11.4 Lubricant & Fluid Analysis
Key Takeaways
- A representative oil sample is taken from a live, turbulent zone in the system while the equipment is at operating temperature, never from the bottom of a sump or a drained line.
- Wear metals identify the failing component: iron points to gears and races, copper to bronze bushings and thrust washers, tin and lead to babbitt, and silicon to dirt ingress.
- A viscosity change of more than about 10% from the new-oil specification is a standard alarm limit, indicating oxidation, fuel or solvent dilution, or a wrong-oil top-up.
- A rising total acid number shows oxidation and additive depletion, while a falling total base number in an engine oil shows the reserve alkalinity is exhausted.
- Particle counting to ISO 4406 measures contamination while spectrographic analysis measures wear and additive chemistry — the two answer different questions and are used together.
Sub-task F-23.05 (Performs fluid analysis procedures) turns a routine oil sample into an early warning of failure. Oil analysis answers three separate questions, and a good program keeps them distinct:
- Is the lubricant still fit for service? — viscosity, acid number, additive levels, oxidation.
- Is the lubricant contaminated? — particle count, water, glycol, fuel, soot.
- Is the machine wearing? — wear metals, ferrous density, particle shape.
Taking a Representative Sample
The single largest source of bad analysis data is a bad sample.
| Rule | Reason |
|---|---|
| Sample with the machine running at normal operating temperature | Wear particles and water stay suspended; a cold, settled system gives a falsely clean result |
| Sample from a live, turbulent zone — a dedicated sampling valve on a pressurized or return line, or a drop-tube in a circulating sump | Static tank bottoms concentrate sludge and water, and drain-port samples are unrepresentative |
| Sample upstream of the filter, downstream of the components | Otherwise the filter removes the very evidence being sampled |
| Flush the sampling valve before filling the bottle | Removes the stagnant slug of oil sitting in the port |
| Use clean, certified sample bottles and cap them immediately | Ambient dust will corrupt a particle count in seconds |
| Sample the same point, the same way, at the same interval every time | Analysis is a trend tool; a single result out of context means very little |
| Label completely — equipment ID, sample point, date, hours on oil, hours on machine, oil brand and grade, any top-ups | The lab cannot interpret without this |
Laboratory Tests and What They Reveal
| Test | Measures | Indicates |
|---|---|---|
| Kinematic viscosity at 40 C | Resistance to flow | Increase: oxidation, contamination with a heavier oil, soot loading. Decrease: fuel or solvent dilution, shear-down of a VI improver, wrong-oil top-up |
| Spectrographic (elemental) analysis | Parts per million of wear metals, contaminants and additives | Which component is wearing, what dirt is present, and whether the additive package is depleted |
| Particle count (ISO 4406) | Particles per millilitre at 4, 6 and 14 micron | Cleanliness against target; the primary hydraulic system health metric |
| Water content (crackle test, Karl Fischer) | Free, emulsified and dissolved water | Seal or cooler leaks, condensation, wash-down ingress |
| Total acid number (TAN) | Acidic by-products | Oxidation and additive depletion; a rising TAN forecasts varnish and corrosion |
| Total base number (TBN) | Reserve alkalinity (engine oils) | Remaining ability to neutralize combustion acids; a falling TBN sets the drain interval |
| Ferrous density / PQ index | Total ferromagnetic material | Large wear particles that spectrographic analysis misses |
| Analytical ferrography | Particle size, shape and colour under a microscope | The wear mechanism — cutting, fatigue spalling, sliding, corrosion |
| FTIR (infrared) | Oxidation, nitration, sulphation, soot, glycol | Chemical degradation and specific contaminants |
The key limitation of spectrographic analysis: most spectrometers only reliably detect particles below about 5–10 micron. A bearing shedding large spalls can show low wear-metal ppm while it is failing, which is exactly why ferrous density and ferrography are added when a problem is suspected.
Reading Wear Metals
| Element | Common source in industrial machinery |
|---|---|
| Iron (Fe) | Gears, shafts, bearing races and rollers, housings, rust |
| Copper (Cu) | Bronze bushings, thrust washers, cages, oil cooler cores |
| Lead (Pb) / Tin (Sn) | Babbitt journal bearings, solder from a cooler |
| Aluminum (Al) | Pistons, thrust bearings, housings, or dirt (as aluminosilicate) |
| Chromium (Cr) | Plated shafts, rings, some bearing steels |
| Silicon (Si) | Dirt ingress — silica dust; also silicone antifoam or sealant |
| Sodium (Na) / Potassium (K) | Coolant leak (glycol inhibitors) or salt water ingress |
| Zinc (Zn), Phosphorus (P), Calcium (Ca), Molybdenum (Mo), Boron (B) | Additives — anti-wear, detergent, dispersant; a decline signals additive depletion |
Interpretation is about rate, not absolute value. A gearbox that reads 40 ppm iron consistently is normal for that gearbox; the same reading in a unit that has always run at 8 ppm is an alarm. Trend every point against its own history, and always correlate with hours on the oil — ppm accumulate over time in a non-drained sump.
A classic diagnostic pair: rising silicon together with rising iron and copper means dirt is entering through a failed breather or seal and is abrading the components. The fix is the ingress path, not the oil.
Water and Contamination Limits
| Contaminant | Typical concern threshold |
|---|---|
| Water in industrial oil | Above roughly 500 ppm (0.05%) is a caution; free water is always actionable |
| Water in hydraulic oil | Tighter still; hazy oil already means the saturation point has been passed |
| Glycol | Any detectable glycol is a serious alarm — it forms abrasive oil balls and destroys bearings |
| Fuel dilution (engines) | Above about 2–5% depending on the engine builder |
| Soot (diesel) | Above the engine maker's limit; increases viscosity and abrasive wear |
A quick field crackle test — a drop of oil on a hot plate at about 150 C — audibly crackles when free or emulsified water is present, giving an immediate indication between lab samples.
Grease Analysis
Grease can be analyzed too, though sampling is harder. Useful tests are consistency (penetration), oil separation (bleed), oxidation, wear metals and contamination. A purged sample taken from a bearing's grease outlet during relubrication tells you what is happening inside the bearing. Look for hardening or dry, crusty grease (oxidation or over-temperature), a soupy consistency (mechanical shear or wrong-grease mixing), and grey or metallic discolouration (wear).
Grease incompatibility is a common self-inflicted failure: mixing a lithium-complex grease with a polyurea grease can collapse the thickener into a fluid that runs out of the bearing. Confirm compatibility before changing grease types, and where in doubt, purge the bearing completely.
Converting Results into Action
| Result | Action |
|---|---|
| Viscosity outside about ±10% of specification | Confirm the correct product was used, investigate contamination or oxidation, change the oil |
| Particle count above target code | Check breathers, seals and filter condition; add or upgrade filtration; consider a kidney-loop flush |
| Water above limit | Find the ingress path — cooler leak, seal, breather, wash-down — then dry or replace the charge |
| Rising wear metals with rising ferrous density | Increase inspection frequency, correlate with vibration and thermography, plan an intervention |
| Silicon rising with wear metals | Fix the ingress path first; oil changes alone will not stop the wear |
| Additive elements declining | The oil is at the end of its useful life regardless of how clean it looks |
Oil analysis is one leg of a condition-monitoring program. Its power multiplies when correlated with vibration analysis and thermography from section 11.1: a gearbox with rising iron, a rising 1x gearmesh vibration and a warm bearing housing is a confirmed finding, not a suspicion.
A millwright takes a gearbox oil sample by draining half a litre from the bottom drain plug after the machine has been shut down overnight. Why is this sample unreliable?
An oil report on a mine conveyor gearbox shows silicon rising from 6 to 45 ppm over three intervals, with iron rising from 20 to 130 ppm and copper from 4 to 22 ppm. What is the correct interpretation and response?
A rolling-element bearing in a large fan is suspected of spalling, but the spectrographic analysis reports only 15 ppm of iron, which is close to its historical baseline. What additional test should be requested and why?