An illustrative machine example

Last month's oil report raised a concern about particles. This month's looks better. The tempting conclusion is that the problem has gone away.

Then a technician mentions one small change: the usual sampling point was inaccessible, so the bottle was filled after the filter instead of before it. Same gearbox. Same oil. Different question.

The laboratory can measure what arrived in the bottle. It cannot reconstruct an unrecorded change at the machine. Before using the result to close an investigation, trace how that bottle was filled.

Start with the question, then choose the point

Imagine an industrial gearbox with a circulating oil supply. Oil returns from the gearbox through a filter to a reservoir. This is a teaching example, not a drawing of a particular installation. The pump, supply line and safety provisions are omitted from the return-path diagram.

Point A is before the return filter. Point B is after it. Point C is at the reservoir bottom. A and B sample the return stream at different stages. C can include material accumulated away from that stream. Calling all three bottles 'gearbox oil' loses the distinction that matters.

Think of counting traffic before and after an exit. Both counts can be correct without describing the same traffic. A filter similarly changes the particle population passing onward. That makes location part of the result, not an administrative detail.

Illustrative return pathSame oil system. Different evidence.
  1. GearboxReturning oil
  2. ABefore filter
  3. Return filterParticle removal
  4. BAfter filter
  5. ReservoirC Bottom location
Conceptual return path only, not a piping or sampling instruction. Pump, supply and safety provisions are omitted. A, B and C are example locations, not interchangeable samples.
ABefore the filterWhat is returning from the gearbox?
Useful for investigating the return stream before this filter removes particles. It does not identify the cause of wear by itself.
BAfter the filterWhat is passing onward?
Useful for examining downstream cleanliness. A lower count here does not prove the gearbox stopped generating debris.
CReservoir bottomWhat has accumulated here?
May support a targeted deposit or free-water investigation. Do not substitute it for the established circulating-oil trend.

Parker describes return-line filtration as capturing debris before it enters the reservoir. The three-position comparison below applies that principle to an illustrative circuit. Parker: Handbook of Hydraulic Filtration.

Same point is only half the comparison

A repeatable location is useful only if the machine state is also understood. A sample following normal circulation and one taken after a long shutdown may contain different distributions of suspended material. Record whether the machine was operating or stopped, its relevant duty, and the delay after shutdown when applicable.

Mobil's sampling guidance stresses consistent location and interval, with oil at operating temperature. That is a sampling objective, not permission to approach moving equipment or open a pressurized connection. The approved equipment and site procedure must define how to obtain a representative sample safely.

If that procedure cannot be followed, report the limitation. Do not quietly choose an easier point and file the bottle under the usual one. Ask the technical owner whether an alternative can answer the intended question, and identify it separately.

Mobil's fundamentals guide explains sample consistency. HSE's maintenance guidance supports planning the work around equipment instructions, competence and safe access. Mobil: Lubricant Analysis Fundamentals; HSE safety guidance.

Do not let the bottle become part of the problem

A correctly chosen point can still produce a misleading sample if dust enters through the cap or residue remains in the sampling equipment. Use the laboratory's specified bottle and collection equipment, protect them before use, and follow the approved port-cleaning and flushing method.

Caterpillar's guide calls for new tubing for each sample and protection of bottles and equipment from contamination. Do not improvise a universal flush quantity, rinse a prepared bottle without instruction, or substitute a workshop container. The method depends on the port, equipment and analysis required.

If contamination is suspected during collection, flag it and arrange a replacement under the approved procedure. Keep that fact with the record. An unexplained 'bad sample' label is not enough to dismiss a concerning machine result.

Caterpillar's sampling guide distinguishes valve and vacuum methods and explains why clean supplies matter. Follow the instructions applicable to your own equipment. Caterpillar: How to Take a Good S.O.S Sample.

Give the analyst the missing half of the sample

The machine name alone will not explain whether an apparent improvement followed a filter change, a top-up or a different sampling point. The bottle and its submission record need to tell the same story.

Record the asset and compartment, exact point, collection time, operating state, oil identity, oil and machine service hours, and recent additions or maintenance. Use 'unknown' when a detail cannot be established. Caterpillar explicitly asks for equipment and fluid service information, changes and top-ups to support interpretation.

A short note such as 'point B used because A was inaccessible' is more useful than a perfectly completed form that names the wrong point. The example below shows how to carry that exception into the review.

Illustrative sample record

A different point needs a visible note

Asset / compartment
GB-07 / circulating oil system
Sample position
B, downstream of return filter
Reason for change
Routine point A inaccessible
Machine state
Record actual duty and collection conditions
Service context
Record oil hours, top-ups and filter work
Review note
Different point: review before comparing with A
GB-07 is fictional. This extract highlights an exception; it is not a complete laboratory submission form or a measured field result.

Your turn

Would you close the investigation?

The previous sample came from A under the established procedure. The new bottle came from B after a filter change and shows fewer particles. A colleague wants to mark the gearbox problem resolved. What would you say?

  1. What changed besides the result?
  2. What can the new bottle actually support?
  3. What evidence would make the next comparison useful?
Compare your answer

Not yet. The point and maintenance history both changed. The result describes oil after the filter under the recorded conditions; it is not a like-for-like improvement in the A trend.

Keep the original concern open for technical review. Confirm the records, then arrange any approved follow-up at the established point and comparable state. Review relevant inspections and operating evidence too. A concerning result still needs a response under the site's rules; uncertainty is not a reason to wait through an unsafe condition.

Before acting on the next report, ask: 'Does this bottle represent the condition we are deciding about?' If the answer is unclear, improve the sampling record before building a confident maintenance story around the number.

Sources and scope

  1. Parker: Handbook of Hydraulic Filtration

    Return-line filtration, printed page 29; fluid sampling, printed pages 32 and 33.

  2. Mobil: Lubricant Analysis Fundamentals

    Taking a representative sample, page 6.

  3. Caterpillar: How to Take a Good S.O.S Sample

    Oil sampling, clean supplies and complete sample labels. Equipment-specific instructions are not universal.

  4. HSE: Maintenance of work equipment

    Safe access, competent personnel, equipment instructions and isolation.

Build your oil-analysis skills

Machinery Lubrication I

Build a foundation in sampling and contamination control.

View course
Machine Lubricant Analyst II

Develop your oil-analysis and interpretation skills.

View course

New field notes

Get one practical lubrication field note each month.

Clear guidance on lubrication, oil analysis and reliability improvement. No daily emails.

Back to insightsExplore coursesView training dates