There is a saying on site: when the test result is wrong, suspect the instrument first. Most of the time the instrument is the last thing that needs suspecting. The sample that goes into it stopped being the oil from that tank long before. Sampling is the first step in the testing chain and the one most easily allowed to go out of control, and it leaves little room for carelessness.
Pitfall one: the container carries water
If the sample bottle is not dry, or if a freshly washed bottle is filled by inverting it, the water is charged to the oil at this step. Water content is measured to GB/T 7600 and the target value is ≤3 ppm, which is a very low threshold; a thin film of water left on the inside wall of the container is enough to lift the reading as a whole. The sample bottle and the cap must therefore be dry, the seal on the bottle mouth must be free of residue, and the bottle must be kept upright in transit so that liquid clinging to the inner wall is not thrown onto the cap.
Pitfall two: left standing open
If the sample is not sealed immediately, or the stopper is not pushed right in, the oil in the open container is doing two things at once: moisture from the air dissolves into the oil, and moisture in the oil passes out into the air. Neither rate is zero, so the measured water content can come out high or low, depending on how long the sample stood, on the temperature and on the relative humidity of the air. The same applies to gas content — leaving the sample open lets the dissolved gas escape, and the ≤0.01% measured to GB/T 423 is only an upper limit requirement; once the sample has stood open, the data no longer represents the true state of the oil in the tank.
Pitfall three: the oil has not been mixed
After the oil has stood in the tank for a while it stratifies: the lower density material floats on top, the higher density material settles below, and the suspended particles sit in a layer somewhere in between. Sampling from there takes whichever layer the height of the suction pipe and the arrangement of the tank happen to select. If the intention is to judge the state of the whole tank, a sample from the middle layer is the representative one; if the intention is to judge the filtration performance downstream of the filter element, the sample should be taken at the outlet of the filtration unit. These two positions cannot replace each other, and the sample has to state where it was taken.
Pitfall four: particles hidden at the bottom of the bottle
If the sampling pipe is pushed down to the place at the bottom of the tank where sediment collects most easily, what is drawn up is not only oil but also the particles and deposits at the bottom. They settle at the bottom of the bottle and are then counted into the cleanliness figure, so the NAS 1638 grading is distorted. Conversely, if the sample is left untreated after sampling and stands in the bottle for a while before being sent for testing, the suspended particles settle to the bottom of the bottle and the result changes in the same way. Breakdown voltage is measured to GB/T 507 and is required to be ≥65 kV; particles and free water both affect it directly, so the uniformity of the sample matters more for this figure than for water content.
| Sampling step | Symptom when it goes wrong | The figure affected most clearly | What to do |
|---|---|---|---|
| Container carries water | Water content reads high across the board | Water content (GB/T 7600) | Keep the sample bottle and cap dry |
| Left standing open | Readings fluctuate and cannot be reproduced | Water content, gas content | Seal immediately after sampling and transport upright |
| Oil not mixed | Large differences between samples from the same tank | Cleanliness, breakdown voltage | Fix the sampling point and state the position |
| Particles at the bottom of the bottle | Cleanliness grade comes out too low | Cleanliness (NAS 1638) | Sample away from the sediment zone |
Preparation list before and after sampling
- Sample bottle dry, seal intact, bottle transported upright
- Fix the sampling point and write on the test request which section the sample came from
- Seal the sample immediately and avoid any time standing open
- When a re-test is needed, keep a reserve sample from the same batch
When a test result comes out abnormal, go back to the sampling step first; whether the container was dry, which layer was sampled and whether the stopper was pushed right in are three things better checked before anything on the instrument is touched.