Verified Compressor Overhaul Projects

Verified Compressor Overhaul Projects

OWNED DECISION: Connect as-found condition, approved overhaul scope, inspection measurements and as-left performance. MANDATORY ASSET: Overhaul Project Evidence Template.

1. Direct engineering answer

A verified compressor overhaul project is a chain from the asset’s pre-work condition and documented failure symptoms through approved disassembly, measurements, disposition of components, controlled reassembly, commissioning and measured as-left acceptance. Evidence must identify exactly what was overhauled, by whom, using which approved procedures and what was actually tested. Completing an overhaul is not proof that rated capacity, original efficiency, guaranteed service life or an OEM warranty has been restored.

3. Scope taxonomy: what does “overhaul” mean?

Distinguish compressor-package overhaul, airend rebuild, reciprocating compressor cylinder/valve/piston overhaul, motor overhaul, engine overhaul, dryer overhaul and component exchange. A refurbished airend installed into an otherwise unchanged package is not equivalent to a complete compressor overhaul. Record inclusions, exclusions, outsourced operations and replaced-versus-reconditioned components. The commercial work order controls the scope, not the marketing headline.

4. Asset identity and operating history

Capture site-approved alias, manufacturer, model, serial, technology, rated pressure and declared capacity basis, package age, total operating hours, loaded hours if available, prior overhaul dates, service history, lubricants, alarm history and recorded duty. Preserve serial-specific OEM instructions and relevant revisions. Historical readings without instrument identities or dates are condition indicators, not verified baseline performance.

5. As-found condition and chain of custody

Before dismantling, record reason for removal, complaints, fault codes, operating pressures, temperatures, motor current or measured kW, discharge condition, noise/vibration trends, oil consumption, visible leaks and contamination. Photograph and tag external connections, safety-critical parts, coupling orientation, wiring and transport condition. Log receipt, preservation, quarantine, packaging damage and any missing accessories. Avoid claiming a failure mode before inspection establishes it.

6. Safe isolation and disassembly evidence

Overhaul records must demonstrate a qualified team followed approved isolation, depressurisation, electrical lockout, cooling and lifting procedures. Record permits, pressure-zero verification, trapped-energy checks, lifting method and critical component identification. Public photographs should omit unsafe practices and confidential plant information.

7. Inspection plan and acceptance authority

Predefine the inspection route and decision owner: visual inspection, dimensional checks, non-destructive testing when required, oil analysis, bearing condition, seals, rotor/cylinder surfaces, housings, gears, coolers, valves and controls. State the serial-specific OEM or approved engineering limit for each measurement; a reading without its acceptance criterion cannot support “within tolerance”. Include measurement location, temperature, instrument calibration, repeatability and who signed the disposition.

8. Airend condition: rotary screw applications

Where relevant, document rotor condition, housing contact, bearing and seal condition, axial/radial clearances, gear condition, inlet/unloader components, separator system and lubricant evidence. A photographed rotor may show visible scoring but not prove acceptable running clearance or internal efficiency. Record measurement method and manufacturer-approved tolerances, and distinguish shop assembly checks from a complete package performance test.

9. Reciprocating compressor condition

Where relevant, record cylinder bore, piston/ring condition, valve seats, rod packing, crankshaft and connecting rod condition, lubrication passages, crosshead alignment and cooler condition. Valve replacement can improve delivery but a bench check alone does not prove package capacity at installed conditions. Include pulsation, temperature and pressure evidence only where tested and approved.

10. Root cause versus observed damage

Separate symptom, damaged component, initiating mechanism and contributing operating conditions. Bearing distress may correlate with lubrication contamination, alignment, overload or heat; one damaged bearing does not independently prove which mechanism caused failure. Record competing hypotheses, test results and unresolved factors. If a root cause remains undetermined, say so rather than assigning blame to operator practice or a supplier.

11. Repair-or-replace decision and commercial limits

Record alternatives considered: targeted repair, full rebuild, exchange airend, replacement package or deferral. Compare authorised scope, parts availability, expected downtime, testability, warranty terms and lifecycle economics. Do not claim “like new” merely because major parts were replaced. State whether components were genuine OEM, approved equivalent, customer supplied or refurbished, based on traceable documents and consent.

12. Parts traceability and configuration control

For each critical replacement, capture part number, supplier, batch/serial where applicable, certificate or purchase reference, material/specification when relevant, installation date and technician. Document substitutions and engineering approvals. A supplier invoice proves procurement, not automatically compatibility, authenticity, statutory conformity or installed condition. Maintain an as-left bill of materials and deviations from the initial quotation.

13. Dimensional measurement integrity

Record measurement instrument type, serial, calibration validity, environmental conditions, datum, units, resolution, repeat measurements and technician. Distinguish actual clearance from nominal drawing clearance and installed fit from bench fit. Temperature expansion and assembly state may matter; no universal acceptable bearing preload, rotor clearance or shaft runout is assumed. Any unverified OEM limit is marked “hold for technical review”.

14. Controlled assembly and quality gates

Use approved assembly instructions, cleanliness control, lubrication, seal installation, alignment, fastener tightening and rotation checks. Capture hold points, inspection signatures, nonconformance reports and rework. An unsigned workshop checklist is an internal note, not a witnessed quality certificate. Identify subcontracted machining or balancing and retain their source records without implying the main contractor performed the work.

15. Balancing, alignment and vibration

If rotating parts were balanced, record rotor identity, balancing report, test conditions and applicable acceptance criterion. For package coupling or belt alignment, record method and as-left readings. A single overall vibration number cannot isolate all faults; measurement location, speed, load, sensor mounting and frequency spectrum may be required. Do not invent a universal vibration limit or claim standards compliance without scoped evidence.

16. Workshop test versus site test

Define the actual tested assembly and boundary. An airend spin test may confirm free rotation or leakage under specific conditions but cannot alone demonstrate rated package flow or electrical specific power. A workshop test may exclude site cooling, piping and demand. Site commissioning tests must state inlet conditions, discharge pressure, load, temperatures, alarms, oil carryover checks where applicable and acceptance signoff.

17. As-left performance metrics

Where measured, record delivered flow at a defined reference basis, discharge pressure, true electrical input, specific power, oil temperature, vibration, discharge temperature, leak condition, noise and operating-state stability. Keep original rated values, pre-work field readings and post-work readings in separate columns. Compare only like-for-like pressure, ambient, control state, instrument and demand conditions; otherwise use a documented adjustment method or withhold quantitative improvement claims.

18. Specific power and the performance boundary

Specific power may be reported as kW per (m³/min) at a stated pressure and flow reference. If power is measured at the compressor motor and flow at a package outlet, identify excluded auxiliaries. If one value is nameplate and another is measured, they are not a verified before/after pair. A lower current reading at reduced flow does not establish higher efficiency. State measurement uncertainty and the load point.

19. Oil, contamination and downstream protection

Record flushing, oil and separator changes, filter status, contamination evidence, condensate handling and post-start monitoring. Where failure debris may have entered the distribution or treatment train, the project scope must state whether downstream inspection was included. An overhaul does not automatically establish ISO 8573 air purity or eliminate contamination at every use point. Use measured sampling only when the claim requires it.

20. Acceptance at return to service

Separate workshop release, installation/reconnection, first start, stable operating trial, customer handover and later performance verification. Record open punch items and temporary restrictions. If the customer could not permit a full-load test, mark that acceptance item deferred. A running machine is not automatically a passed rated-duty performance test.

21. Persistence, recurrence and follow-up

Define a dated recheck appropriate to the agreed scope: vibration trend, oil analysis, temperature, leakage, operating hours, alarm history, production acceptance and any repeat flow/power tests. One successful restart cannot substantiate “permanent fix”, “lifetime extension” or “zero future failures”. Track recurrence and corrective action, including unfavourable evidence.

22. Causal attribution and confounders

An overhaul may coincide with cleaning coolers, replacing filters, changing lubricant, resetting controls, reducing pressure or repairing leaks. These interventions may independently change energy and flow. Preserve a dated change ledger; avoid attributing every difference to airend refurbishment. Where no pre-work performance test was possible due to failure, explicitly say a measured before/after efficiency comparison is unavailable.

23. Measurement uncertainty and data integrity

For each instrument record calibration, range, resolution, uncertainty, sampling interval, logger synchronisation and missing data. Distinguish measurement uncertainty from process variation and model uncertainty. Do not report a 1% capacity recovery when the combined comparison uncertainty exceeds the claimed difference. Keep raw files, derived calculations, formula version, reviewer and chart annotations.

24. Warranty, OEM and certification boundaries

Document actual supplier warranty scope, exclusions, dates and claimant rights. Rebuilding a compressor does not renew the original OEM warranty unless expressly documented. A third-party overhaul is not automatically OEM-authorised, certified, factory-standard or equivalent to a new machine. Refer to relevant OEM manuals, applicable pressure/electrical safety obligations and measurement standards only when the project’s edition and scope are established.

25. Evidence permissions and commercial disclosure

Record the customer/data owner, publication consent for site name, logos, images, measured outcomes, service reports and technical drawings, plus whether Reddy supplied parts, performed work, supervised or merely documented it. Anonymisation does not automatically remove contractual confidentiality obligations. Publish only approved claims and provide a correction/withdrawal path.

26. Mandatory asset — Overhaul Project Evidence Template

Controlled blank record: each field requires source ID, input/evidence, method/workflow, measured result or status, accountable owner, open item and disposition. Values are not pre-populated as real customer results. Status options: NOT RECEIVED, RECEIVED, VERIFIED, EXCEPTION, NOT TESTED, NOT APPLICABLE, REJECTED, WITHDRAWN.

27. Evidence status and claim rules

28. Worked case A — Clearance measurement

Illustrative only: an inspection sheet records a rotor axial clearance of 0.19 mm but has no OEM limit or temperature. The number documents a measurement, not pass/fail. The technical reviewer must retrieve the applicable serial-specific tolerance and validate the measurement method before approval.

29. Worked case B — Failed machine has no valid energy baseline

Illustrative only: an airend seized before any flow/power logging. After overhaul, the package runs at 7 bar(g) with a recorded 8.5 m³/min. This is an as-left observation; no pre-overhaul specific-power improvement can be quantified from the failed condition. Compare with OEM rating only after reconciling conditions and test method.

30. Worked case C — Specific-power comparison

Illustrative only: baseline true power 92 kW at 11.0 m³/min and after true power 90 kW at 11.6 m³/min, both at comparable pressure and reference conditions. Specific power changes from 8.36 to 7.76 kW/(m³/min), an illustrative reduction of about 7.2%. This is not a verified real-world saving: instrument uncertainty, demand, auxiliaries and matched conditions must be reviewed.

31. Worked case D — Pressure confounder

Illustrative only: pre-work power is 95 kW at 8.0 bar(g); post-work power is 88 kW at 7.0 bar(g). It is invalid to attribute the entire 7 kW difference to overhaul. The pressure setpoint changed; obtain matched-pressure measurements or a defensible adjustment model.

32. Worked case E — Vibration measurement

Illustrative only: a sensor reads 5.1 mm/s before repair and 2.8 mm/s after repair at different speeds and loads. The lower number may be encouraging but is not directly comparable without sensor location, frequency response, machine operating condition and acceptance criterion. Record both readings and uncertainty.

33. Worked case F — Replacement part traceability

Illustrative only: a bearing invoice lists a supplier part number but no installed serial or work-order link. The record can substantiate purchase, not installation or OEM approval. Require a technician-signed installation trace and authorised compatibility disposition.

34. Worked case G — Oil contamination recurrence

Illustrative only: metallic particles are found before overhaul; the package passes first start, but oil analysis after 60 operating hours still shows abnormal debris. The record must preserve the adverse follow-up and investigation. “Permanent contamination elimination” is unsupported.

35. Worked case H — Workshop test versus rated duty

Illustrative only: an airend spins freely in the workshop and passes a low-pressure leakage check. The compressor has not yet undergone full package testing. Publish “workshop checks completed within documented scope,” not “restored rated compressor output”.

36. Worked case I — Uncertainty budget

Illustrative only: independent standard uncertainty for two flow measurements is 0.18 m³/min each. The standard uncertainty of their difference is sqrt(0.18²+0.18²)=0.255 m³/min before demand and pressure variability. An observed 0.12 m³/min difference is not robust evidence of a capacity change.

37. Worked case J — Warranty claim

Illustrative only: a third-party workshop provides a six-month workmanship warranty. This does not imply the OEM original warranty has been renewed or that all package components are covered. The published wording must match the signed warranty terms.

38. Worked case K — Partial overhaul

Illustrative only: the work order replaces inlet valve, seals and separator but does not open the airend. The project may be a valid targeted refurbishment, but calling it a “complete airend overhaul” misstates scope. Show the actual component boundary.

39. Worked case L — Follow-up persistence

Illustrative only: commissioning records normal discharge temperature; after four weeks a cooler fouls and high-temperature trips return. The initial as-left result remains dated evidence, but any persistent-reliability claim requires new verification and transparent correction.

40. Before/after evidence matrix

41. Claim and evidence ledger

42. Publication architecture and media

Recommended public sequence: answer-ready overview; anonymised approved asset and symptom; as-found evidence; approved scope; inspection/disposition matrix; as-left assembly and test boundary; quantified outcome with conditions; exceptions; persistence; customer consent and source notes. Caption each image with stage and date, remove sensitive serials where required, and show chart axes, pressure/flow units and uncertainty. A slider of “before and after” photos without measurements is visual documentation, not proof of performance recovery.