Different values from OELCHECK and another laboratory

Our presses are in use globally. In our operating manual, we recommend semi-annual trend analyses for hydraulic and transmission oils. One of our service technicians has had a hydraulic oil tested in our laboratory in the USA. We have now also sent the sample to OELCHECK. In both laboratory reports some values differ marginally, but the wear element iron can be seen clearly in both of them. What could have caused this?


Examples of precision specifications




Measurement result

Possible range of values

      OELCHECK External
from to
E DIN 51399-1
ASTM D5185
0,25* mean value
0,52* mean value 0,80
43 36 35 45
E DIN 51399-1
ASTM D5185
0,20*mean value
4,3*mean value0,50
386 422 361 447
E DIN 51399-1
ASTM D5185
0,15*mean value
0,083*mean value1,1
193 221 192 223
Acid value
DIN 51558
0,15*mean value
0,141*(mean value + 1)
0,88 1,12 0,86 1,14
Water content
DIN EN ISO 12937
ASTM D6304
0,06877*mean value0,5
0,4243*mean value0,6
140 180 120 200



When they are processed in the USA they are not only processed according to different standards, but also in terms of comparability, not reproducibility. Reproducibility and comparability are often, depending on the value measured, dependent on the concentration or, in the case of element determination, also on the element in question. The results are more scattered when analysing comparability than for reproducibility and again at low concentrations more than for high concentrations. In the DIN this is illustrated for the element iron as an example: At a concentration in the range of 1-10 mg/kg fluctuations in the results of 15% from the mean value for reproducibility (in the same laboratory) and deviations of 60% from the mean value for comparability (different laboratories) are tolerable. At higher concentrations, from 10-1,000 mg/kg, the results for reproducibility may only vary by 5% and those for comparability by 25%. The numbers show that with a mean value of 50 mg/kg, values of 44 mg/kg and 56 mg/kg are considered to be valid, if the results from two different laboratories are compared. Conclusion: If a particular sample is analysed in different laboratories, it is a question of a comparable, but not reproducible process. Also comparable values determined according to a standard can in part differ considerably from each other. The OELCHECK laboratory in Germany has markedly improved the tolerance range for comparability, because in both locations analysis is carried out with identical standards, equipment types and calibration standards. In that way the values are of the same standard as they would be if they were within the narrow range for reproducibility. By the way: In case of doubts regarding a diagnosis by an external laboratory you can, for a small fee, have the diagnosis compiled by OELCHECK tribologists.


OELCHECKER Summer 2015, page 8