An oil analysis report runs to twenty or thirty lines of numbers, a red or yellow status circle in one corner and a one-line recommendation. At most plants the reader stops at the color of the circle. Six parameters are enough to tell you what condition the machine is in and what to do next; everything else on the page is there to explain those six. The example used here is a rotary screw air compressor, but the same method of interpretation applies to gearboxes, hydraulic systems and refrigeration compressors, each with its own limits.
One result on its own means little. It becomes information only when set against new oil and against earlier samples from the same machine, which is why a good report lines up the last three or four samples side by side. The plant's part of the job is sampling correctly: the same sampling point every time, the machine at operating temperature, the stagnant oil in the sampling valve flushed off first, a clean bottle, and machine hours and oil hours written on the label. A badly taken sample makes every number that follows worthless.
| Parameter | What it measures | What a rise or fall means | What to do |
|---|---|---|---|
| Viscosity at 40 °C (mm²/s) | How thick the oil is, checked against new oil of the same product | Up by more than 10 %: oxidation, loss of light ends by evaporation, or a heavier oil has been added. Down by more than 10 %: a thinner oil has got in, or cleaning solvent was left in the machine | Find the contamination source; if there is none, the oil is at the end of its service life – schedule a change |
| Total acid number, TAN (mg KOH/g) | Acidity of the oil, read as the increase over new oil of the same product (every product has its own starting TAN) | A slow, steady rise is normal. A sharp rise between two samples means the oil is running hot or its antioxidant is depleted. Rising TAN together with rising viscosity means deposits are about to form | Check operating temperature and cooling; change the oil once the increase passes the laboratory's limit for that product |
| Water (ppm, Karl Fischer) | Dissolved and suspended water in the oil | High: condensation from running at low load, condensate not drained regularly, humid surroundings, or a leaking cooler. Water emulsifies the oil, causes rust and promotes foaming | Drain condensate, review the load cycle, check the cooler; change the oil if the level is very high |
| Iron, Fe (ppm) | Wear of rolling bearings, airend rotors and housing | A steady low-level rise is normal wear. A sudden jump means a bearing or the airend is failing, especially when large particles are also reported | Check vibration and bearing temperatures; resample after 250–500 hours; plan an overhaul if the rise continues |
| Copper, Cu (ppm) | Wear of copper cooler tubes or bronze bushings, or the oil attacking copper at high temperature | Rising together with water: the cooler is corroding. Rising together with high TAN: the oil has degraded and is attacking copper | Inspect the cooler; change the oil if TAN is high |
| Silicon, Si (ppm) | Dust drawn in from outside | Rising together with aluminum and particle count: the intake air filter is leaking or overdue, and dust is abrading the surfaces | Replace the air filter, inspect the intake line; resample after 500 hours |
One further line is worth a look: the particle count to ISO 4406 (three figures such as 21/18/14) gives overall cleanliness and usually climbs before iron does.
The metals on the report are measured by ICP spectroscopy. The method counts elements, not molecules: iron, copper and aluminum are wear; silicon is dust; zinc, phosphorus, calcium and magnesium are elements of the additive package. The additive lines are where reports are most often misread. When antioxidant and anti-wear additives are "used up", their molecules change chemically but the elements stay in the oil, so zinc and phosphorus concentrations hardly fall. Do not take a drop in zinc or phosphorus as proof that the additives are gone; judge remaining additive life by TAN against new oil, viscosity and infrared spectroscopy (FTIR) for oxidation. A clear change in zinc, phosphorus or calcium against new oil mostly tells you something else: the oil has been diluted or mixed with an oil carrying a different additive package.
| Parameter | Normal | Monitor | Act |
|---|---|---|---|
| Viscosity vs. new oil | ±5 % | ±5–10 % | beyond ±10 % |
| TAN vs. new oil | small, steady increase between samples | clear increase between two consecutive samples | rising fast, or beyond the laboratory's limit for that product |
| Water, ppm | below 200 | 200–500 | above 500 |
| Iron, ppm | below 20 | 20–50 | above 50 or a sudden jump |
| Copper, ppm | below 10 | 10–30 | above 30 |
| Silicon, ppm | below 10 | 10–20 | above 20 |
These are general guide values for screw air compressors running on good-quality mineral or PAO oil. The governing limits are the compressor manufacturer's and those of the laboratory that issued the report. The water and TAN limits above do not apply to PAG and ester oils: those base oils start from a different TAN, absorb moisture and can hydrolyze, so use the oil manufacturer's and the laboratory's own limits for them. Refrigeration compressors need far lower water; gearboxes tolerate more iron.
Viscosity up, TAN up, metals normal: the oil is aged, the machine is fine. Change the oil and look at operating temperature. Water high, calcium high, viscosity up: contamination from the surroundings, typical of cement plants and wood processing; drain the low points, change the oil and keep dust and moisture out of the intake. Iron jumping, particle count up, oil parameters normal: the machine is wearing and the oil is not the problem; inspect the rolling bearings before the failure becomes a major one. The case study on the CPI-9370-46 field trial publishes the data table for three representative compressors from the trial, so you can check this method against real numbers.
This article is general guidance; the specific conditions of each machine must be assessed by an engineer.
