11. Capacity, Constraints, Maintenance and Asset Productivity

Introduction and Context

Manufacturing performance depends not only on the availability of machines, labour, materials and technology, but also on how effectively these resources are coordinated. A plant may possess sufficient installed capacity and still fail to meet demand because capacity is unavailable at the required time, is located at the wrong process stage, or is restricted by a bottleneck. Similarly, a machine with high technical capability may contribute little economic value if it frequently fails, produces defects, or is operated at a speed below its designed rate.

Capacity management, constraint management, maintenance and asset productivity are therefore interconnected dimensions of production management. Capacity planning determines what a system can produce and when. Constraint analysis identifies the factors that limit system output. Maintenance preserves the capability of assets, while productivity measures reveal how effectively those assets convert inputs into saleable output. The managerial challenge is to optimise the performance of the entire production system rather than maximise the utilisation of individual machines in isolation.

This chapter integrates capacity planning, the Theory of Constraints (TOC), throughput accounting, Overall Equipment Effectiveness (OEE), Total Productive Maintenance (TPM), reliability economics and digital asset management. The emphasis is on decision-making: where to invest, what to improve first, how to interpret performance measures, and how to balance short-term throughput with long-term asset health.

Learning Objectives

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

1. Distinguish design, effective and actual capacity and calculate capacity-related measures.

2. Analyse bottlenecks and explain the logic of the Theory of Constraints.

3. Apply throughput accounting to operational and investment decisions.

4. Calculate and interpret OEE and its loss categories.

5. Explain TPM, reliability-centred maintenance and major maintenance policies.

6. Evaluate maintenance alternatives using life-cycle cost and reliability economics.

7. Integrate capacity, maintenance and productivity decisions at system level.

8. Assess the role of analytics, sensors and digital technologies in asset productivity.

Capacity as a System Property

Capacity is the maximum rate at which a production system can generate output under specified conditions. It is not a single universal number. A machine may have a nameplate capacity of 100 units per hour, but practical output may be lower because of product mix, changeovers, quality losses, labour availability, maintenance requirements or material shortages.

Three distinctions are particularly useful. Design capacity is the theoretical maximum output under ideal conditions. Effective capacity is the maximum output achievable under normal operating conditions, including planned breaks, changeovers, preventive maintenance and product variation. Actual output is what the system really produces during a period.

Capacity utilisation is commonly expressed as:

Utilisation = Actual Output / Design Capacity × 100

Efficiency is commonly expressed as:

Efficiency = Actual Output / Effective Capacity × 100

These measures must be interpreted carefully. A plant can show high utilisation of design capacity while failing to achieve customer requirements if the relevant effective capacity is lower. Conversely, low utilisation may be economically rational when demand is seasonal or when spare capacity is deliberately retained for flexibility.

Capacity is also time-dependent. Short-term capacity may be increased through overtime, additional shifts, subcontracting, temporary labour or reduced setup time. Long-term capacity normally requires capital investment, facility expansion, process redesign or technology change. The appropriate decision depends on demand uncertainty, lead-time requirements, investment cost and the strategic importance of internal capability.

A major analytical principle is that system capacity is governed by the constraining stage. If four linked operations can produce 80, 100, 65 and 90 units per hour respectively, the line’s effective theoretical flow is approximately 65 units per hour, assuming material availability and balanced scheduling. Increasing the capacity of the 100-unit operation does not increase total output unless the constraint is changed.

Identifying and Managing Constraints

A constraint is anything that prevents an organisation from achieving a higher level of performance relative to its goal. In a manufacturing context it may be a machine, labour skill, material, policy, information flow, market condition or quality requirement. A constraint is therefore broader than a bottleneck, although a bottleneck is a common physical form of constraint.

The Theory of Constraints provides a disciplined improvement logic known as the five focusing steps:

1. Identify the constraint. Determine which resource or policy most limits system throughput.

2. Exploit the constraint. Obtain maximum useful output from the existing constraint without major investment. Examples include reducing downtime, ensuring quality at source, prioritising valuable orders and eliminating unnecessary interruptions.

3. Subordinate other activities. Align non-constraint resources with the constraint’s requirements. Producing more than the constraint can process merely increases work-in-process inventory.

4. Elevate the constraint. If the constraint remains limiting, add capacity through overtime, outsourcing, additional equipment, redesign or investment.

5. Repeat the process. Once one constraint is broken, another will normally emerge. Improvement is continuous rather than a one-time project.

This sequence challenges the common practice of improving every workstation simultaneously. If a non-bottleneck machine is improved by 20 per cent while the bottleneck remains unchanged, system throughput may not change. The improvement may even increase inventory, congestion and handling cost.

A practical constraint study begins with a process map showing routing, processing times, setup requirements, queue levels, downtime and quality losses. Capacity should be measured in the product mix actually demanded, not merely in standard hours. The constraint may shift by product family, shift, season or maintenance condition. For this reason, constraint management requires regular review.

Throughput Accounting and Economic Decisions

Traditional cost accounting often allocates fixed overhead to products and encourages managers to maximise local efficiency or machine utilisation. Such measures can produce undesirable decisions, such as manufacturing items that do not generate cash contribution or building inventory simply to absorb fixed costs.

Throughput accounting, associated with TOC, focuses on the rate at which the system generates money through sales. Its central measures are:

• Throughput (T): sales revenue minus totally variable cost, commonly direct material cost.
• Operating Expense (OE): money spent to convert inventory into throughput, including labour, supervision, utilities and depreciation.
• Investment (I): money tied up in inventory, equipment and other assets required to operate.

A simplified system relationship is:

Net Profit = Throughput − Operating Expense

Return on Investment = (Throughput − Operating Expense) / Investment

When a bottleneck exists, decisions should consider throughput per unit of constrained resource. Suppose Product A yields a contribution of $40 and requires two bottleneck hours; its throughput per constraint hour is $20. Product B yields $30 and requires one bottleneck hour; its throughput per constraint hour is $30. If market demand exists for both, Product B should receive priority, even though its unit contribution is lower.

Throughput accounting is not a replacement for all financial accounting. It is a decision-support perspective, particularly useful for product mix, overtime, outsourcing, process investment and scheduling. It also requires care: the definition of totally variable cost must be consistent, demand must be considered, and quality, strategic capability and customer commitments must not be ignored.

Measuring Asset Effectiveness: OEE

Overall Equipment Effectiveness measures how effectively a specified asset performs relative to its ideal productive potential. OEE combines three factors:

OEE = Availability × Performance × Quality

Availability reflects time lost to failures, setup, adjustment and other stoppages:

Availability = Operating Time / Planned Production Time

Performance reflects speed losses and minor stoppages:

Performance = (Ideal Cycle Time × Total Count) / Operating Time

Quality reflects the proportion of good output:

Quality = Good Count / Total Count

For example, a machine has 480 minutes of planned production time. It loses 60 minutes to breakdowns and setups, produces 800 units during the remaining 420 minutes, has an ideal cycle time of 0.5 minutes per unit, and produces 32 defective units. Availability is 420/480 = 87.5 per cent. Performance is (0.5 × 800)/420 = 95.2 per cent. Quality is 768/800 = 96 per cent. OEE is approximately 80.0 per cent.

OEE is valuable because it separates different sources of loss. A low score may result from poor maintenance, excessive changeovers, slow operating speed or defects. The aggregate number alone is insufficient; improvement requires analysis of the underlying losses. OEE should also be used for comparable assets and clearly defined time bases. Comparing a simple repetitive machine with a highly flexible, frequently changing process without adjustment may lead to misleading conclusions.

OEE is an equipment-centred measure. It does not directly measure labour productivity, schedule adherence, customer service, inventory or profitability. A high OEE on a machine producing unwanted inventory is not necessarily good performance. OEE is most useful when connected to bottleneck status, throughput and business priorities.

Maintenance Strategies and Asset Life Cycles

Maintenance is the set of technical, managerial and administrative actions intended to retain or restore an asset to a condition in which it can perform its required function. The correct maintenance policy depends on failure behaviour, consequence of failure, detectability, repairability and economics.

Corrective maintenance restores equipment after failure. It may be appropriate for low-cost, non-critical assets whose failure has limited consequences. Its risk is uncontrolled downtime and secondary damage.

Preventive maintenance is performed at predetermined intervals, such as replacement after a specified number of operating hours. It can reduce age-related failures but may also cause unnecessary intervention if the replacement interval is not technically justified.

Predictive or condition-based maintenance uses evidence about asset condition—such as vibration, temperature, oil quality, acoustic emission or electrical signatures—to determine when intervention is required. It is particularly useful when degradation can be detected before functional failure.

Proactive maintenance addresses the underlying causes of failure through redesign, precision installation, lubrication control, contamination prevention and operating discipline.

Reliability-centred maintenance (RCM) begins with the functions required from an asset, identifies functional failures and failure modes, evaluates their consequences, and selects technically and economically appropriate tasks. It avoids the assumption that every machine should receive the same preventive schedule.

TPM expands maintenance responsibility beyond the maintenance department. Operators undertake routine cleaning, inspection, lubrication and early abnormality detection, while maintenance specialists handle technical restoration, planning and improvement. TPM commonly includes focused improvement, planned maintenance, quality maintenance, training and safety. Its cultural contribution is as important as its technical practices: equipment condition becomes a shared production responsibility.

Reliability Economics and Replacement Decisions

Reliability concerns the probability that an asset performs its required function for a specified time under stated conditions. Availability depends not only on reliability but also on maintainability and support resources. A highly reliable asset that takes weeks to repair may have lower operational availability than a moderately reliable asset that can be restored quickly.

Important indicators include:

• Mean Time Between Failures (MTBF): average operating time between inherent failures.
• Mean Time To Repair (MTTR): average active repair time.
• Inherent availability: MTBF / (MTBF + MTTR).
• Failure rate: frequency of failures over a defined operating period.

Maintenance economics should consider life-cycle cost rather than purchase price alone. Life-cycle cost includes acquisition, installation, energy, labour, spares, planned and unplanned downtime, quality losses, safety consequences, disposal and residual value.

A replacement decision should compare the expected cost of retaining an existing asset with the cost of acquiring and operating an alternative. Relevant questions include whether failure frequency is increasing, whether spare parts remain available, whether the asset limits throughput, whether a new asset reduces energy and quality costs, and whether technological obsolescence creates strategic risk. Discounted cash-flow analysis may be used for major investments, while simpler annual-equivalent-cost comparisons may support routine decisions.

A low-cost repair is not necessarily economical if it restores an unreliable asset that repeatedly disrupts a bottleneck. Conversely, replacing an asset merely because it is old may destroy useful economic life. The correct decision is based on future expected cost and operational consequence.

An Integrated Management Framework

Capacity, constraints, maintenance and productivity should be managed as a connected control system. A useful managerial cycle is:

1. Translate demand and customer requirements into required capacity by product and time period.

2. Map the process and identify the resource that limits required throughput.

3. Measure the constraint’s availability, speed, quality, queue and downtime losses.

4. Protect the constraint through maintenance, quality at source, appropriate buffers and disciplined scheduling.

5. Prioritise improvement using throughput per constraint hour and life-cycle economics.

6. Recalculate capacity after each major change because the constraint may move.

7. Review financial, operational, safety and employee outcomes together.

Buffers have an important role in this framework. A time or material buffer before a constraint protects it from starvation; a finished-goods buffer may protect customer service. Buffers should not become excuses for poor process control. Their size should reflect variability, replenishment time and the cost of interruption.

Illustrative Case: The Constrained Heat-Treatment Cell

Background

The following is an illustrative case. A manufacturer of industrial valves operates machining, washing, heat treatment and final assembly. Heat treatment is performed in one furnace cell and is required for most high-margin products.

Situation and Problem

Customer demand has increased, but monthly shipments are consistently below the plan. The machining department reports high utilisation, while the heat-treatment cell experiences frequent waiting, changeovers and unplanned stoppages. Finance proposes purchasing another machining centre because machining appears to have the highest unit cost. Operations management suspects that the furnace cell is the true constraint.

Analysis

The furnace has 600 planned production hours per month. Breakdown and adjustment losses consume 72 hours, leaving 528 operating hours. Its ideal output is 10 batches per hour, but speed losses reduce effective production to 9 batches per hour. Of 4,752 processed batches, 190 require rework or are rejected. The approximate OEE is:

Availability = 528/600 = 88 per cent.

Performance = (10 × 4,752)/5,280 = 90 per cent.

Quality = 4,562/4,752 = 96 per cent.

OEE = 88 × 90 × 96 = 76.0 per cent approximately.

The team also calculates throughput per furnace hour. Product X generates $1,200 throughput per batch and requires 1.5 furnace hours, or $800 per furnace hour. Product Y generates $900 per batch and requires 0.75 furnace hours, or $1,200 per furnace hour. Demand exists for both products, but the schedule gives excessive furnace time to Product X because its unit contribution is higher.

Approach and Solution

The company applies the five focusing steps. First, the furnace is confirmed as the constraint. Second, it is exploited by standardising loading patterns, performing inspections before the shift, moving non-essential cleaning outside production time, reducing changeover duration and giving priority to products with higher throughput per furnace hour. A small buffer of approved work is maintained before the furnace to prevent starvation.

Third, machining and washing are subordinated to the furnace schedule rather than producing at maximum local utilisation. Fourth, the company elevates the constraint by outsourcing selected low-margin heat-treatment work during peak periods and evaluating an additional furnace only after short-term losses have been removed. TPM routines are introduced, including operator cleaning, temperature checks and abnormality tags. Condition monitoring is added to furnace fans and burners.

Outcome

The immediate improvement comes not from buying equipment but from increasing useful furnace availability, improving loading discipline and changing the product sequence. The case demonstrates that the apparent cost centre is not necessarily the economic constraint. Investment is considered only after the organisation has exploited existing capacity and validated future demand.

Learning Points

The case highlights five lessons. First, system throughput is governed by the constraint. Second, local utilisation can conflict with total performance. Third, OEE identifies losses but does not by itself determine product priority. Fourth, maintenance and scheduling decisions must protect the bottleneck. Fifth, capital expenditure should follow disciplined constraint analysis rather than precede it.

Digital Technologies and Emerging Practice

Industrial Internet of Things technologies enable continuous collection of vibration, temperature, energy, cycle-time and quality data. Analytics can detect abnormal patterns, estimate remaining useful life and recommend maintenance windows. Digital dashboards can combine OEE, downtime codes, maintenance work orders, production schedules and throughput information.

Artificial intelligence can assist anomaly detection and predictive maintenance, but its value depends on data quality, sensor reliability, failure history and organisational response. A prediction that maintenance personnel cannot act upon has little operational value. Digital systems should therefore be integrated with standard work, spare-parts planning, skills development and escalation rules.

Digital twins and simulation can evaluate capacity alternatives, buffer sizes, maintenance schedules and product mixes before physical implementation. Energy monitoring is also becoming central to asset productivity because a machine may appear productive while consuming excessive energy during idle, warm-up or low-load operation.

Emerging practice is moving from isolated machine metrics toward asset health management across the life cycle. This includes connected equipment, remote diagnostics, service-based contracts, circular-economy principles, remanufacturing and design for maintainability. Cybersecurity and data governance become increasingly important as production assets are connected to enterprise and external networks.

Challenges, Risks and Common Mistakes

Several errors repeatedly weaken capacity and maintenance programmes. One is treating utilisation as the primary objective for every machine. Another is calculating OEE with inconsistent definitions or using planned downtime selectively to improve reported performance. A third is performing preventive maintenance at arbitrary intervals that create unnecessary interventions.

Organisations may also overinvest in sensors while neglecting basic lubrication, cleaning, calibration and failure coding. Predictive models can produce false alarms or missed failures when data are incomplete. Excessive focus on mean measures may conceal rare but catastrophic failures. Safety-critical assets require consequence-based controls even when their average downtime is low.

There are organisational risks as well. Production and maintenance may compete over equipment access, managers may reward output at the expense of asset deterioration, and operators may be blamed for failures caused by poor design. Effective governance requires shared performance measures, clear ownership, competent personnel and transparent escalation.

Ethically and professionally, production managers must not bypass safety interlocks, defer legally required inspections or continue operating assets that create unacceptable risks. Maintenance records should be accurate because they support safety, regulatory compliance, warranty claims and future investment decisions. Environmental requirements, including leakage control, emissions, energy use and responsible disposal of oils and components, are part of responsible asset management.

Best-Practice Principles

A mature asset-productivity system generally follows these principles:

• Manage the flow of the entire system, not isolated departmental utilisation.
• Identify and protect the current constraint.
• Use OEE diagnostically, not as a simplistic ranking score.
• Combine operator care with specialist maintenance expertise.
• Select maintenance policies according to failure consequences and evidence.
• Evaluate assets using life-cycle cost and throughput impact.
• Integrate production, maintenance, quality, finance and safety data.
• Standardise downtime definitions, data collection and review routines.
• Treat improvement as a repeating cycle because constraints migrate.
• Ensure that productivity gains do not compromise safety, quality, sustainability or workforce dignity.

Chapter Summary

Capacity management establishes what a production system can deliver, but effective capacity is limited by the most restrictive resource or policy. The Theory of Constraints directs managers to identify, exploit, subordinate and elevate the constraint, then repeat the process when the constraint moves. Throughput accounting supports decisions by focusing on throughput, operating expense, investment and throughput per unit of constrained resource.

OEE integrates availability, performance and quality to diagnose equipment losses. It is most useful when linked to bottleneck analysis and financial outcomes. TPM creates shared responsibility for equipment condition, while preventive, predictive, proactive and reliability-centred maintenance provide alternative approaches based on failure behaviour and consequence. Reliability economics extends the analysis over the asset life cycle and supports repair, replacement and investment decisions.

The central managerial lesson is that asset productivity is a system outcome. High machine utilisation, low maintenance cost or impressive OEE is not automatically valuable unless it improves safe, reliable and profitable customer-serving flow.

Key Terms

Capacity: Maximum output rate achievable by a resource or system under specified conditions.

Design Capacity: Theoretical maximum capacity under ideal operating conditions.

Effective Capacity: Realistic maximum capacity under normal operating conditions.

Bottleneck: A resource whose capacity restricts the output of the overall process.

Constraint: Any factor that prevents a system from achieving a higher level of performance.

Theory of Constraints: A management approach that focuses improvement on the system’s limiting factor.

Throughput: Sales revenue less totally variable cost, usually direct material cost.

Operating Expense: Resources consumed to convert inventory into throughput.

Investment: Money tied up in inventory, equipment and operating assets.

OEE: Product of availability, performance and quality for a defined asset and time period.

Availability: Proportion of planned production time during which equipment is operating.

Performance: Comparison between actual operating speed and ideal speed.

Quality: Proportion of total output that meets requirements without defect.

TPM: A company-wide approach that involves operators and specialists in maintaining and improving equipment.

Reliability: Probability that an asset performs its required function for a stated time and condition.

Maintainability: Ease and speed with which an asset can be restored after failure.

MTBF: Mean operating time between failures.

MTTR: Mean active time required to repair an asset.

Life-Cycle Cost: Total cost of acquiring, operating, maintaining and disposing of an asset.

Condition-Based Maintenance: Maintenance triggered by evidence of actual equipment condition.

Review Questions

Short Answer Questions

1. Distinguish design capacity, effective capacity and actual output.

2. What are the five focusing steps of the Theory of Constraints?

3. Define throughput, operating expense and investment in throughput accounting.

4. State the formula for OEE and identify its three components.

5. Explain the difference between preventive and condition-based maintenance.

Long Answer Questions

1. Explain why improving a non-bottleneck resource may fail to improve total system throughput.

2. Discuss how OEE should be interpreted and explain its limitations as a management measure.

3. Compare corrective, preventive, predictive, proactive and reliability-centred maintenance.

4. Explain how throughput accounting can influence product-mix and capital-investment decisions.

5. Discuss the role of TPM in integrating production operators, maintenance specialists and management.

Analytical / Application Questions

1. A line has four stages with capacities of 90, 75, 110 and 80 units per hour. Identify the constraint and recommend two actions that follow the first two focusing steps.

2. A machine has planned production time of 500 minutes, operating time of 425 minutes, ideal cycle time of 0.4 minutes, total output of 950 units and 38 defective units. Calculate availability, performance, quality and OEE.

3. Product A generates $50 throughput and uses 2.5 hours of bottleneck time. Product B generates $36 throughput and uses 1.5 hours. If demand exists for both, determine the priority and explain any additional information required before finalising the schedule.

22. Multiple Choice Questions with Answers

Question 1

Which measure most directly identifies the economic priority of products when a bottleneck exists?

A. Unit selling price

B. Throughput per unit of constrained resource

C. Allocated fixed cost per unit

D. Machine utilisation of the non-bottleneck process

Correct Answer: B. Throughput per unit of constrained resource

Question 2

A system has stage capacities of 120, 95, 110 and 100 units per hour. Which stage initially constrains system flow?

A. The first stage

B. The second stage

C. The third stage

D. The fourth stage

Correct Answer: B. The second stage

Question 3

If availability is 90%, performance is 80% and quality is 95%, the approximate OEE is:

A. 68.4%

B. 72.0%

C. 76.0%

D. 85.5%

Correct Answer: A. 68.4%

Question 4

Which action best represents exploiting a constraint?

A. Purchasing equipment for every department

B. Keeping the bottleneck supplied with approved work and reducing its avoidable stoppages

C. Producing maximum inventory at non-bottleneck stations

D. Reducing inspection at the bottleneck regardless of quality risk

Correct Answer: B. Keeping the bottleneck supplied with approved work and reducing its avoidable stoppages

Question 5

Throughput accounting defines throughput primarily as:

A. Sales revenue minus totally variable cost

B. Sales revenue minus all manufacturing costs

C. Standard cost multiplied by output

D. Gross profit divided by total assets

Correct Answer: A. Sales revenue minus totally variable cost

Question 6

Which maintenance policy uses vibration or temperature evidence to determine when intervention is required?

A. Corrective maintenance

B. Calendar-only maintenance

C. Condition-based maintenance

D. Emergency replacement only

Correct Answer: C. Condition-based maintenance

Question 7

A machine produces 1,000 units, of which 40 are defective. What is its quality rate?

A. 4%

B. ?

C. 96%

D. 104%

Correct Answer: C. 96%

Question 8

The main purpose of subordinating non-constraints in TOC is to:

A. Maximise local utilisation at every workstation

B. Align other activities with the capacity and requirements of the constraint

C. Eliminate all inventory from the system

D. Increase product variety regardless of bottleneck load

Correct Answer: B. Align other activities with the capacity and requirements of the constraint

Question 9

Which factor is most important in a life-cycle cost decision?

A. Purchase price alone

B. Future operating, maintenance, downtime and disposal costs

C. Supplier brand recognition alone

D. Book value at the date of purchase

Correct Answer: B. Future operating, maintenance, downtime and disposal costs

Question 10

A high OEE on a non-bottleneck machine may fail to improve business performance because:

A. OEE never includes quality

B. The machine may be producing output that the system constraint cannot process

C. Availability is always excluded from OEE

D. High OEE automatically increases operating expense

Correct Answer: B. The machine may be producing output that the system constraint cannot process