Overhaul Versus Replacement Decision Tool

Restore the right asset. Replace the right limitation. Compare the complete cost.

An air compressor should be overhauled when the identified damage is repairable, the remaining package can support the required duty, and the complete restoration pathway offers acceptable reliability and lifecycle value. Replacement becomes stronger when repair cannot restore the required function, major supportability gaps remain, future duty has changed, or a properly installed replacement has a better evidence-based whole-life outcome. Age, motor size and a repair-to-purchase-price percentage are not sufficient decision rules. [2, 3]

This tool turns condition findings, operating records and comparable quotations into an editable decision worksheet. It compares continued service, targeted correction, overhaul, exchange and complete replacement without confusing the price of an airend with the cost of restoring an operating compressed-air system. Its output is a documented recommendation with assumptions, unresolved evidence and an acceptance plan—not an automatic purchase instruction.

1. The decision this page owns

The existing lifecycle, airend-diagnosis and cutover pages remain the detailed engineering references. [1, 2, 3, 4, 5]

Keep three conclusions separate. First, determine whether the equipment can continue operating under the site's approved conditions. Second, determine which interventions are technically credible. Third, compare the forward costs and operational consequences of those interventions. A financially attractive option must not override an unresolved integrity defect or an inability to meet essential duty.

The least expensive quotation and the best intervention are not necessarily the same. A low overhaul price can exclude removal, transport, damaged rotors, reinstallation and testing. A replacement price can exclude electrical work, dryer changes and production continuity. The worksheet makes those differences visible before comparing totals.

2. Compare pathways, not just two quotations

Do not assume that all five pathways are available. An exchange unit is not a confirmed option until identity, supply and terms are verified. Similarly, continued operation is not a default alternative for equipment that has already been withdrawn from service. Record excluded pathways and the reason rather than manufacturing a complete-looking comparison.

3. Freeze the required duty and comparison boundary

Describe the required delivered air flow, minimum dynamic pressure, air quality, operating schedule and resilience at the same point of use for every option. Specify whether the unit is base-load, trim, standby or portable project equipment. Include committed production changes, but keep speculative expansion separate from the approved base case. [2, 4]

The comparison boundary may be the compressor package, the compressor room or the complete air service. Package boundaries are useful for equipment performance; a system boundary is needed when treatment, sequencing or distribution changes materially affect energy and continuity. State what is common and what changes.

A replacement that supplies less air cannot claim an efficiency advantage merely because it draws less power. An overhaul that restores nameplate capacity may still be unsuitable for a plant whose demand has increased. Use flow, pressure and quality together, and retain the flow reference conditions. Do not equate inlet volume, free-air delivery and actual compressed volume without the appropriate conversion basis.

4. Establish the condition evidence before pricing the remedy

Start with exact compressor and airend identification, running and loaded hours, maintenance history, lubricant history, alarms, failure events and recent configuration changes. Then relate symptoms to measurements: pressure and flow, package input power, temperature, differential pressure, oil-analysis findings, vibration and inspection results where available. [3]

A low-pressure complaint can originate in demand growth, leakage or distribution restriction. High temperature can arise from cooling or ventilation. Abnormal vibration can involve the drive, mounting or connected piping. These findings do not prove that the airend requires overhaul. Preserve the diagnostic distinction between the observed symptom, the suspected mechanism and a confirmed finding.

Record evidence quality alongside each conclusion. A signed inspection report, a comparable operating measurement and an unverified verbal estimate are not interchangeable. If dismantling is necessary to establish repairability, separate the authorised inspection stage from the final repair commitment. Specify who approves additional work after inspection and the cost of stopping at that point.

5. Apply technical eligibility before economic ranking

Use three states: eligible, ineligible and unresolved. An unresolved option may retain a clearly labelled scenario calculation, but it cannot be declared the preferred executable option. A composite score must never allow favourable price or efficiency to cancel an essential technical failure.

6. Define what the overhaul will actually restore

An overhaul quotation should identify the assembly being worked on, inspection included, mandatory and conditional replacements, retained components, cleaning, balancing or specialist processes where applicable, reassembly, installation and testing. Do not infer rotor machining, factory tolerances or an OEM-equivalent rebuild from the word “overhaul.” Exact methods and limits must be supported for the identified equipment. [3]

Separate three performance statements: the pre-intervention measured condition, the supplier's proposed recovery and the acceptance value to be demonstrated. Replacing bearings and seals is not, by itself, proof that lost capacity or poor specific power will be restored. The expected improvement must match the diagnosed mechanism.

The retained motor, starter or variable-speed drive, cooler, separator system, sensors, controller and enclosure remain part of the option. Identify the liabilities not addressed by the overhaul. An airend can be restored while an obsolete controller still creates a serious supportability exposure. Record the complete retained-package condition rather than presenting the rebuilt component as a new compressor.

7. Define the installed replacement, not the catalogue machine

For replacement, include the selected package and any necessary dryer, filters, drainage, cooling, ventilation, foundation, lifting, pipework, electrical supply, protection, controls integration and instrumentation. Check access and installation space before treating a quoted machine as installable. Separate physical delivery from commissioned availability. [4]

Include removal, transport, preservation or disposal of the old equipment. Record any trade-in or sale credit once, net of the costs needed to realise it. When a vendor quote already deducts trade-in, do not deduct it again in the worksheet. Quote exclusions should become explicit owner costs or unresolved items—not disappear from the comparison.

A new compressor does not automatically solve inadequate storage, a restricted header or incompatible sequencing. Compare targeted system improvements where those conditions drive the problem. The DOE's compressed-air guidance supports a system approach addressing demand, pressure, storage, controls, leaks and equipment rather than equipment substitution alone. [6]

8. Use one study period and a consistent cost basis

Select a common study period reflecting the plant's actual planning need. The useful extension expected from overhaul and the service life of a replacement are inputs to that common-period model, not separate denominators chosen to favour an option. Include a planned later intervention when a pathway cannot serve the full period without one.

State the currency, base date, cash-flow timing and discount-rate basis. Constant-price cash flows require a real discount rate; cash flows including general inflation require a consistent nominal rate. Do not mix these approaches. Discount future renewal costs and terminal residual values at their actual modelled dates. These are general lifecycle-cost principles; the US federal policy settings in the NIST reference are not prescribed inputs for an Indian business. [8]

Past purchase expenditure and completed repairs are normally sunk costs for this forward comparison. Preserve them as history, but do not add them to only one alternative to justify further spending. Future cancellation charges, disposal costs and realisable sale proceeds are different: they can change the decision and belong in the relevant future cash flows.

9. Calculate energy from comparable operating states

Use measured package electrical input or credible candidate performance data at the required conditions. Motor nameplate kW is not measured package consumption. A comparison should identify pressure, delivered flow, speed or load state, auxiliaries and the source of the data. CAGI performance data sheets provide a structured basis for comparing supported compressor performance; a data-sheet point still needs to be matched to the plant's duty. [7]

Annual energy, E = sum over states j of (Pj × hj)

Annual energy cost = sum over states j of (Pj × hj × cj)

Here Pj is package input power in kW, hj is annual hours in that state and cj is the applicable energy charge in INR/kWh. Include loaded, unloaded, standby and other material states without overlapping hours. Add treatment or other auxiliary energy only when it is outside the already measured boundary. Do not divide measured electrical input by motor efficiency a second time.

Use time-of-use charges where they matter. Demand charges require a separate tariff-based calculation; they cannot be inferred by multiplying saved kWh by an energy rate. A model based on one weighted average power is acceptable for screening when the averaging period and representativeness are recorded. Do not apply full-load specific power indiscriminately to every part-load hour.

10. Keep maintenance, outage cost and continuity distinct

Separate routine maintenance, planned renewal, corrective repair, temporary supply and production-service loss. Explain what changes and why. [5]

For a defined event class, a screening estimate is expected events per year multiplied by the incremental cost per event. That frequency is not a percentage probability and may exceed one. For a single mutually exclusive scenario, use its probability and consequence instead. Do not add overlapping scenarios as though they were independent. Unknown frequency remains unknown, not zero.

Use lost air-service hours after standby or temporary supply, not simply compressor repair hours, to estimate production loss. Agree the loss valuation with the plant; sales revenue per hour is not automatically avoidable loss. Avoid counting the same labour, parts, scrap or interruption twice. Keep serious safety, quality and essential-service concerns visible outside monetary totals even when a defensible financial value is unavailable.

11. Lifecycle-cost equation and sign convention

The worksheet uses the following forward-cost model for each eligible option i. Costs are positive; credits reduce cost. [2, 8]

LCCi = C0i + sum from t = 1 to N of [(Ei,t + Mi,t + Ri,t + Bi,t) / (1 + r)^t] − Vi,N / (1 + r)^N

C0 is net initial installed/intervention cost. E is energy expenditure. M is routine and corrective maintenance not counted elsewhere. R is planned renewal or later capital intervention. B is other explicitly modelled continuity or business-interruption cost. V is the net residual value at the end of the study period, after relevant disposal costs. N is the common period in years and r is the annual discount rate entered as a decimal.

Annual items occur at year-end in this version. Record any year-zero outage cost within C0. Use dated cash flows for material mid-year payments rather than forcing them into a misleading annual bucket. A shared cost may be omitted from the difference calculation only when its amount and timing are genuinely identical; document that exclusion.

Residual value is not automatically book value, original purchase price or an arbitrary percentage. State whether it represents resale, exchange recovery or supported remaining service value. Test its influence when it is uncertain. A negative net terminal value represents a disposal liability and therefore increases lifecycle cost under the equation.

12. Interpret the result without false precision

Among technically eligible, comparable options, lower modelled lifecycle cost is the financial preference under the stated assumptions. Define the net advantage of replacement as LCCoverhaul minus LCCreplacement: positive favours replacement; negative favours overhaul. Display the absolute difference, the principal drivers and any input ranges capable of reversing it.

Simple payback is only a screening indicator. For a positive incremental initial investment and positive, uniform annual savings, divide that investment by annual savings. This ignores discounting, later renewals and terminal-value differences. If savings are zero or negative, show “no simple payback on these assumptions”; do not display a negative payback as an attractive result.

A narrow difference between uncertain options is not a precise engineering answer. The appropriate recommendation may be to obtain an inspection disposition, measure part-load power or confirm installation costs before committing. The output should state which missing fact is decision-critical, rather than hiding uncertainty behind extra decimal places.

13. Test the assumptions that can reverse the choice

Run sensitivity cases for annual hours, demand profile, energy charge, installed cost, performance recovery, future intervention timing, downtime and residual value. Keep the same service boundary in every case. One-at-a-time sensitivity identifies influential inputs; combined scenarios test plausible changes that occur together.

For example, lower production can reduce operating hours but leave calendar maintenance or licence costs unchanged. A future production increase can change both running hours and the number of machines required. Avoid combining low maintenance, high residual value and exceptional efficiency into an optimistic replacement case unless that combination is supported.

Where performance data are incomplete, publish a range or a conditional scenario rather than one recommended model. Link each adjustable assumption to its measurement, quotation or engineering basis. A user-selected “conservative” label is not evidence that the selected number is conservative.

14. Useful life, supportability and recurrence

Do not assume that an overhaul resets the age of the complete package. Distinguish the expected useful extension of the repaired assembly from the condition and support horizon of retained equipment. Record available parts, controller support, software access, specialist resources and the consequence of another failure. [2, 3]

Another rebuild can reproduce the same failure when the upstream mechanism remains unchanged. Conversely, replacing a healthy airend because the true restriction is elsewhere destroys value without addressing the problem.

A short-term overhaul followed by planned replacement can be a legitimate continuity strategy. Model both interventions, both interruption periods and the terminal condition. Do not hide the later replacement outside the chosen comparison window or assume five years of restored service when the evidence supports only a shorter bridge.

15. Worked case A — higher-use plant, replacement leads

Illustrative calculation only. These figures are invented to demonstrate the method, not Reddy quotations, measured customer results, prevailing tariffs or guaranteed equipment performance. Assume both options can supply the same required service for five years. Use constant-price INR, a purely illustrative 10% real discount rate, year-end annual costs, no escalation and no tax or financing effects. Weighted input powers are assumed to represent the same annual demand profile. Outside the stated hours, both machines are assumed de-energised; unchanged common auxiliary costs are excluded.

Energy is 372,000 versus 312,000 kWh/year. Annual energy expenditure is therefore INR 3,348,000 versus INR 2,808,000. Adding the separately stated maintenance and interruption items gives recurring annual costs of INR 3,528,000 and INR 2,883,000. No future renewal is included in this case because full-period feasibility is an explicit assumption.

Replacement leads financially on these inputs, not merely because it is new. Confirm the assumed power profile, installed scope, interruption reduction and service capability before adopting that conclusion. The simple payback is supplementary; the lifecycle result also includes timing and residual value.

16. Worked case B — lower utilisation reverses the result

Retain case A's costs, horizon and technical assumptions but reduce both options to 1,000 operating hours per year. For this sensitivity test only, maintenance and expected interruption allowances remain unchanged. A real low-use plant would need those allowances reviewed rather than copied automatically.

Annual energy expenditure becomes INR 558,000 for overhaul and INR 468,000 for replacement. Including the unchanged non-energy items gives INR 738,000 and INR 543,000 per year. Their lifecycle costs become INR 3,416,555 and INR 3,522,121. Overhaul is now lower by INR 105,566. This small margin should be viewed against quotation and residual-value uncertainty.

With the other inputs fixed, annual savings equal 90H + 105,000, where H is annual operating hours. The crossover solves: 1,000,000 = (90H + 105,000) × 3.79078677 + 250,000 × 0.62092132. The result is approximately 1,309 hours/year. This is a scenario-specific threshold, not a universal replacement rule. Different maintenance, tariff, residual or duty assumptions move it.

17. Worked case C — a two-year overhaul bridge is not five-year renewal

Return to case A's 6,000-hour duty, but assume inspection and supportability evidence justify only a two-year overhaul bridge. Compare immediate replacement with overhaul now followed by replacement at the end of year 2. The later replacement is assumed to cost a net INR 1,650,000, including its own changeover scope, in constant-price terms.

The staged path pays INR 650,000 initially, incurs INR 3,528,000 in each of years 1 and 2, pays the later replacement cost at year 2, then incurs INR 2,883,000 in each of years 3 to 5. Its replacement is three years old at the end, so this example assumes a supported terminal value of INR 500,000 rather than INR 300,000. These terminal values remain illustrative assumptions.

Do not treat this result as proof that staged intervention is always wrong. Immediate replacement may be unavailable within the required restoration window. The staged option can still be necessary for continuity, but its extra cost should be visible. If the later changeover requires additional rental or service interruption not included above, add it at the relevant date.

18. Worked case D — repair hours are not production-loss hours

Assume a defined failure event makes the asset unavailable for 96 hours. Verified standby and temporary arrangements maintain the required production service for 80 of those hours. The residual air-service interruption is therefore 16 hours, not 96. Suppose the agreed incremental production-loss valuation is INR 12,000/hour, event-specific temporary supply costs INR 80,000 and additional logistics cost INR 30,000.

Incremental event cost = 16 × 12,000 + 80,000 + 30,000 = INR 302,000

If a supported scenario assumes 0.25 such events per year, its expected annual cost is INR 75,500. A lower frequency of 0.10 gives INR 30,200; a higher frequency of 0.50 gives INR 151,000. These are scenario calculations, not measured failure forecasts. Apply a probability model instead where the input represents one event's likelihood rather than an annual frequency.

Using all 96 hours as production loss would overstate this example's loss component by INR 960,000. Conversely, assuming standby exists without verifying capacity, pressure and quality can understate interruption. Keep direct repair costs in the maintenance line unless intentionally included here, and never count the same temporary-supply invoice in both initial cost and recurring event cost.

19. Reusable overhaul-versus-replacement worksheet

Copy the following tables for an actual assessment. Replace bracketed entries with project evidence; do not leave them in a published customer example. A blank cost or unknown technical state is not zero. Record a source reference and date for each decision-driving input.

A completed worksheet should allow another engineer to reproduce the result without guessing what “repair cost,” “power” or “five-year life” meant. Attach quotations and measurements rather than copying only their most favourable values.

20. Turn the calculation into a decision record

Write the recommendation in operational terms: the required service, technically eligible options, preferred option under the base assumptions, expected benefit, remaining uncertainty and the event that would change the recommendation. State whether the output is a screening comparison, a quotation-supported evaluation or an authorised execution decision.

For example: “Replacement has the lower modelled five-year cost at the agreed 6,000-hour duty. Proceed to final selection only after confirming installed scope and representative performance. Reopen the comparison if annual hours fall below the tested crossover range or if the overhaul inspection materially changes its cost.” This is an illustrative decision statement, not a recommendation for a particular plant.

Record the approving roles, funding constraints, supplier commitments and next evidence required. Technical approval, commercial acceptance and permission to interrupt production are separate decisions. Do not interpret a worksheet signature as permission to dismantle equipment or change live controls.

21. Define acceptance before the intervention

Specify what the selected option must demonstrate after work: required pressure and delivered flow, air-quality performance, stable controls, acceptable condition indicators, package input power at comparable duty, and a complete as-left record. Use equipment- and project-specific limits, not invented universal tolerances. [3, 4]

For overhaul, verify the affected assembly and the retained package interfaces. For replacement, test integration with other compressors, treatment and the production demand profile. A machine running unloaded in the workshop does not demonstrate complete installed service. Plan temporary supply, restoration contingencies and the point beyond which rollback is no longer possible.

Compare actual outcomes against the assumptions used here, distinguishing short acceptance tests from longer-term recurrence evidence. Preserve the old identity and records when equipment is retired; update the current asset register so a removed machine is not still counted as available standby capacity.

22. Questions customers and engineers ask

Should a compressor be replaced because it is old?

Not on age alone. Compare condition, required duty, recurring failures, supportability and forward cost. Age can explain exposure or obsolescence, but it does not establish repairability or prove poor performance. [2]

Is there a fixed repair-cost percentage that decides replacement?

This tool uses no universal percentage. A quotation ratio omits installation, energy, remaining condition, future renewals and continuity. A company may use a screening threshold, but it must not replace the technical and lifecycle comparison.

Does a rebuilt airend make the entire compressor new?

No. The rebuilt scope may leave the motor, controls, cooling and other components unchanged. Record exactly what is renewed and what liabilities remain. [3]

Can an energy-efficient replacement still be the wrong choice?

Yes. Low utilisation, installation constraints, incompatible duty or inadequate support can change the decision. Case B shows an illustrative financial reversal when operating hours fall.

Can an overhaul improve air output and energy performance?

It may where the verified deterioration is within the repaired scope. Require a stated recovery basis and acceptance test; do not infer a fixed efficiency improvement from an overhaul label. [3]

How should an exchange airend be compared?

Treat it as its own option. Include exact compatibility, assembly condition, core charges and credits, delivery, installation, commissioning and actual supplier obligations. Do not treat “exchange” as synonymous with new.

What happens when remaining useful life is uncertain?

Show more than one justified horizon or renewal scenario within a common study period. Identify the evidence needed to narrow the uncertainty. Do not default every overhaul to the replacement's assumed life.

Should earlier repair spending be included again?

Usually not in forward differential cost: it is historical expenditure. Include future costs and credits that change between options, while preserving past spending as useful maintenance evidence.

Can the tool guarantee fewer failures or a payback period?

No. It can calculate outcomes from supplied assumptions. Reliability improvement needs a supported mechanism and post-intervention evidence; payback depends on actual duty, costs and implementation.

What data should be collected first?

These determine whether more detailed modelling is worthwhile.

23. Reddy Compressors assessment enquiry

Request an overhaul-versus-replacement assessment. Provide the compressor and airend nameplates, the operating problem, service and failure history, available measurements, current duty and quotations. Include the required restoration date and any restriction on interrupting production.

The assessment should identify the evidence still needed, compare technically credible pathways and define an installed scope and acceptance basis. Parts, exchange assemblies, rental support, specialist processes and supplier obligations remain subject to confirmation for the particular equipment and project. No fixed response time, savings percentage or restoration-to-new claim is implied.

Choose the intervention that restores the required service—not simply the quotation with the smaller number.