When a machine is delivered or a maintenance job is finished, the easiest thing to do is to take one sample, send it for testing and consider the job done once the result comes back qualified. That proves the oil in the tank is qualified at this moment, but it does not prove how much work the machine actually did. If the question is how the run went, the result after treatment is not enough; the difference between before and after is what carries the answer.

A comparison is only meaningful if the two samples are comparable. If the sample before treatment is taken from the bottom of the tank and the sample after treatment from the oil outlet, the difference between the two results carries the difference in sampling point with it. If the oil is at 20 °C before and 40 °C after, the change in viscosity comes into the result as well. Same sampling point, same container, similar oil temperature, similar volume processed and similar running time: once those are in place, the difference starts to mean something.

The sampling step decides whether the comparison means anything

The same sampling point, the same container, a similar oil temperature, a similar volume processed and a similar running time: only when those are in place does the difference say anything. Take the sample from the same place every time, use a container that is dry, clean and of the same specification, and keep the oil temperature close; the volume processed and the running time have to line up as well, because a tank half way through a batch and a tank that has been through the whole batch do not hold the same water content.

There is one more condition that is easily missed: the waiting time between the two samples has to be the same. Once a sample leaves the machine it keeps exchanging with the air, and both the dissolved gas and the moisture change; the longer the container stands, the larger the gap between the measured result and the state of the oil at the moment the machine stopped. Send the sample for testing as soon as it is taken, and keep the container and the sealing method the same both times.

Which way each of the four indicators moves

Get the direction right before comparing, otherwise an improvement is easily read as a deterioration. Water content and gas content are better the lower they are, so both should fall between the two samples. Cleanliness is read as a class number, and the class number has to fall for the oil to be cleaner: NAS 1638 class 5 is a requirement, not a figure where bigger is better. Breakdown voltage is better the higher it is; the requirement on the ZYD-100 is ≥65 kV (GB/T 507), so it should rise after treatment.

ZYD-100 two-stage vacuum oil purifier
A before-and-after comparison of the four indicators only means something on the same machine and the same sampling point
IndicatorReference standardRequirementWhat to look at in the comparison
Water contentGB/T 7600≤3 ppmHow far it fell, and whether it stays inside the requirement
Gas contentGB/T 423≤0.01%The size of the drop, and whether it keeps falling as the run goes on
Breakdown voltageGB/T 507≥65 kVHow much it rose, and whether it falls back again
CleanlinessNAS 1638class 5Whether the class number falls; the change around an element change is the clearest

Where to look when the comparison falls short of expectations

If the water content and the gas content have clearly fallen and the cleanliness has improved, but the breakdown voltage has not risen noticeably, that combination usually means the machine did its work and that something else in the oil is at work, such as an acid value or oxidation products that have already formed. A single degassing run cannot resolve that; it has to be judged from the state of the oil itself rather than by turning back to suspect the machine.

The other way round, if two or three of the four figures have hardly moved, the problem is most often in the sampling or the testing rather than in the machine: the wrong sampling point, a residue in the container, a sample that changed while it waited. The ZYD-100 processes 6000 L/h, so a tank takes hours to get through, and if the rhythm of sampling and transport does not match the rhythm of the process, the two samples are simply not the same batch of oil and the comparison means nothing.

Structure and cross-section comparison of filter elements made from different materials
Changing to elements of another material has a clearer effect on cleanliness than on water content
  • Write the sampling time, the volume processed, the oil temperature and the sampling point of both samples on one sheet.
  • Keep one sample before and one after every element change: the effect of the change is most visible in the cleanliness figure.
  • Data across several batches is more convincing than a single comparison; an occasional outlier has to be ruled out with the batches around it.
A qualified result after treatment only proves that this tank of oil is qualified; the difference between before and after is what proves how much the machine did.