Oil Analysis: When to Monitor, Investigate or Act

A gearbox report comes back with iron flagged at 48 mg/kg. Change the oil, take another sample or raise a work order? The flag starts the review. It does not choose the response.

Start with three questions: what changed, how quickly, and does the machine show the same problem? Check the history and whether the sample represents the oil in service. Then choose a response that accounts for the consequence of waiting.

Read the history behind the flag

Put the earlier reports beside the latest one. Do they refer to the same asset, compartment, lubricant and sample point?

Next, check the operating hours between samples and the hours on the oil. Note oil additions, changes, filter work and recent repairs. Ask operations whether the load or duty changed. These are part of the result, not background details to fill in later.

These details can change the meaning of the same number. The example below holds the sampling method and operating intervals constant so the difference in history is visible.

Two histories, one latest iron resultIllustrative data, not plant results or alarm limits. Iron in mg/kg.

Iron concentration at four consecutive samples for two illustrative gearboxes
History	April	May	June	July
Stable	44	46	45	48
Rising	12	18	30	48

Assume equal operating intervals, the same oil charge and test method, and no recorded oil additions or filtration changes within each history.

The rising history warrants prompt investigation. The stable history may support monitoring only if that concentration is an established acceptable baseline for that asset and the other evidence agrees. Neither sequence, by itself, establishes a damaged component or an oil change requirement.

STLE’s discussion of trending explains why sample timing and sump volume matter. A wear metal concentration reflects debris entering the oil, debris being removed and the oil volume carrying it. Adding clean oil can lower the concentration without repairing a wearing surface. Record that addition before interpreting an apparent improvement.

Give the report a response and a deadline

Use the laboratory’s recommendation alongside the machine’s approved limits and operating evidence. The categories below are a practical review method, not a universal severity scale. STLE describes how limits and interpretation depend on the equipment, lubricant and duty.

Monitor

The sample is credible, the trend fits an acceptable baseline, and no related result or operating observation calls for intervention.

Record the next review: Keep the approved sampling schedule unless the review justifies a change. State what new result or symptom would trigger escalation.

Investigate

A meaningful change has appeared, the sample is questionable, or the results disagree. There is a specific uncertainty to resolve.

Assign the next check: Name the question, the evidence needed, its owner and a due time based on machine risk. Agree with operations how the asset will be managed while evidence is gathered.

Act

An approved response criterion has been reached, or the combined evidence establishes a condition that requires corrective work.

Authorise the work: Control the cause and the exposure. The response may be contamination removal, a repair, an oil change or an outage. Its urgency comes from the asset risk and procedure.

“Monitor” still requires a next review. “Investigate” is incomplete without a question. “Act” needs a work order and an agreed way to verify the result.

Select evidence that can change the decision

A repeat sample checks whether a finding can be reproduced. Another test may answer a different question. Choose deliberately: advanced analysis can include wear debris examination, water separation, oxidation reserve and other properties that a routine package may not cover.

Iron is rising

Routine ICP elemental analysis has a particle size limitation. The public scope of ASTM D5185 says larger particles can give low results. A modest iron value therefore cannot exclude damaging larger debris.

Ask the analyst whether ferrous debris measurement, microscopy or filter debris examination would help. Compare the result with vibration, the components present and recent work. Iron alone cannot identify which gear or bearing is wearing, or establish the wear mechanism.

Water is reported

Check the method and units before comparing results. A percentage of water saturation and a concentration in mg/kg describe different quantities. Karl Fischer titration measures water content, but the procedure and possible interferences still matter. Ask the laboratory whether they affect this oil.

Investigate plausible entry paths: a cooler leak, damaged seals, the breather, washdown or contaminated transfer equipment. Confirmed water requires a response suited to the fluid and asset. Removing water while leaving the entry path open invites recurrence.

The particle count has increased

Compare the sample point and conditions with the earlier sample. Recent filter work, a different sampling location or a dirty bottle can change the result. Pall explains that water, air and dark fluid can affect the suitability of optical particle counting.

Ask whether the method and sample preparation suit the fluid before treating every counted event as a solid wear particle. If contamination is confirmed, inspect ingress routes, filter condition and bypass arrangements. Choose filtration against the asset target and lubricant requirements.

Viscosity has moved

Compare results at the same test temperature and with the correct unused lubricant reference. Check product identity and any recent addition first. A viscosity change can have several explanations, including mixing, oxidation or contamination; the direction of change alone does not identify the cause.

Review related results with the laboratory and lubricant supplier. For an engine, for example, suspected fuel dilution needs evidence relevant to fuel as well as viscosity. For an industrial oil, the useful confirmation tests depend on the fluid and suspected contaminant.

Make a repeat sample answer a question

Ask the laboratory whether it can rerun the retained sample. That can check the analytical result, but it cannot establish that the original bottle represented the machine. A fresh sample from the controlled point addresses a different uncertainty.

For the fresh sample, use the approved bottle, point, flushing method and machine state. Record the time and any work done since the first sample. Sampling after an oil change or filtration gives useful information about the intervention, but it is no longer a direct confirmation of the original condition.

Tell the laboratory what you are trying to distinguish. “Please confirm the water result before we investigate a possible cooler leak” is more useful than “Please retest.” Where the operating risk already requires action, arrange confirmation alongside that action.

Write work that someone can close

“Oil report abnormal, check gearbox” leaves the next person to reconstruct the review. Attach the report and comparison history, identify the uncertain finding, and say what evidence is required.

Example investigation brief

Illustrative gearbox case. Adapt the work and timing to the actual asset.

Finding: Iron rose across three comparable samples. No oil addition or filter change is recorded.
Question: Is the trend consistent with active component wear, and is larger debris present?
Assigned work: The reliability engineer will review the trend with the analyst, select the needed debris examination and compare it with vibration and maintenance history.
Operating decision: Operations and maintenance will record the permitted machine state, escalation criteria and review deadline.
Closure evidence: Record what the investigation established, the corrective work authorised, and the next sample or inspection needed to verify the outcome.

Replace the role with a person’s name and put the due time into the work system. If the response includes an oil change, retain the reason, product and quantity. A lower iron result immediately afterwards may reflect dilution; it does not, by itself, prove that wear has stopped.

Keep the action open until the agreed evidence has been reviewed. Record an inconclusive result as inconclusive, with the next decision date. That record is what lets the next reviewer tell whether the machine improved or only the report changed.

Review a difficult oil report with VAS

Bring the report, available sample history and machine operating record. We can help define the next check and a practical maintenance response.

Discuss a report

Sources and scope

This guide applies published technical principles to maintenance review. The examples and response categories are editorial guidance. Use current OEM instructions, laboratory advice and site procedures to set limits, deadlines and work methods. ISO 17359 provides the broader framework for a condition monitoring programme.

STLE: Limitless possibilitiesWhy oil age, sump volume and sampling consistency affect a trend.
STLE: Oil analysis flagging limitsHow operating context, OEM criteria and related results affect interpretation.
ASTM D5185: Elemental analysisPublic scope of the 2018 edition, including the particle size limitation of routine ICP analysis.
ASTM D6304: Water by Karl Fischer titrationPublic scope of the 2025 edition and its stated analytical interferences.
Pall: Fluid monitoring methodsWhy water, air and fluid appearance matter when selecting a particle measurement method.
ALS: Advanced oil analysisExamples of additional tests available beyond a routine package.
ISO 17359: Condition monitoring guidelinesGeneral programme guidance. It does not prescribe the response categories used in this article.

Laboratory photograph: U.S. Air Force photo by Michelle Gigante, DVIDS image 4127901, public domain. The source identifies the fluid as F-16 hydraulic fluid. It is not a client sample or this article's worked example. The appearance of U.S. Department of War visual information does not imply or constitute endorsement.