Industry reliability guide
Data Centres & Critical Power
Standby engines, stored fuel, cooling systems, transformers and mechanical utilities must remain ready despite low run hours and high failure consequence.
Data Centres & Critical Power programmes should connect equipment criticality, credible failure modes, representative samples, routine and triggered tests, response ownership and maintenance feedback.
Operating context
Design the programme around how the assets fail.
Data-centre fluid programmes are readiness programmes. Long idle periods can promote fuel water bottoms, microbial growth, condensation and corrosion, while short monthly runs may not fully stabilise engines or evaporate moisture. Link oil, coolant, fuel and transformer evidence to generator load tests, fuel turnover, cooling performance and critical-power maintenance.
Use this guide to scope an oil-analysis programme; then confirm equipment-specific limits, methods and safety controls with the laboratory, OEM and site engineering team.
Equipment-to-test map
What to test—and why.
Routine panels create comparable history. Triggered tests answer a focused question after a trend change, inspection finding or operating event.
| Equipment | Fluids | Primary risks | Routine tests | Triggered / advanced tests | Sampling approach |
|---|---|---|---|---|---|
| Standby diesel generators | Engine oil, coolant and stored diesel | Condensation, fuel degradation, coolant leakage, corrosion and start-readiness risk | kinematic viscosity, elemental analysis icp oes, base number, fuel dilution, coolant ph | soot, glycol contamination, pq index, analytical ferrography | Sample engine oil after a representative loaded run from a dedicated live gallery; record idle time and load-test history. |
| Bulk and day-tank fuel systems | Diesel fuel | Water bottoms, microbial debris, oxidation, particulate contamination and wrong-product delivery | karl fischer water, density, flash point, cold filter plugging point | distillation, sulfur, particle count | Use defined upper, middle, lower and bottom locations for storage investigations; keep routine acceptance sampling separate. |
| Transformers, chillers and cooling utilities | Transformer oil, compressor oil, coolant and hydraulic fluid | Moisture, dielectric deterioration, refrigerant/process contamination and compressor wear | dissolved gas analysis, transformer oil breakdown voltage, karl fischer water, acid number | elemental analysis icp oes, ftir oil condition monitoring, pq index | Use approved gas-tight transformer procedures and dedicated live compressor or coolant points under electrical and pressure safety controls. |
Programme priorities
Controls that make the data useful.
- 01Treat low-hours equipment as a distinct operating regime
- 02Integrate fuel-tank condition with generator readiness tests
- 03Trend each generator, tank and transformer separately
- 04Escalate rapid DGA, moisture or stored-fuel changes through qualified critical-power procedures
Maintenance outcomes
Decisions the programme should improve.
- 01Improved emergency-start confidence
- 02Earlier control of fuel and coolant deterioration
- 03Condition-based tank treatment and filtration
- 04Better evidence for critical-power maintenance windows
Programme architecture
Build a closed evidence-to-action loop.
A test panel becomes a reliability programme only when the sample, interpretation, response and field outcome remain connected to the same asset and compartment.
| Stage | Control | Evidence created | Decision |
|---|---|---|---|
| 1 · Baseline | Confirm asset, compartment, fluid, method, point and operating state | Comparable new or stable in-service reference | Define normal variability and initial interval |
| 2 · Routine surveillance | Collect at a fixed risk-based interval from the documented point | Trend of fluid condition, contamination and wear | Continue, monitor or request focused confirmation |
| 3 · Exception diagnosis | Use mechanism-specific tests, secondary points, filters and companion technologies | Independent evidence for or against the failure hypothesis | Protect, inspect, correct or revise the hypothesis |
| 4 · Verification | Repeat comparable evidence after maintenance | Post-action condition and rate of change | Close the work order or escalate |
| 5 · Programme learning | Record inspection findings, false positives and missed detections | Asset-specific failure signatures and improved rules | Change point, panel, interval, limit or procedure |
Interpretation ladder
Translate severity into a defined response.
These are workflow classes—not universal pass/fail limits. Asset consequence, rate of change and site safety procedures control the urgency.
| Class | Typical evidence pattern | Minimum review | Possible response |
|---|---|---|---|
| Stable / Normal | Comparable results remain within the justified baseline and no corroborating distress is present | Confirm route and context are complete | Continue the planned interval |
| Monitor | A small deviation, unusual rate of change or incomplete contextual explanation | Check sample, top-up, maintenance and companion indicators | Clarify data, add one focused test or shorten the interval |
| Abnormal / Investigate | Unsatisfactory fluid, contamination or wear evidence with a credible developing mechanism | Confirm promptly using independent evidence | Inspect, control the source and plan corrective work |
| Critical / Urgent review | Rapid, severe or corroborated change with material equipment or safety consequence | Immediate qualified operational and engineering review | Apply site procedures; protect people and equipment before diagnosis continues |
Connected test references
Complete test library →Open the laboratory method context.
TEST REFERENCEkinematic viscosityPrinciple, sampling, interpretation and limitations→TEST REFERENCEelemental analysis icp oesPrinciple, sampling, interpretation and limitations→TEST REFERENCEbase numberPrinciple, sampling, interpretation and limitations→TEST REFERENCEfuel dilutionPrinciple, sampling, interpretation and limitations→TEST REFERENCEcoolant phPrinciple, sampling, interpretation and limitations→TEST REFERENCEsootPrinciple, sampling, interpretation and limitations→TEST REFERENCEglycol contaminationPrinciple, sampling, interpretation and limitations→TEST REFERENCEpq indexPrinciple, sampling, interpretation and limitations→TEST REFERENCEanalytical ferrographyPrinciple, sampling, interpretation and limitations→TEST REFERENCEkarl fischer waterPrinciple, sampling, interpretation and limitations→TEST REFERENCEdensityPrinciple, sampling, interpretation and limitations→TEST REFERENCEflash pointPrinciple, sampling, interpretation and limitations→TEST REFERENCEcold filter plugging pointPrinciple, sampling, interpretation and limitations→TEST REFERENCEdistillationPrinciple, sampling, interpretation and limitations→TEST REFERENCEsulfurPrinciple, sampling, interpretation and limitations→TEST REFERENCEparticle countPrinciple, sampling, interpretation and limitations→TEST REFERENCEdissolved gas analysisPrinciple, sampling, interpretation and limitations→TEST REFERENCEtransformer oil breakdown voltagePrinciple, sampling, interpretation and limitations→TEST REFERENCEacid numberPrinciple, sampling, interpretation and limitations→TEST REFERENCEftir oil condition monitoringPrinciple, sampling, interpretation and limitations→