When an oil test report arrives, the part that is most often got wrong is not reading the conclusion but reading the numbers. ppm used where mg/kg belongs, a volume fraction used where a mass fraction belongs, the unit of the breakdown voltage written as kV or as V, all of these will let two reports that both look acceptable lead to opposite conclusions.
What ppm means here
In oil testing, ppm is a millionth by mass: 1 ppm is the equivalent of 1 milligram of water in a kilogram of oil. Converting that to a volume fraction requires the density of the oil, and the density changes the result, so 3 ppm and 3 mg/kg mean the same thing in this context while 3 ppm and 3 mg/L do not. This is the most common misreading in a report.
The water content is tested to GB/T 7600 and the limit for the ZYD-100 is ≤3 ppm (GB/T 7600). The gas content is tested to GB/T 423 with a limit of ≤0.01%, and the breakdown voltage is tested to GB/T 507 with a limit of ≥65 kV. The units and the scales of these three are completely different, and when they sit in the same table care is needed not to read across a line.
Several sets of numbers that are easily misread
The table below puts together the groups of values that often appear both on one report and on one machine nameplate. The point is the units and the scale, not the conclusion to remember.
| Indicator | Test basis | Unit and meaning | Common misreading |
|---|---|---|---|
| Water content | GB/T 7600 | A millionth by mass, in the mg/kg range | Treating ppm as a volume fraction, which after a conversion by density gives the opposite conclusion |
| Gas content | GB/T 423 | A percentage by volume | Reading 0.01% as 1%, a difference of a hundred times |
| Breakdown voltage | GB/T 507 | kV | Reading kV as V, or taking a single result for the average |
| Filtration accuracy | Machine rating | μm | Converting to a mesh count and treating it as the same indicator |
| Cleanliness | NAS 1638 | A class number | Treating the class number as a measurement and comparing the numbers directly |
The order to read a report in
- Start with the units column and confirm what unit each number is in before looking at the numbers themselves.
- Check the standard number of the test basis, because the measuring conditions for the same indicator differ from one standard to the next.
- Note whether the figure on the report is a single result, an average or a maximum; this matters most for the breakdown voltage.
- Confirm that the sample number, the sampling time and the sampling point on the report belong to the machine you are dealing with.
- Compare the range of oils the method applies to; a conclusion cannot be carried over to an oil outside that range.
- If anything is unclear, ask the testing party about the sampling and the measuring conditions rather than converting the values yourself and drawing a conclusion afterwards.
Reading the report against the machine rating
A test report describes the state of the oil, while the machine rating describes the capability of the machine, and a judgement needs both of them together. The filtration accuracy of the ZYD-100 is ≤1 μm, the cleanliness corresponds to NAS 1638 class 5, the breakdown voltage is ≥65 kV (GB/T 507) and the gas content is ≤0.01% (GB/T 423). These figures describe the state that the machine should reach after processing, not any single number standing on its own in a report.
The ppm on a report and the ppm on a machine nameplate have to be brought into line before they are compared at all. Subtracting two numbers whose units do not match produces a difference that means nothing, and averaging two values taken to different standards means just as little.
Besides the report, the running data of the machine is a kind of report as well: the vacuum level, the pressure drop, the current and the oil temperature all have their own units and their own normal ranges. Only when they are read together with the test report can it be told whether the change in the oil came from the treatment itself or from the sampling and the measuring.
The numbers themselves are rarely the problem; the problem is the units, the test basis and the link between the sample and the machine.
