The most common misunderstanding when an oil test report arrives is to read it as a description of the oil in its present state. The report does describe a state, but the state it describes is the oil in one sample, at one sampling point, at one moment. Take away those limits and the figures in the report lose their reference.
Which moment the report describes
The test methods are fixed: water content by GB/T 7600, gas content by GB/T 423, breakdown voltage by GB/T 507, cleanliness by NAS 1638. The same oil on the same instrument with the same method gives a reproducible result. But a fixed method is not a fixed sample. A test report is really three pieces of information put together: a sample, a sampling moment and a set of readings.
Without any of the last two the report is incomplete. A report that says pass but does not state the sampling date cannot be traced to a particular batch of oil. A report that gives the date but not the sampling point cannot say whether it represents the top of the tank or the oil outlet. These details take only a few lines on the report, yet they decide whether those lines of figures can be used at all.
What happens to the oil over three months
Whether the oil is held in an open tank, a closed tank or the tank of the machine, the direction of the change differs, but the following points are common to all of them:
- Oil in contact with the air redissolves moisture from that air, and the water content follows the ambient humidity and the time in storage.
- The dissolved gas in the oil returns to equilibrium while it stands, and the gas content reading moves towards that equilibrium value.
- Oxidation continues throughout the storage period, and the acid value and the colour change with it; this is a curve that only goes one way.
- If the oil surface in a container is left open, particles falling in and moisture condensing arrive at the same time.
The direction and the size of these changes depend on the storage conditions and on the oil itself, and the differences are large, so an old report cannot be used to calculate the state today. A water content of ≤3 ppm measured three months ago only says that the oil on that day was within the limit; whether the oil in the same tank is still within it today can only be answered by sampling again.
Where an older report may still be quoted
| Use | Can the old report be quoted | Reason |
|---|---|---|
| Acceptance record when a batch of oil is received | Yes | It records the state on arrival |
| Judging whether the treatment improved the oil | Yes, as a comparison only | The difference between the two reports is informative |
| Deciding whether the machine may be put into service | No | The current state has to be reflected |
| Acceptance for return to service after transformer maintenance | No | The result of a fresh sample after treatment governs |
| Judging whether the oil has to be changed | No | Oxidation and moisture uptake continue all the time |
Three things that make a report traceable
Test data loses its currency very often not because the figures are wrong but because the context is missing. Write the following three into the sampling record and the report stays usable for a long time:
The sampling date and time. This one sets how fresh the data is.
The sampling point. The top of the tank, the middle of the tank and the oil outlet do not give the same oil; once the position changes, the meaning of the data changes with it.
The oil temperature and the state of the machine at the moment of sampling. A sample taken while running and a sample taken after the machine has stood represent two different states, and the oil temperature also affects the breakdown voltage reading.
Those three, together with the test values themselves, make up a piece of data that can be reused. Leave out any one of them and the data proves only the instant at which a test happened.
A test report has no expiry date, but every set of figures answers only for the moment it belongs to.
