Maintenance Economics, Analytics and Managerial Decisions

Introduction and context

Total Productive Maintenance (TPM) is often presented as a technical system for improving equipment availability, performance and quality. At postgraduate level, however, maintenance must also be understood as an economic decision system. Every intervention consumes labour, materials, energy, production capacity and managerial attention. Conversely, insufficient or poorly timed maintenance creates failures, safety exposure, quality losses, environmental consequences and customer-service problems. The central managerial question is therefore not whether maintenance is costly, but whether a particular maintenance policy creates more value than its alternatives over the asset life.

Maintenance economics integrates reliability engineering, operations management, financial analysis, risk management and strategic decision-making. It supports decisions such as:

• How much should be allocated to preventive and predictive maintenance?
• What is the full economic cost of downtime?
• Should an asset be repaired, upgraded, replaced or outsourced?
• When should a replacement occur?
• Which improvement projects should be funded when capital is constrained?
• How should safety, environmental and reputational risks be included in financial analysis?

A narrow accounting view records maintenance expenditure. A managerial view evaluates lifecycle value, uncertainty and consequences. The latter is essential because the lowest immediate cost may produce the highest total cost.

Learning objectives

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

1. Distinguish maintenance expenditure, downtime cost, risk cost and lifecycle cost.
2. Construct a maintenance budget linked to asset criticality and business objectives.
3. Evaluate maintenance alternatives using present value, NPV, ROI and equivalent annual cost.
4. Analyse repair, replacement, upgrade and outsourcing decisions.
5. Incorporate uncertainty, risk, sensitivity and scenario analysis into maintenance decisions.
6. Prioritise improvement portfolios under financial and operational constraints.
7. Critically evaluate the limitations and governance requirements of maintenance analytics.

1. The economic logic of maintenance

1.1 Cost categories

A useful economic model separates at least five categories of cost:

1. Direct maintenance cost: labour, spares, contractors, tools, inspections and maintenance support.
2. Downtime cost: lost contribution, restart losses, labour waiting, scrap, expediting and customer penalties caused by unavailability.
3. Risk cost: the expected economic effect of safety incidents, environmental releases, legal claims, regulatory action and reputational damage.
4. Capital cost: acquisition, installation, commissioning, modification and financing of equipment.
5. Residual or end-of-life cost: decommissioning, disposal, site restoration and salvage value.

The distinction between cost and cash flow is important. Depreciation is an accounting allocation, not normally a cash payment. It may affect tax, but it should not be counted as a cash maintenance cost unless the analysis specifically concerns accounting profit.

A basic annual maintenance cost model is:

Here, each term is an annual cost in currency units. The model is useful for comparison, but it must be adapted to the decision. For example, energy cost may be relevant to an equipment replacement decision but not to a short-term lubrication policy.

1.2 Downtime cost

Downtime cost is frequently underestimated because production value is confused with profit. If a failed machine prevents production that can be recovered later without overtime, the true short-run loss may be limited to labour, restart and quality costs. If demand is lost permanently, the cost may approximate the lost contribution margin.

A practical formulation is:

where:

• t_d = downtime duration, usually hours;
• CM = contribution margin lost per hour;
• L_idle = cost of idle labour and support resources per hour;
• C_restart = restart and stabilisation cost;
• C_scrap = quality loss associated with the event;
• C_penalty = contractual or service penalty per hour; and
• V_recovered = value recovered through rescheduling, inventory or alternate capacity.

The terms must not be double-counted. For example, if the contribution margin already includes normal labour, idle labour should not be added again.

If failures occur randomly with an expected frequency of λ events per year and mean downtime per event is τ, expected downtime is:

This is an expected-value relationship, not a guarantee. It is appropriate for planning when event frequency and duration are reasonably stable. It becomes unreliable when failures are strongly dependent, seasonal or affected by changing operating conditions.

Worked example 1: estimating downtime cost

A bottling line has a contribution margin of $1,800 per operating hour. During a failure, idle labour costs $220 per hour, and restart and scrap losses average $2,400 per event. The line experiences six failures per year, each lasting 2.5 hours. Rescheduling recovers $500 per event.

Expected annual downtime cost is:

This estimate assumes that the contribution margin is genuinely lost and that the recovery value is not already reflected in the margin. If production can be fully recovered without customer or overtime consequences, the estimate would be materially lower.

2. Lifecycle costing and the time value of money

Maintenance decisions extend across years, so comparing nominal costs in different years is misleading. A dollar today can be invested and therefore is worth more than a dollar received several years later. Discounted cash-flow analysis converts future cash flows into present value.

The present value of a future cash flow is:

where CF_t is the net cash flow in year t, r is the annual discount rate expressed as a decimal, and t is the number of years from the base date. The formula assumes end-of-period cash flows and a constant discount rate.

Net present value is:

where C₀ is the initial cash outflow and n is the analysis horizon. A project is financially attractive relative to a zero-value alternative when NPV is positive. For mutually exclusive alternatives, the preferred choice is normally the one with the highest NPV, subject to risk, capacity and strategic constraints.

Lifecycle cost should include acquisition, operation, maintenance, downtime, energy, modification, training, compliance and disposal. A general lifecycle cost expression is:

Here SV is salvage value. The expression is a decision model rather than a universally complete accounting identity; relevant cost categories depend on the asset and boundary of analysis.

Equivalent annual cost

When alternatives have unequal lives, simply comparing their NPVs may be unfair. Equivalent annual cost (EAC) converts a present value into a uniform annual amount over the asset life:

The bracketed term is the capital-recovery factor. EAC assumes a constant discount rate and comparable renewal or replacement conditions. It is useful for comparing machines that provide equivalent service but have different useful lives.

Worked example 2: repair or replacement

An existing machine can be retained for three years. Its immediate repair costs $30,000, followed by annual maintenance and downtime costs of $42,000, $48,000 and $55,000. It has no salvage value at the end. A replacement costs $190,000 immediately, has annual maintenance and downtime costs of $18,000 for three years, and has a salvage value of $35,000 at the end of year three. Use a discount rate of 10%.

The present value of retaining the machine is:

The present value of replacement is:

On this three-year boundary, retaining the machine has the lower present cost by approximately $64,721. This conclusion does not mean that retention is always preferable. The replacement may have a longer life, greater capacity, lower safety risk or better strategic fit. A complete analysis should use a common service horizon or EAC when asset lives differ.

3. Budgeting maintenance as a managerial control system

A maintenance budget should not be an arbitrary percentage of replacement asset value. It should reflect workload, failure exposure, regulatory obligations, asset criticality, operating context and improvement plans.

3.1 Budget architecture

A robust budget separates:

• planned preventive and predictive work;
• corrective work and emergency response;
• statutory inspection and compliance;
• spares and inventory carrying cost;
• contractor and specialist services;
• condition-monitoring technology and data infrastructure;
• training and competency development;
• shutdown or turnaround expenditure; and
• reliability improvement and capital projects.

The budget should distinguish controllable from uncontrollable costs. A maintenance manager may control planning efficiency but not a sudden regulatory inspection requirement. Variance analysis should therefore investigate causes rather than punish all overspending equally.

3.2 Budgeting methods

Zero-based or activity-based budgeting estimates resources from the maintenance plan, labour standards, job plans and known inspections. It improves transparency but requires reliable work history.

Driver-based budgeting links cost to operating drivers such as production hours, equipment population, failure rate, number of statutory inspections or planned shutdown scope.

Risk-based budgeting assigns greater resources to assets whose failure consequences exceed accepted risk. It is especially appropriate for process industries, utilities, healthcare and transport.

Rolling forecasts update expectations as failure data, production plans and supplier conditions change. They are preferable to treating an annual budget as a fixed prediction.

A useful performance measure is not simply maintenance cost per unit. It is maintenance cost relative to the value and risk controlled. Examples include maintenance cost per operating hour, planned-work percentage, emergency-work cost, cost per unit of availability and avoided-loss value from improvement projects.

4. Cost–risk–performance trade-offs

Maintenance decisions usually involve three competing dimensions:

• Cost: expenditure and resource consumption;
• Performance: availability, throughput, quality, energy efficiency and delivery reliability; and
• Risk: probability and consequence of undesirable events.

Reducing preventive maintenance may lower direct cost but increase failure probability. Increasing inspection frequency may reduce uncertainty but consume labour and cause intrusive downtime. A technically optimal policy is therefore not necessarily economically optimal.

Expected risk cost can be represented as:

where p_i is the probability of consequence scenario i, C_i is its monetary consequence, and m is the number of scenarios. This is an expected-value model; it does not adequately represent catastrophic aversion, correlated events or moral obligations by itself.

A risk matrix can support screening, but colour-coded matrices should not substitute for quantitative analysis where credible estimates exist. Low-probability, high-consequence events may be understated by ordinal scoring. Safety and environmental decisions may also be constrained by law or company risk appetite, making financial optimisation impermissible.

Reliability-centred decision logic

Reliability-centred maintenance asks what functions an asset performs, how it can fail, what causes failure, what consequences follow and which task is technically and economically justified. The principal task choices are:

• condition-based task;
• scheduled restoration or replacement;
• failure-finding task for hidden protective functions;
• redesign or engineered modification; or
• run-to-failure where consequences are acceptable and repair is economical.

The economic analysis must respect technical applicability. Scheduled replacement is not automatically effective for every failure mode; it is most defensible where failure probability increases predictably with age and the task restores useful reliability.

5. Replacement timing and renewal decisions

Replacement timing is a dynamic decision. Keeping an old asset avoids immediate capital expenditure but may increase operating cost, failure risk and obsolescence. Replacing too early sacrifices remaining useful life and creates transition costs.

Relevant signals include:

• increasing corrective maintenance and downtime;
• declining efficiency or quality capability;
• spare-parts obsolescence;
• loss of vendor support or cybersecurity exposure;
• inability to meet capacity or regulatory requirements;
• residual value and market availability; and
• the opportunity cost of capital.

A replacement model compares the present cost of retaining the current asset for one or more years with the cost of replacing now or later. If the equipment is expected to be renewed indefinitely, the equivalent annual cost or an infinite-horizon renewal model is more appropriate than a single short horizon.

Managerial judgement remains important. A mathematically optimal replacement date may be unacceptable if it creates a concentrated shutdown, depends on an unreliable supplier or exposes the organisation to unacceptable safety risk.

6. Capital investment, ROI and portfolio selection

6.1 NPV and ROI

Return on investment is commonly stated as:

ROI is easy to communicate but ambiguous unless the time period, treatment of taxes, depreciation, working capital and risk are specified. It is not a substitute for NPV because it ignores the timing of cash flows.

For maintenance improvements, net benefit may include avoided downtime, reduced spares, lower energy use, reduced contractor cost and quality improvement. Benefits should be incremental: the analysis compares the future with the project against the credible future without it.

Payback period is the time required for cumulative net cash benefits to recover the initial investment. It is useful as a liquidity screen but ignores benefits after payback and, in its simple form, the time value of money.

6.2 Portfolio decisions under capital constraints

Maintenance improvement projects compete for limited funds. A portfolio should balance quick-return projects with strategic, compliance and resilience investments. Ranking solely by ROI can favour small projects and exclude essential risk-reduction work.

A practical portfolio process is:

1. Define the business problem and baseline.
2. Verify the failure mechanism and benefit logic.
3. Estimate capital, operating cost and implementation disruption.
4. Calculate NPV, risk-adjusted benefit and sensitivity.
5. Identify dependencies, resource requirements and regulatory constraints.
6. Select a portfolio rather than isolated projects.
7. Establish benefit ownership and post-implementation review.

Where projects are divisible and independent, a benefit-to-investment ratio can assist screening. Where projects interact, such as common sensors, shared shutdowns or mutually exclusive equipment choices, a formal optimisation or scenario comparison is required.

7. Uncertainty, sensitivity and decision analysis

Maintenance data are uncertain. Failure rates may be estimated from small samples; downtime consequences vary with demand; benefits may depend on operator adoption. A single-point NPV can create false precision.

7.1 Sensitivity analysis

Sensitivity analysis changes one input at a time or several inputs systematically to identify value drivers. Typical variables include discount rate, failure frequency, repair duration, energy price, project cost, benefit realisation and equipment life.

A break-even value is particularly useful. For example, an improvement is justified if the annual avoided cost exceeds the annualised investment. If the annualised investment is $60,000 and direct benefits are $35,000, the project requires at least $25,000 of additional verified annual benefit to break even.

7.2 Scenario analysis

Scenario analysis evaluates coherent combinations of assumptions, such as low demand, normal demand and capacity-constrained demand. It is superior to arbitrary independent changes when variables are linked.

7.3 Expected monetary value and decision trees

If alternatives have mutually exclusive uncertain outcomes, expected monetary value is:

where P_j is the probability of outcome j, V_j is its value, and probabilities sum to one. EMV is appropriate for repeated decisions or risk-neutral comparisons. It should not be used mechanically for irreversible, catastrophic decisions where organisational risk appetite is not risk-neutral.

Decision trees are useful when decisions occur sequentially. For example, an organisation may first install condition monitoring, observe evidence, and then decide whether to refurbish or replace. The value of information can be assessed by comparing the expected value with and without the additional information, less the cost of obtaining it.

8. Outsourcing and contracting decisions

Outsourcing is not simply a comparison between internal wages and contractor invoices. The relevant comparison is total economic cost and capability risk.

Internal provision may include labour, supervision, training, tools, stores, scheduling, systems, management overhead, standby capacity and employee benefits. Outsourcing may include mobilisation, contract management, travel, profit margin, minimum call-out charges, knowledge transfer, switching costs and vendor lock-in.

A sound make-or-buy analysis should address:

• critical knowledge and asset-specific competence;
• response time and service-level requirements;
• contractor safety performance and financial resilience;
• data ownership and cybersecurity;
• quality assurance and regulatory accountability;
• dependence on a single supplier;
• incentives created by the contract; and
• reversibility if the arrangement fails.

Outcome-based contracts can align payment with availability or reliability, but measurement must be robust. If a contractor is rewarded for reducing work orders without regard to risk, undesirable deferral may result. Contracts should define failure reporting, escalation, spares ownership, access to condition data and audit rights.

9. Analytics, digitalisation and AI in maintenance economics

Computerised maintenance management systems (CMMS), enterprise resource planning systems, historians, sensors and condition-monitoring platforms provide the data foundation for economic analysis. Useful analytical outputs include failure-mode cost, mean time to repair, schedule compliance, backlog ageing, spare criticality and cost by asset hierarchy.

Predictive analytics can estimate failure probability or remaining useful life, but a prediction has economic value only when it changes a decision. The value chain is:

AI may support anomaly detection, work-order classification, spare forecasting and optimisation of inspection intervals. It does not remove the need for engineering judgement. Models can fail because of biased historical maintenance, changing operating regimes, data leakage, sensor drift or confusing correlation with causation.

Governance should require traceable data, model validation, human accountability, cybersecurity, access control and periodic review. A predictive model that generates excessive false alarms may increase intrusive maintenance and reduce value. Performance should therefore be evaluated using economic and operational outcomes, not only prediction accuracy.

10. Illustrative case: economics of a packaging-line improvement portfolio

Background

The following is a fictional illustrative case. A food manufacturer operates a packaging line whose unplanned stoppages cause lost contribution, overtime and customer-service disruption. The plant has $250,000 available for reliability improvements during the financial year.

Problem

Three proposals have been prepared:

Assume end-of-year benefits, no salvage value, and a discount rate of 10%.

Analysis

For a uniform annual benefit, the present value factor is:

For Project A, PVAF at 10% for four years is approximately 3.170. Therefore:

For Project B, the five-year factor is approximately 3.791:

For Project C:

All three projects have positive NPV individually. However, A plus B costs $230,000 and produces the highest combined NPV, approximately $123,018, assuming benefits are independent. B plus C costs $210,000 and produces approximately $108,148. A plus C costs $160,000 and produces approximately $71,410. Funding all three is impossible because the total cost is $300,000.

Approach and solution

The recommended initial portfolio is A plus B, subject to a shutdown plan for the conveyor redesign and a controlled pilot for the monitoring system. The plant should not treat the calculated benefits as guaranteed. It should define baseline downtime, alarm-response compliance, avoided failures and benefit-realisation targets. Project C may be reconsidered if stockout risk is shown to be a major contributor to downtime or if additional funds become available.

The analysis should also test downside assumptions. If Project A realises only 50% of its expected benefit, its NPV becomes:

This sensitivity result shows that Project A requires strong adoption and credible benefit capture. Project B is less sensitive to benefit reduction, but its installation disruption must be included in the baseline comparison.

Learning points

1. Positive individual NPVs do not determine the best portfolio when capital is constrained.
2. Benefits must be incremental, measurable and assigned to accountable managers.
3. Sensitivity analysis identifies projects requiring pilots or safeguards.
4. Technical feasibility and implementation disruption must accompany financial metrics.
5. Risk-reduction and compliance projects may require funding even when NPV is not positive.

11. Limitations, ethics and governance

Economic models simplify reality. Common limitations include poor failure data, unstable production demand, omitted indirect costs, optimism bias, double-counting of benefits and inappropriate discount rates. Analysts should document boundaries, assumptions, data sources, uncertainty ranges and excluded effects.

Ethical issues arise when financial pressure encourages deferral of safety-critical maintenance, manipulation of downtime records or suppression of bad news. No positive ROI should justify violating statutory requirements or exposing workers and communities to unacceptable risk. Where consequences include injury or environmental harm, monetisation may support comparison but must not imply that human welfare is merely a financial variable.

Governance should include independent review of major business cases, separation of technical assurance from project sponsorship, audit trails for cost and benefit calculations, post-investment reviews and escalation rules for deteriorating risk. Maintenance analytics should support transparent decisions rather than conceal managerial preferences behind apparently objective numbers.

12. Best-practice decision architecture

An effective maintenance-economic decision follows a disciplined sequence:

1. Define the asset function, decision boundary and required service level.
2. Establish the baseline using verified operational and financial data.
3. Identify failure modes, consequences and regulatory constraints.
4. Generate technically credible alternatives.
5. Estimate lifecycle cash flows and risk-adjusted consequences.
6. Discount cash flows consistently and use a common service basis.
7. Test assumptions with sensitivity and scenario analysis.
8. Compare alternatives using NPV, EAC, risk and strategic fit.
9. Select and govern a portfolio under resource constraints.
10. Measure realised performance and update the model.

The highest-quality decisions combine quantitative analysis with engineering judgement, front-line knowledge and explicit risk governance. Maintenance economics is therefore not an exercise in reducing maintenance expenditure. It is the disciplined allocation of resources to achieve reliable, safe and economically sustainable operations.

Chapter Summary

Maintenance economics evaluates maintenance choices through total lifecycle value rather than direct expenditure alone. Downtime cost should reflect lost contribution, idle resources, restart, quality, penalties and recoverable production without double-counting. Lifecycle costing includes acquisition, operation, maintenance, downtime, risk, disposal and salvage value, with future cash flows discounted to present value.

NPV is generally the strongest primary measure for mutually exclusive investments, while EAC supports comparison of unequal lives. ROI and payback are useful communication or screening measures but do not replace discounted cash-flow analysis. Maintenance budgets should be driver-based, risk-informed and linked to work content and business objectives.

Replacement, outsourcing and improvement decisions require comparison of technically credible alternatives. Cost, risk and performance must be considered jointly, particularly where safety, environmental or regulatory consequences limit financial optimisation. Sensitivity, scenario, expected-value and decision-tree methods make uncertainty visible. Digital and AI tools can improve prediction and prioritisation, but their value depends on data quality, decision integration, human accountability and governance.

Key Terms

Lifecycle cost: Total cost of acquiring, operating, maintaining and disposing of an asset.

Downtime cost: Economic consequence of equipment unavailability.

Contribution margin: Sales revenue less variable costs attributable to output.

Expected risk cost: Probability-weighted economic consequence of uncertain events.

Present value: Current value of a future cash flow after discounting.

Net present value: Present value of benefits and costs including the initial investment.

Equivalent annual cost: Uniform annual value equivalent to an asset’s present cost.

ROI: Net return divided by investment, under a specified convention.

Payback period: Time required to recover an initial investment from cash benefits.

Risk-based budgeting: Allocation of resources according to risk exposure and consequence.

Renewal decision: Decision about retaining, replacing or upgrading an asset.

Sensitivity analysis: Examination of how results change when assumptions vary.

Scenario analysis: Evaluation of coherent alternative future conditions.

Expected monetary value: Probability-weighted value of possible outcomes.

Value of information: Benefit obtained from information that improves a decision.

Outsourcing: Transfer of selected activities to an external provider.

Benefit realisation: Verification that planned operational and financial benefits occur.

Predictive maintenance: Maintenance triggered by evidence about developing degradation.

Capital-recovery factor: Factor converting present value into an equivalent annual amount.

Maintenance portfolio: Group of improvement or investment projects selected under constraints.

Review Questions

Short Answer Questions

1. Why is contribution margin usually more appropriate than sales revenue for estimating short-run downtime cost?
2. Distinguish maintenance expenditure from lifecycle cost.
3. State two limitations of simple payback analysis.
4. What is the purpose of equivalent annual cost?
5. Why should predictive-maintenance models be evaluated by decision outcomes as well as prediction accuracy?

Long Answer Questions

1. Explain how a risk-based maintenance budget can be constructed and governed.
2. Compare NPV, ROI, payback period and EAC as tools for maintenance investment decisions.
3. Discuss the technical and economic conditions under which run-to-failure, preventive and predictive maintenance may each be appropriate.
4. Develop a framework for deciding whether a maintenance activity should be outsourced.
5. Critically evaluate the ethical and governance risks of using financial optimisation for safety-critical maintenance.

Analytical / Application Questions

1. A machine experiences eight failures per year, each lasting 1.75 hours. Contribution margin is $1,250 per hour, idle labour is $150 per hour, and restart and scrap cost $1,600 per failure. No production is recoverable. Calculate expected annual downtime cost and explain which assumptions should be validated.
2. An improvement costs $120,000 and produces annual net benefits of $38,000 for five years. Using a 10% discount rate, calculate its NPV and evaluate how the decision changes if benefits are only 60% of forecast.
3. A plant must choose between retaining an old asset, replacing it immediately or outsourcing its function. Design a decision model that includes lifecycle cash flow, capacity, safety risk, supplier dependence, uncertainty and reversibility.

Multiple Choice Questions with Answers

Question 1

Which cost best represents the short-run economic value of production lost during an equipment failure?

A. Total sales revenue per hour

B. Contribution margin per hour, adjusted for recoverable production

C. Annual depreciation expense

D. Book value of the equipment

Correct Answer: B. Contribution margin per hour, adjusted for recoverable production

Question 2

What does NPV primarily account for that simple payback does not?

A. The number of maintenance technicians

B. The physical dimensions of the asset

C. The timing and discounting of cash flows

D. The colour of the risk matrix

Correct Answer: C. The timing and discounting of cash flows

Question 3

An asset has a present lifecycle cost of $100,000 over four years. Which method converts this value into a uniform annual cost?

A. Equivalent annual cost

B. Mean time between failures

C. First-pass yield

D. Failure-finding analysis

Correct Answer: A. Equivalent annual cost

Question 4

If six failures occur annually and each causes 2.5 hours of downtime, expected annual downtime is:

A. 2.4 hours

B. 8.5 hours

C. 15 hours

D. 30 hours

Correct Answer: C. 15 hours

Question 5

Which statement about ROI is correct?

A. It always accounts for the time value of money

B. It is meaningful only when its time period and conventions are specified

C. It is identical to NPV in all investment decisions

D. It excludes all maintenance benefits

Correct Answer: B. It is meaningful only when its time period and conventions are specified

Question 6

What is a major purpose of sensitivity analysis in maintenance investment appraisal?

A. To eliminate all uncertainty from the decision

B. To identify assumptions that strongly influence the result

C. To replace technical feasibility assessment

D. To guarantee that a project will succeed

Correct Answer: B. To identify assumptions that strongly influence the result

Question 7

Which factor is especially important in an outsourcing decision but may be missed by comparing wages with contractor invoices?

A. Asset-specific knowledge and supplier dependence

B. The colour of maintenance uniforms

C. The historical purchase price alone

D. The number of product brands in the market

Correct Answer: A. Asset-specific knowledge and supplier dependence

Question 8

A positive NPV for each project does not necessarily mean all projects should be funded because:

A. NPV cannot be used for maintenance projects

B. Capital and implementation resources may be constrained

C. Positive NPV means benefits are guaranteed

D. Projects with positive NPV have no risk

Correct Answer: B. Capital and implementation resources may be constrained

Question 9

Expected monetary value is calculated by:

A. Adding all possible outcomes without probabilities

B. Multiplying each possible value by its probability and summing

C. Dividing initial cost by asset age

D. Using only the worst-case outcome

Correct Answer: B. Multiplying each possible value by its probability and summing

Question 10

Why must safety-critical maintenance not be governed solely by financial optimisation?

A. Safety consequences may be constrained by legal, ethical and risk-acceptance requirements

B. Safety has no relationship to maintenance

C. Financial analysis always overestimates safety benefits

D. Safety-critical assets never require budgets

Correct Answer: A. Safety consequences may be constrained by legal, ethical and risk-acceptance requirements