The breakdown voltage is the indicator that most often holds up acceptance testing on site. The machine has finished its run, the oil looks clear and the water content has come down, and yet the breakdown voltage will not move up towards 65 kV (GB/T 507). The quickest way out at that point is to run the machine longer or to treat the oil a second time, but both of these only postpone the question. First it has to be established whether the machine failed to achieve the figure, or whether the oil itself is no longer within a range that can be recovered.
Three sets of comparisons are enough to settle where the responsibility lies. The order cannot be changed, because each step further on costs more.
First comparison: has the water content actually come down?
This is the most important and also the most easily skipped step. Take a sample before treatment and a sample after treatment, and measure the water content once in each case with the same method, on the same day and with the same instrument. If the figure before treatment is 200 ppm and it is still 180 ppm after treatment, then the breakdown voltage not rising is the result and not the cause, and there is still work to be done on the equipment side. Only once the water content has really come down to within 3 ppm (GB/T 7600) does a discussion of the breakdown voltage make sense.
If this step is skipped, the conclusion goes completely off course: a machine whose dehydration does not come anywhere near the target is judged a conforming unit, or a batch of oil is judged non-conforming oil.
Second comparison: how much does the result change on another machine with the same batch of oil?
Take the same batch of oil, split it into two portions, process one with this machine and the other with another machine in normal condition, at the same temperature and for the same length of time, and then measure the breakdown voltage in each case. If the results from the two machines are close, the difference is not in the equipment but in this batch of oil. If one of them is clearly higher, some part of the process on this machine still has a gap. Conversely, if the breakdown voltage does not rise on the other machine either, where the responsibility lies is clear.
When this comparison is made, the process conditions on both sides have to be brought into line: the temperature, the feed rate, the running time and the state of the pre-filtration. Changing the machine without changing the conditions is just as unreliable.
Third comparison: is a repeated measurement of the same sample stable?
The breakdown voltage is extremely sensitive to the condition of the sample. Whether the sample bottle is clean and dry, whether the cap is sealed, whether the sampling line was flushed and drained of residual liquid beforehand, and whether the sample has stood in the air for too long all change the reading directly. Measure the same sample twice in succession: if the two readings differ widely, the problem lies in the sampling and the measurement and has nothing to do with either the machine or the oil. This comparison takes the least time and yet often settles the question.
What it means when the water content is within limits and the breakdown voltage still does not rise
At that point the responsibility lies essentially with the oil itself. Water is only one of the variables that affect the dielectric strength; three other kinds of contaminant are at work:
- Particles: metal swarf and conductive particles in particular align in the electric field and bridge the electrodes, forming a continuous path and causing the breakdown voltage to fall sharply.
- Oxidation products: oil that runs for a long time at high temperature produces polar molecules, which adsorb onto the charged surfaces and change the dielectric characteristics of the oil, which shows up as a slow and continuous decline.
- Dissolved gas and other polar substances: dissolved gas and substances containing sulphur or copper change the overall withstand capability of the oil, and they are a sign that the oil has already aged.
How to proceed
| Result of the comparison | Points to | Next action |
|---|---|---|
| The water content has barely fallen | Dehydration on the equipment side is not achieving the target | First work through the oil inlet, the condensate draining and the vacuum level in the order given in the water content topic |
| The two measurements differ widely | The sampling or the measurement | Repeat the sampling, and standardise the flushing, the draining of the residual liquid, the sealing and the time before the sample is sent for testing |
| The result is clearly better on another machine | The process conditions on this machine are not sufficient | Compare the temperature, the feed rate, the running time and the pre-filtration |
| The result is just as poor on another machine | A problem with the oil itself | Measure the cleanliness and the gas content as well, and decide on the basis of how far the oil has aged whether it can still be used |
In the end there are only two ways forward. If the oil can still be recovered, bring the cleanliness and the gas content into the required range as well and measure again. If the oil has aged to the point where the water content, the cleanliness and the gas content all fail, further treatment only tidies up the figures, and oil that has to be replaced has to be replaced.
When the breakdown voltage does not rise, first confirm that the water content has come down. Without that step, every conclusion after it is a guess.