Oil analysis answers three questions from one sample: is the oil still fit? (viscosity, oxidation, additives, TAN/TBN), what's getting in? (water, dirt, ISO 4406 particle count), and what's wearing off? (wear metals by spectrometry, plus debris shape under ferrography).
The wear-metal fingerprint locates the fault: iron points to steel (gears, shafts, rolling bearings), copper + tin to bronze bushings, tin + lead to white-metal journal bearings, silicon to dirt ingress, sodium + boron to a coolant leak.
Cleanliness is destiny. Most wear is caused by contamination, so getting the ISO 4406 code down with better filtration and sealing can multiply bearing life several times over — at almost no cost.
1 · The machine's blood test
Of all the condition-monitoring techniques, oil analysis is the one that sees inside the machine. A blood test doesn't require surgery; an oil sample doesn't require a strip-down. And because every wearing surface sheds particles into the same oil, a single sample carries evidence from the whole lubricated system — often flagging trouble on the early part of the P-F curve, before vibration or heat appear.
It is the natural partner to vibration: vibration is best at structural and dynamic faults (imbalance, misalignment, late-stage bearing damage), while oil analysis is best at the chemistry — lubricant degradation, contamination, and the very first wear particles. Critical machines get both.
2 · Three questions, one sample
Is the oil still fit for service?
The lubricant itself degrades. The tests: viscosity (the single most important oil property — drift of more than ~10% is a red flag), oxidation and TAN (total acid number, rising as the oil ages), TBN (total base number, the reserve alkalinity in engine oils, falling as it's used up), and additive levels (anti-wear, anti-oxidant — they deplete).
What's getting in?
Water (by crackle test or Karl Fischer — even a fraction of a percent destroys the oil film and corrodes), dirt (silicon), and overall particle count graded as an ISO 4406 cleanliness code. Contamination is the dominant cause of wear, which is why this part of the report matters most.
What's wearing off?
Elemental spectrometry measures wear metals in parts per million (ppm); particle counting sizes the debris; and ferrography looks at the shape of the particles under a microscope to tell rubbing wear from cutting, fatigue or severe sliding. The metals present are a fingerprint of which component is wearing.
3 · The key tests at a glance
| Test | Measures | Flags |
|---|---|---|
| Viscosity | Oil thickness vs grade | Wrong/degraded oil, fuel dilution, oxidation |
| Elemental (ICP) | Wear metals, contaminants, additives (ppm) | Which component is wearing; dirt; additive depletion |
| Particle count / ISO 4406 | Solid cleanliness code | Contamination — the biggest wear driver |
| Water (Karl Fischer) | Water content | Leaks, condensation; film breakdown, corrosion |
| TAN / TBN | Acidity / alkaline reserve | Oil oxidation/ageing; acid attack |
| Ferrography | Debris size & shape | Wear mode — rubbing, cutting, fatigue, sliding |
4 · Reading the wear-metal fingerprint
This is where oil analysis turns into diagnosis. Because each machine component is made of characteristic alloys, the pattern of elevated wear metals points to the source. Pick a sample below and see which component the fingerprint accuses.
Interactive 1 — Wear-metal fingerprint
Live modelElemental spectrometry (ppm)
5 · Cleanliness is destiny
The single most actionable thing oil analysis gives you is the ISO 4406 cleanliness code — three numbers counting particles larger than 4, 6 and 14 µm. It matters because contaminant particles are what wear most machines out: they indent and abrade the very surfaces the oil film is trying to protect, dragging bearings into the boundary regime from Part 6.
The payoff is large and cheap. Cleaning up the oil — better filtration, sealed breathers, clean top-up practice — moves the operating point along the contamination factor in the ISO 281 life equation, and can multiply bearing and component life several times over. The model shows the effect:
Interactive 2 — Oil cleanliness & bearing life
Live modelRelative bearing life vs oil cleanliness
≈ 2^((20 − code)/4) relative to an ISO 4406 4µm code of 20, reflecting the contamination life factor in ISO 281 modified life. Real factors depend on bearing type, lubricant film and particle size distribution — but the direction and rough magnitude are real.6 · Garbage in, garbage out: sampling
Oil analysis is only as good as the sample. A careless sample produces confident nonsense, so the discipline matters:
- Sample from the same point every time — a live zone in the running system (e.g. a sampling valve in the return line), not the bottom of the sump or a drain.
- Sample while hot and circulating, so debris is suspended and representative.
- Use clean bottles and clean technique — it is shamefully easy to put more dirt in than you measure.
- Keep the interval consistent and label everything — oil analysis is a trend, and a trend needs comparable points.
It's the trend, not the number. A single iron reading of 40 ppm means little; iron climbing 5 → 12 → 30 ppm over three samples means a lot. Establish each machine's baseline, set caution and critical limits against it, and act on the rate of change. Then close the loop with a root cause analysis when a sample turns bad.
Key takeaways
- One sample, three answers — oil condition, contamination, and wear.
- Wear metals are a fingerprint — Fe steel, Cu+Sn bronze, Sn+Pb babbitt, Si dirt, Na+B coolant.
- ISO 4406 cleanliness drives life — cleaning up the oil multiplies bearing life cheaply.
- Sample properly — same point, hot, clean, consistent interval; it's the trend that counts.
- Pair it with vibration — chemistry and dynamics together catch the most, earliest.