Many customers treat oil analysis as a health check report: when an indicator is out of limits, the oil is changed. That use wastes the most valuable information on the report. The real purpose of the report is to answer the questions that cannot be seen on site, and when it gives no answer, it usually means the sampling method itself has blocked the answer.
The sampling point decides whether the report is of any use
For the same machine, a sample taken from the running line and a sample taken from the bottom of a settling tank produce the reports of two different machines. The oil in service is flowing and mixed, and it reflects the state of the oil the machine is actually in contact with at that moment; the oil at the bottom of a settling tank has undergone sedimentation and stratification and reflects the state of the storage stage. The two sets of figures differ not because the analysis is wrong, but because they are describing two different things.
The common problem is that the sampling step itself erases the information: the sample bottle has not been cleaned, water from the previous sample is left in the bottom, the container is not sealed properly and takes in air on the way to the laboratory. All of this makes the water content and the particle count results deviate from the true values, and these two are exactly the indicators most likely to be used directly to draw a conclusion.
| Sampling location | What it reflects | Common misreading |
|---|---|---|
| Sampling point on the running line | The oil the machine is actually in contact with at that moment | Assuming that it represents the condition of the oil in the tank |
| Sample from the bottom of a settling tank | The sedimentation and stratification result of the storage stage | Assuming that it represents the condition of the oil circuit inside the machine |
| Sample after filtration | The quality of the oil after the treatment stage | Comparing it directly with the data from before the treatment |
| Sample from a new oil drum | The initial condition of the oil as received | Assuming that it is necessarily conforming |
What the report can answer and what it cannot
The report is good at answering questions about trends: whether the water content has risen or fallen since last quarter, whether the particle count class has got worse, and how far the breakdown voltage is from the target. All of these need historical data from the same measuring point and the same method for comparison; a single result only marks one point in time.
It is not good at locating the fault. When the gas content rises the laboratory can report a figure, but which component is discharging and which section of the sealing is leaking has to be established on site. When the water content rises, whether the breather is damp, a seal has failed or the water was carried in with the top-up oil also has to be established on site. The report gives the direction; the specific cause is found by people.
The report and the site actions verify each other
A workable practice runs like this: fix the measuring points first, fix the test items second, then sample and send for analysis on a fixed cycle. When the data is out of line, run one round of treatment with the purification equipment first, send for analysis once before and once after the treatment, and compare the two results by the same method. This step is a two-way verification: the report verifies whether the treatment was effective, and the treatment result in turn verifies whether the measuring points on the report were chosen correctly.
- Sampling before and after the treatment must be done at the same measuring point and within the same time window, otherwise the two sets of data cannot be compared.
- The test bases on the report should be aligned with the acceptance bases of the equipment; the usual approach is to correspond to GB/T 7600, GB/T 423, GB/T 507 and NAS 1638.
- If the data has not improved after one round of treatment, suspect the sampling and the test method before suspecting the capability of the equipment.
- Laying the past reports out in chronological order and reading the trend explains far more than a single report can.
It should be emphasised that within the customer's organisation the laboratory is more like a third-party reference. What it provides is a reading that is not influenced by the site operations, and the customer uses it to judge whether their own assessment is right. Treating it as a substitute for an authoritative conclusion tends to cause problems instead.
The value of the report lies not in the figures themselves but in whether the sample was taken at the right measuring point and whether it was taken in the same place by the same method as the previous one.