Maintenance Fundamentals and Maintenance Strategies

Introduction and Context

Maintenance is the coordinated set of technical, managerial and administrative activities used to retain an asset in, or restore it to, a state in which it can perform its required function. In production management, maintenance is not merely the repair of broken machines. It is a business process that protects safety, product quality, delivery reliability, energy performance, regulatory compliance and the total cost of ownership.

A production system contains interacting assets: machines, tools, utilities, control systems, buildings and information systems. Failure of one item may reduce the output of an entire line. Consequently, maintenance decisions should be based on the function and consequences of failure, not only on the physical condition of an individual component.

This chapter explains deterioration and failure, introduces major maintenance strategies, and develops a framework for organizing and selecting maintenance work according to risk, criticality, cost and production requirements.

Learning Objectives

By the end of this chapter, the learner should be able to:

1. Explain how equipment deteriorates and distinguish failure, fault, defect and breakdown.
2. Interpret common failure patterns and their implications for maintenance planning.
3. Describe the roles, structures and information flows of a maintenance organization.
4. Compare reactive, corrective, preventive, predictive, condition-based and proactive maintenance.
5. Select an appropriate strategy using risk, asset criticality, failure modes, cost and production requirements.
6. Apply maintenance performance measures and recognize limitations of maintenance data.
7. Evaluate maintenance decisions from technical, operational, ethical and regulatory perspectives.

1. Fundamental Concepts

1.1 Asset function and required performance

An asset is maintained for the service it provides. Its function should therefore be stated in measurable terms, such as producing 500 units per hour, maintaining pressure between specified limits, or stopping within a required distance. A functional failure occurs when the asset cannot perform a required function to the specified performance standard.

A useful distinction is:

• Failure: loss of the ability to perform a required function.
• Fault or defect: an abnormal condition that may or may not yet have caused functional failure.
• Breakdown: a failure that makes continued operation impossible or unacceptable.
• Degradation: gradual reduction in performance or condition.
• Repair: action that restores function after failure.
• Maintenance: the broader process of inspection, servicing, adjustment, restoration and improvement.

The same physical condition can have different consequences. A noisy fan in a non-critical office may be tolerated temporarily, while the same condition in a dust-controlled pharmaceutical process may require immediate intervention.

1.2 Equipment deterioration

Deterioration results from physical, chemical, electrical, thermal, environmental and human influences. Common mechanisms include:

• Wear: progressive material loss caused by sliding, rolling, impact or erosion.
• Corrosion: chemical or electrochemical attack that reduces material integrity.
• Fatigue: crack initiation and growth under repeated stress cycles.
• Creep: time-dependent deformation under sustained load, especially at elevated temperature.
• Fouling and deposition: accumulation that restricts flow or impairs heat transfer.
• Insulation breakdown: deterioration caused by heat, moisture, contamination or electrical stress.
• Loosening and misalignment: loss of mechanical integrity caused by vibration, thermal cycling or poor installation.
• Obsolescence: loss of support, compatibility, cybersecurity or economic viability even when the equipment still operates.

Deterioration may be accelerated by overload, poor lubrication, contamination, incorrect operating procedures, unsuitable spare parts, inadequate foundations, poor design or an unfavourable environment. Maintenance personnel should therefore avoid assuming that replacement alone solves a recurring problem.

1.3 Reliability, maintainability and availability

Reliability is the probability that an item performs its required function for a specified time under stated conditions. Maintainability describes the ability to be retained in, or restored to, a functioning state using prescribed procedures and resources. Availability expresses the proportion of time that an asset is capable of performing its required function.

For a stable system with repairable equipment, a commonly used inherent-availability approximation is:

Here, A_i is inherent availability, MTBF is mean time between failures, and MTTR is mean time to repair. Both time measures must use the same units. The approximation assumes a repairable item, statistically stable operating conditions, and that only active repair time is included; waiting for parts, access or management authorization is excluded. Operational availability is usually lower because it includes delays such as logistics and administration.

For observed operating data:

These measures are meaningful only when the population, observation period, failure definition and treatment of planned stoppages are clearly specified. A high MTBF does not prove that an asset is well managed if failures are under-reported.

Worked example. A pump operates for 7,200 hours and experiences six functional failures. The total active repair time is 48 hours. Thus, MTBF = 7,200 ÷ 6 = 1,200 hours and MTTR = 48 ÷ 6 = 8 hours. The estimated inherent availability is 1,200 ÷ (1,200 + 8) = 0.9934, or approximately 99.34%. This result excludes waiting for a mechanic or spare part; a plant manager should not interpret it as the actual proportion of scheduled production time available.

2. Failure Behaviour and Failure Patterns

2.1 The bathtub curve

The bathtub curve is a conceptual representation of a population’s conditional failure rate over time. It commonly has three regions:

1. Early-life failures: a decreasing failure rate caused by design errors, manufacturing defects, installation mistakes or commissioning problems.
2. Useful life: an approximately stable random-failure rate.
3. Wear-out period: an increasing failure rate caused by fatigue, wear, corrosion, ageing or depletion of life-limited components.

The curve is not a universal law. Many modern complex systems show random or multiple failure patterns rather than a clear wear-out phase. Replacing equipment solely because it is old may therefore provide little benefit unless age-related deterioration is demonstrated.

2.2 Failure modes and consequences

A failure mode is the specific manner in which an item fails, such as bearing seizure, seal leakage, sensor drift or circuit-breaker failure. A failure cause is the underlying reason, such as contamination, overload or incorrect calibration. A failure effect is what is observed locally, while a failure consequence is the broader effect on safety, production, quality, environment, cost or reputation.

Failure Modes and Effects Analysis (FMEA) structures this reasoning by asking what can fail, why it can fail, what happens, how it can be detected and what action is appropriate. Risk-based methods may rank each failure mode using severity, occurrence and detectability scores, but numerical rankings should support—not replace—engineering judgment. A low-frequency event with catastrophic consequences must not be dismissed merely because its calculated score is modest.

2.3 Random and age-related failures

A time-based replacement policy is most defensible when the probability of failure increases predictably with age and replacement restores a meaningful level of reliability. It is less defensible when failures are random, when condition can be monitored more effectively, or when replacement introduces installation errors.

Failure patterns should be established from evidence: work orders, inspection findings, sensor histories, warranty data, operating conditions and expert knowledge. Data quality is critical. If technicians record symptoms inconsistently, statistical conclusions can be misleading.

3. Maintenance Organization and Work Management

3.1 Purpose and structure

A maintenance organization converts business requirements into safe, planned and controlled work. Typical roles include:

• Maintenance manager: establishes policy, budget, standards and performance objectives.
• Reliability or maintenance engineer: analyses failure, improves maintainability and develops strategies.
• Planner: defines job scope, labor, tools, permits, procedures and estimated duration.
• Scheduler: coordinates work with production, materials and access windows.
• Supervisor: directs field execution and verifies quality.
• Craft personnel and technicians: inspect, diagnose, repair, test and document equipment.
• Stores and procurement personnel: control spares, suppliers and material availability.
• Operations personnel: operate equipment, report abnormalities and participate in basic care.
• Safety, quality and environmental specialists: verify legal and organizational controls.

A centralized organization can standardize skills, tools and planning across a plant. A decentralized organization can respond rapidly to local production needs. A hybrid structure often combines a central reliability, planning and materials function with area-based execution teams.

3.2 The maintenance work process

A disciplined work-management cycle normally includes:

1. Identification of an abnormality or improvement opportunity.
2. Screening for immediate safety, environmental or production risk.
3. Work request approval and prioritization.
4. Technical planning, including job steps, permits, isolation, labor, tools and materials.
5. Scheduling with production and other departments.
6. Safe execution and quality control.
7. Testing, recommissioning and handover.
8. Close-out in the computerized maintenance management system (CMMS).
9. Review of causes, costs, delays and learning.

The CMMS should be treated as a management-control system, not simply a store of electronic work orders. Asset hierarchy, criticality, failure codes, labor hours, parts and completion notes must be standardized sufficiently to support analysis.

3.3 Planning, scheduling and backlog

Planning determines how work should be performed; scheduling determines when resources will perform it. Planned work normally has lower execution risk and shorter duration than unplanned work because tools, permits and parts are prepared in advance. A backlog is not automatically unhealthy: it must be classified by priority, age, risk and readiness. A large backlog of low-risk work may be acceptable, while one overdue safety-critical inspection is unacceptable.

Maintenance and operations should agree on equipment release, production windows, temporary controls and return-to-service criteria. Maintenance efficiency cannot be judged independently of production stability and safety.

4. Maintenance Strategies

4.1 Reactive or run-to-failure maintenance

Reactive maintenance allows an item to operate until it fails, after which corrective action is taken. It can be appropriate when failure has low consequences, repair is quick, redundancy exists, and deterioration cannot be economically monitored. Examples include inexpensive, non-critical lighting components or replaceable office equipment.

It is unsuitable where failure can cause injury, environmental release, major quality loss, extensive secondary damage or a long production outage. “Run to failure” should be a deliberate strategy with prepared spares and controls, not the absence of a strategy.

4.2 Corrective maintenance

Corrective maintenance restores an item after a defect or failure has been identified. It includes emergency repairs, but also planned correction of a known defect during a suitable production opportunity. Corrective work may be deferred when risk is controlled, or immediate when continued operation is unsafe or unacceptable.

Corrective maintenance is often more efficient than preventive replacement when the defect is observable and consequences are limited. Root-cause analysis is necessary when the same defect recurs.

4.3 Preventive maintenance

Preventive maintenance is performed at predetermined intervals or according to specified criteria to reduce the probability of failure or degradation of function. It includes lubrication, inspection, cleaning, adjustment, calibration, overhaul and scheduled replacement.

Intervals should be based on manufacturer information, failure history, engineering analysis, legal requirements and operating context. Excessive preventive work creates cost, consumes parts and may introduce human error. Insufficient preventive work increases failure risk. A task should have a clear failure mechanism, a defined acceptance criterion and an action that changes risk.

4.4 Predictive and condition-based maintenance

Condition-based maintenance (CBM) initiates action when evidence of deteriorating condition reaches a defined threshold. Predictive maintenance is often used as a more analytical form of CBM, using condition data, trend analysis, models or machine learning to estimate future condition or remaining useful life.

Typical techniques include vibration analysis, infrared thermography, oil analysis, motor-current analysis, ultrasound, acoustic emission, performance testing and online process monitoring. Effective deployment requires a detectable relationship between the measured variable and functional failure, suitable sensor placement, reliable data and sufficient time to plan intervention.

A warning threshold should not be confused with a failure limit. For a monitored variable x, an organization may define a normal operating band and an intervention threshold, but the thresholds must be validated for the particular asset and operating state. Alarm fatigue, sensor drift, poor sampling and false positives can undermine confidence.

4.5 Proactive maintenance

Proactive maintenance addresses the causes of failure rather than repeatedly treating symptoms. It includes precision alignment and balancing, contamination control, lubrication improvement, design modification, operator care, root-cause analysis and elimination of recurring defects. Reliability-centred maintenance (RCM) is a structured decision process that selects technically feasible and economically justified tasks based on functions, functional failures, failure modes and consequences.

Total Productive Maintenance (TPM) extends responsibility for basic equipment care to operators while retaining specialist support. Activities may include cleaning, inspection, tightening, lubrication, abnormality tagging and small improvements. TPM is not a transfer of skilled maintenance work to untrained operators; competency, boundaries and escalation rules are essential.

4.6 Strategy comparison

These categories can coexist. A motor may receive routine lubrication, online vibration monitoring, corrective repair of identified defects and a proactive alignment improvement.

5. Selecting a Maintenance Strategy

5.1 Risk and criticality

Criticality is the relative importance of an asset to safety, environment, quality, delivery, cost and legal compliance. A practical assessment asks:

• What functions are required?
• What happens if the function is lost?
• Is there redundancy or a safe fallback?
• How quickly can the asset be restored?
• Can deterioration be detected before failure?
• What resources and skills are available?
• Is the strategy compatible with production windows?

Risk may be expressed conceptually as a combination of likelihood and consequence. A numerical risk score can support prioritization, but risk is not always reliably represented by multiplying two subjective numbers. High-consequence hazards require explicit controls, independent review and compliance with applicable law and standards.

5.2 Economic evaluation

The relevant comparison is not simply preventive-maintenance cost versus repair cost. It includes lost production, quality losses, energy effects, spare parts, labor, inspection, safety controls, environmental consequences and the probability and duration of failure.

For a proposed task, the expected annual cost can be represented conceptually as:

This expression is a decision aid, not a precise prediction. Probability and consequence must refer to the same period and include clearly defined cost components. Where data are uncertain, sensitivity analysis should be used rather than false precision.

5.3 Decision logic

A suitable sequence is:

1. Eliminate unsafe operation and meet mandatory legal or inspection requirements.
2. Define the asset’s required functions and performance standards.
3. Identify credible failure modes and consequences.
4. Determine whether failure is detectable and whether intervention time exists.
5. Select a technically feasible task: condition monitoring, scheduled restoration, scheduled discard, failure finding, redesign or deliberate run-to-failure.
6. Compare total life-cycle cost, resources and production compatibility.
7. Define task frequency, acceptance limits, escalation criteria and records.
8. Review performance after implementation.

A task should be removed or revised if it does not control a credible failure mechanism, produces no actionable information or creates disproportionate risk during execution.

6. Applied Case: Illustrative Packaging-Line Conveyor

Background

A food-packaging plant experiences repeated stoppages on a conveyor driven by a geared motor. The conveyor is important because one failure stops filling and packing downstream. The plant initially replaces the gearbox every 12 months, regardless of condition.

Problem

Work-order records show that only some gearboxes fail through wear. Several failures involve lubricant contamination and coupling misalignment. Replacement consumes a planned shift, while unplanned failures cause approximately four hours of lost production and urgent procurement.

Analysis

The reliability engineer conducts a simplified FMEA. Loss of conveyor drive has high delivery consequence, but the line has limited short-term buffering. The gearbox has observable precursors: rising vibration, abnormal temperature, oil debris and coupling misalignment. The annual replacement policy is therefore not sufficiently linked to actual condition. It also introduces installation and alignment errors.

Approach and solution

The team adopts a combined strategy:

• daily operator checks for leakage, noise and abnormal temperature;
• monthly visual inspection and coupling checks;
• quarterly vibration and oil-debris analysis;
• precision alignment after motor or gearbox removal;
• improved sealing and contamination-control practices;
• a planned spare gearbox and critical seal kit;
• a defined alarm-and-action matrix in the CMMS;
• root-cause review for every repeat failure.

The plant retains a preventive inspection because condition monitoring alone cannot detect every failure mode. Replacement is now condition-based or failure-directed rather than automatically annual.

Outcome

In this illustrative case, the plant reduces emergency stoppages and avoids unnecessary annual gearbox replacement. The principal gain is not a single technology; it is the integration of operator care, condition data, precision work, spares readiness and learning from failure. Results must be confirmed with comparable operating hours, production demand and failure definitions.

Learning points

1. Criticality should influence the level of control and preparation.
2. A scheduled replacement is not automatically preventive excellence.
3. Condition monitoring is effective only when measurements lead to timely action.
4. Repeated failure usually requires investigation of causes, not repeated replacement.
5. Maintenance performance must be evaluated with operations, safety and quality data together.

7. Performance Measures, Governance and Contemporary Practice

Useful indicators include planned-work percentage, schedule compliance, emergency-work percentage, maintenance cost per unit, overdue safety work, repeat-failure rate, PM compliance, stock-outs, MTBF and MTTR. No single indicator is sufficient. For example, high PM compliance can coexist with poor reliability if tasks are badly designed; low maintenance cost can reflect unsafe deferral.

Safety governance requires isolation and lockout/tagout, permit-to-work controls, competency verification, contractor management, confined-space controls, working-at-height precautions and post-maintenance testing. Environmental obligations may include leak prevention, waste-oil handling, refrigerant control and pollution-response readiness. Records must be truthful: falsifying inspection results or closing work without completion can expose workers and the organization to serious harm.

Digital CMMS and industrial Internet of Things technologies can improve asset history, mobile reporting, remote monitoring and analysis. Artificial intelligence can assist anomaly detection, failure classification and work prioritization, but it does not remove the need for engineering validation. Models may fail when operating regimes change, sensors drift or historical data contain hidden maintenance bias. Human approval, cybersecurity, access control, data governance and transparent escalation remain necessary.

Best practice is characterized by clear asset functions, risk-based strategy selection, competent people, standardized work, reliable data, planned materials, strong operations-maintenance cooperation and continuous review. Maintenance maturity develops from reacting to failures, through planning and prevention, toward reliability improvement and learning.

Chapter Summary

Maintenance management protects the function of production assets and must be connected to safety, quality, delivery and life-cycle cost. Equipment deteriorates through wear, corrosion, fatigue, contamination, thermal and electrical stresses, poor operating conditions and obsolescence. Failure analysis distinguishes functions, failure modes, causes, effects and consequences.

Reliability concerns successful operation for a specified time, maintainability concerns restoration, and availability combines operating and repair performance. The bathtub curve is a useful conceptual model but not a universal description of every asset. Maintenance organizations require clear responsibilities, disciplined work management, competent execution and trustworthy CMMS data.

Reactive, corrective, preventive, condition-based, predictive and proactive strategies serve different circumstances. Selection should consider safety and environmental consequence, asset criticality, failure detectability, age dependence, data quality, maintainability, cost, redundancy and production requirements. A mature program combines strategies, investigates recurring causes and reviews results using balanced performance measures. Digital monitoring and AI can enhance decisions, but they require sound engineering, cybersecurity, governance and human judgment.

Key Terms

Asset: A physical or digital item that provides a required function.

Availability: The proportion of time an item is capable of performing its required function.

Condition-based maintenance: Maintenance initiated by evidence of actual equipment condition.

Corrective maintenance: Maintenance performed to restore an identified defect or failure.

Criticality: The relative importance of an asset based on consequences of failure.

Degradation: Progressive reduction in performance or condition.

Failure mode: The specific manner in which an item fails.

Failure rate: The frequency or conditional likelihood of failure under stated conditions.

FMEA: A structured analysis of failure modes, causes, effects and controls.

Maintainability: The ability to restore an item using stated procedures and resources.

Maintenance: Activities that retain or restore an asset’s required function.

MTBF: Mean time between failures for a repairable item under defined conditions.

MTTR: Mean time to repair, usually calculated from total repair time and number of repairs.

Predictive maintenance: Data-based maintenance that forecasts deterioration or future failure.

Preventive maintenance: Maintenance performed at predetermined intervals or criteria.

Proactive maintenance: Actions that eliminate or control underlying causes of failure.

Reactive maintenance: Maintenance performed after functional failure, often called run-to-failure.

Reliability: Probability of performing a required function for a specified time and condition.

Risk: The combination of the likelihood and consequence of an undesired event.

TPM: Total Productive Maintenance, involving organization-wide equipment care and improvement.

Review Questions

Short Answer Questions

1. Distinguish a fault, a functional failure and a breakdown.
2. State two assumptions behind the inherent-availability approximation.
3. Why is the bathtub curve not a universal law for all equipment?
4. Identify four roles commonly found in a maintenance organization.
5. Give two circumstances in which run-to-failure may be appropriate.

Long Answer Questions

1. Explain the main mechanisms of equipment deterioration and show how operating conditions can accelerate them.
2. Compare preventive, condition-based, predictive and proactive maintenance, including their strengths and limitations.
3. Describe a complete maintenance work-management process from defect identification to work-order close-out.
4. Explain how FMEA and criticality analysis support maintenance-strategy selection.
5. Discuss the benefits and risks of using CMMS, sensors and artificial intelligence in maintenance management.

Analytical / Application Questions

1. A critical pump has an MTBF of 1,500 hours and an MTTR of 10 hours. Calculate its estimated inherent availability and state what the result excludes.
2. A machine has repeated bearing failures despite monthly replacement. Develop an investigation and strategy-selection approach that considers causes, condition monitoring, production risk and life-cycle cost.
3. A plant manager proposes annual replacement of every motor because “old motors are unreliable.” Critically evaluate this proposal and recommend the data and decision process needed before adopting it.

Multiple Choice Questions with Answers

Question 1

Which statement best defines a functional failure?

A. A reduction in purchase price of an asset

B. Loss of the ability to perform a required function

C. Any noise produced by rotating equipment

D. Completion of a preventive maintenance task

Correct Answer: B. Loss of the ability to perform a required function

Question 2

Which deterioration mechanism involves progressive material loss caused by relative motion?

A. Wear

B. Creep

C. Obsolescence

D. Insulation breakdown

Correct Answer: A. Wear

Question 3

A pump operates for 6,000 hours and has five failures. What is its MTBF?

A. 120 hours

B. 1,000 hours

C. 1,200 hours

D. 30,000 hours

Correct Answer: C. 1,200 hours

Question 4

Which maintenance strategy is triggered by measured evidence of equipment condition?

A. Reactive maintenance

B. Condition-based maintenance

C. Unplanned replacement

D. Run-to-failure maintenance

Correct Answer: B. Condition-based maintenance

Question 5

Why can automatic age-based replacement be ineffective?

A. All equipment improves with age

B. Many failures are random and do not become more likely predictably with age

C. Replacement never requires labor

D. Age is always unrelated to equipment condition

Correct Answer: B. Many failures are random and do not become more likely predictably with age

Question 6

What is the principal purpose of a CMMS in maintenance management?

A. To eliminate the need for technicians

B. To replace all physical inspections

C. To control work information, planning, history and analysis

D. To guarantee that failures cannot occur

Correct Answer: C. To control work information, planning, history and analysis

Question 7

Which activity is most clearly proactive maintenance?

A. Replacing a failed bearing with an identical bearing

B. Waiting for a motor to stop before ordering parts

C. Improving alignment and contamination control to remove recurring failure causes

D. Closing a work order without inspection

Correct Answer: C. Improving alignment and contamination control to remove recurring failure causes

Question 8

A machine has MTBF = 900 hours and MTTR = 9 hours. Using the inherent-availability approximation, what is availability?

A. 90.0%

B. 98.0%

C. 99.0%

D. 99.9%

Correct Answer: C. 99.0%

Question 9

Which factor should have the strongest influence on strategy selection for a safety-critical asset?

A. Paint colour

B. Consequence of failure and required risk controls

C. Technician preference alone

D. The age of the CMMS software

Correct Answer: B. Consequence of failure and required risk controls

Question 10

What is a major limitation of predictive maintenance based on artificial intelligence?

A. It never uses data

B. It cannot be affected by changing operating conditions

C. Models may produce unreliable results when data quality or operating regimes change

D. It always costs less than inspection

Correct Answer: C. Models may produce unreliable results when data quality or operating regimes change