frequency, downtime cost or repair success rate changes. Critical decisions should include a documented assumption register.

Best-Practice Implementation Roadmap

1. Establish an asset hierarchy and controlled spare-parts catalogue.
2. Clean BOMs, eliminate duplicate records and verify technical revisions.
3. Conduct cross-functional criticality assessment.
4. Segment items by demand, lead time, repairability and obsolescence risk.
5. Set differentiated service and replenishment policies.
6. Introduce work-order issue, barcode control and cycle counting.
7. Build closed-loop repair and quarantine processes.
8. Define vendor measures and joint corrective-action routines.
9. Extend TPM audits to utilities, warehouses, offices and contractors.
10. Review outcomes using availability, downtime waiting, inventory health, repair turnaround, supplier performance and total cost.

The goal is not maximum inventory or minimum inventory. It is the controlled availability of the right capability at the right time, supported by reliable information and economically defensible decisions.

Chapter Summary

Spare parts are reliability resources whose value depends on technical suitability, availability, quality and timing. Effective management combines criticality assessment with ABC, demand-variability, lead-time, repairability and obsolescence analysis. Reorder-point and safety-stock formulas are useful under stated assumptions, but intermittent and critical spares often require empirical, scenario-based or base-stock policies.

Inventory accuracy, BOM governance, preservation, repair loops and work-order consumption are foundational controls. Repair-versus-replace decisions should include direct cost, downtime, risk, life-cycle economics and technical compliance. Supplier management must measure usable maintenance performance, including OTIF, conformity, repair turnaround and corrective action. TPM principles extend naturally to utilities, offices, warehouses and supply partners through ownership, standardization, abnormality detection and continuous improvement. Digital tools and AI can improve visibility and prediction, but require sound data, human governance and cybersecurity.

Key Terms

Spare part: An item used to restore an asset or support process.

Criticality: The consequence and urgency associated with failure or unavailability.

Insurance spare: A low-demand item held because its failure consequence is high.

Repairable: An item that can be restored to a defined serviceable condition.

Rotable: A repairable component exchanged between an asset and a repair cycle.

ABC analysis: Classification by annual usage value.

XYZ analysis: Classification by demand stability or variability.

Reorder point: Stock position at which replenishment is triggered.

Safety stock: Additional stock held against uncertainty.

Service level: A defined probability or performance target for meeting demand.

BOM: Bill of materials linking components to an asset or assembly.

Obsolescence: Loss of future usefulness, support or compatibility.

Quarantine: Controlled segregation of material pending inspection or decision.

Repair loop: Managed process from failed-item removal through restoration and return to stock.

OTIF: On-time-in-full delivery performance.

Life-cycle cost: Total cost incurred over an asset or component’s useful life.

Criticality override: A rule preventing value-based classification from ignoring consequence.

Asset supply network: Internal and external organizations that provide asset materials and services.

Master-data governance: Controls ensuring accurate, consistent and authorized data.

Review Questions

Short Answer Questions

1. Why is spare-parts demand usually more difficult to forecast than production-material demand?
2. Distinguish ABC classification from criticality classification.
3. What information should a controlled spare-parts catalogue contain?
4. Why must repairable items have a closed-loop process?
5. Identify four measures suitable for evaluating a maintenance supplier.

Long Answer Questions

1. Explain how TPM principles can be applied to a maintenance warehouse.
2. Discuss the assumptions and limitations of reorder-point and safety-stock calculations.
3. Develop a framework for managing obsolescence in an industrial plant.
4. Explain the technical, economic and governance factors in repair-versus-replace decisions.
5. Discuss how supplier development and network resilience contribute to equipment availability.

Analytical / Application Questions

1. A critical electronic module has intermittent demand, a 12-week lead time and no equipment bypass. Design a policy covering stock, repair, supplier qualification and obsolescence review. Justify each element.
2. A plant has high inventory value but frequent maintenance stockouts. Diagnose at least five possible causes and propose a measurement-led improvement plan.
3. Using the formulas in this chapter, calculate a reorder point when daily demand mean is 0.4 units, daily standard deviation is 0.5 units, mean lead time is 8 days, lead-time standard deviation is 1.5 days and z = 2.05. State assumptions and interpret the result.

Multiple Choice Questions with Answers

Question 1

Which classification factor should override a purely value-based ABC decision when a low-cost item can cause a major safety consequence?

A. Annual usage value

B. Criticality

C. Supplier discount

D. Warehouse location

Correct Answer: B. Criticality

Question 2

In the reorder-point model, what does safety stock primarily protect against?

A. Uncertainty in demand and replenishment

B. The purchase price of the item

C. The asset’s rated power

D. The accounting depreciation method

Correct Answer: A. Uncertainty in demand and replenishment

Question 3

Which condition most limits the direct use of EOQ for an intermittent insurance spare?

A. The item has a catalogue number

B. Demand is highly irregular and infrequent

C. The item is stored indoors

D. The supplier provides a quotation

Correct Answer: B. Demand is highly irregular and infrequent

Question 4

What is the principal purpose of a controlled bill of materials for spare parts?

A. To increase the number of catalogue records

B. To link approved components to asset configurations

C. To eliminate all preventive maintenance

D. To replace supplier contracts

Correct Answer: B. To link approved components to asset configurations

Question 5

Which activity is part of a closed-loop repair process?

A. Discarding every failed component immediately

B. Returning a repaired and tested item to serviceable stock

C. Issuing parts without a work order

D. Changing revisions without engineering approval

Correct Answer: B. Returning a repaired and tested item to serviceable stock

Question 6

Which measure evaluates whether supplier orders arrive on time and in the required quantity?

A. MTBF

B. OTIF

C. NPV

D. MTTR

Correct Answer: B. OTIF

Question 7

Which cost is most directly associated with a stockout of a production-critical spare?

A. Downtime exposure

B. Catalogue formatting

C. Shelf label colour

D. Annual depreciation of the warehouse

Correct Answer: A. Downtime exposure

Question 8

Why should technical approval be required before using a substitute spare?

A. Substitutes always cost more

B. Compatibility, safety and performance may differ

C. Substitutes cannot be tracked

D. Suppliers cannot provide documentation

Correct Answer: B. Compatibility, safety and performance may differ

Question 9

A repair decision should generally compare repair and replacement using which perspective?

A. Unit purchase price only

B. Life-cycle cost, risk and downtime

C. Catalogue description length

D. Warehouse shelf capacity only

Correct Answer: B. Life-cycle cost, risk and downtime

Question 10

Which is an appropriate application of TPM in an office or support function?

A. Ignoring abnormalities because no production machine is present

B. Standardizing checks and escalation for critical information systems

C. Allowing uncontrolled master-data changes

D. Removing all preventive inspections

Correct Answer: B. Standardizing checks and escalation for critical information systems