The basis for judging breakdown voltage is GB/T 507, and the indicator for the ZYD series is a breakdown voltage of ≥65 kV. On site one situation with this test is common: the same cup of oil reads above 70 kV the first time and drops into the sixties when it is measured again after an interval. Most people suspect that the oil has deteriorated; far fewer suspect the instrument or the way the measurement was taken.
If both measurements used the same electrode and the same cup of oil, with cleaning and drying in between, then exactly one thing changed between them: the surface of the electrode.
What the layer on the electrode surface does
Breakdown voltage measures the ability of the oil to withstand an electric field, and the measurement takes place in the gap between two electrodes in the oil cup. When the sample is poured into the cup, the oil wets the electrode surface; after it stands for a while, a very thin film of oil and a trace of material precipitated out of the oil has formed on the electrode. The field does not fall evenly across the whole plane. As soon as the electrode surface carries a little roughness or a little deposit, the field concentrates at that spot, and the discharge that should have developed evenly across the whole gap is drawn forward to that point. The reading the instrument then gives is too low.
That explains a common observation: when the reading is low it is often not a little low but low by a margin, and repeating the measurement takes it further down. Each test leaves another layer of deposit on the electrode.

| Observation | Possible cause | How to tell them apart | What to do next |
|---|---|---|---|
| The second reading is clearly lower than the first | Residue or contamination on the electrode surface | Retest the same cup of oil with a cleaned electrode | If the reading recovers, the problem was in the measurement |
| The reading stays low after changing the electrode | The oil itself: water content has risen, particles are present, there are traces of previous breakdown | Take a fresh sample and test the water content by GB/T 7600 | Check the oil first, the machine afterwards |
| Two readings of the same cup differ widely | There are bubbles in the oil | Retest after the sample has stood | Let it stand until the bubbles have risen |
| Readings from different batches cannot be compared | The oil temperatures differ | Record the oil temperature for both readings | Compare at one and the same oil temperature |
What state the electrode has to reach
There are three common steps. Rinse both electrodes with a solvent compatible with the oil so that no trace of the previous cup is left behind; wipe the surface dry with a soft lint-free cloth so that no fibre remains; then rinse once more with the oil to be tested, so that the electrode surface sits in the same medium as the measurement. Do not touch the metal face with your fingers during the process: hold the insulator, because a single fingerprint on the electrode is enough to shift the reading.
There is one more step that is easily overlooked: the cleaned electrode has to dry until no liquid film remains. A breakdown voltage test carried out with water present pulls the reading down as well, and the result then looks very much like an oil with excessive water content, which is easy to misjudge.
What a complete test has to record
Write four things on the same line: the sampling point and the sampling time, the oil temperature, whether the electrode was cleaned, and the reading together with the standard it is judged against. The four indicators of the ZYD series are separate from one another: water content ≤3 ppm (GB/T 7600), gas content ≤0.01% (GB/T 423), breakdown voltage ≥65 kV (GB/T 507), cleanliness NAS 1638 class 5. The first three come from a sampled test, and the fourth reflects the result of the filtration. Read on its own, the breakdown voltage line means little; set it beside the water content in the same table and the problem can usually be located either in the measurement or in the oil.
- When one instrument tests a series of samples, clean the electrode between samples instead of repeating a measurement to settle the reading.
- When an old oil cup is used for a new sample, the inner wall has to be cleaned too; a deposit on the wall acts exactly like a deposit on the electrode.
- When the reading is written down, note the value on the dial and the standard number together; a value separated from its basis cannot be judged as pass or fail.
When the breakdown voltage reading drops, first decide whether the oil has deteriorated or the electrode has become dirty. Retesting the same cup of oil with a cleaned electrode takes two minutes and separates the two.