There was a series of papers presented at ASTM in December 2005 showing RPVOT was very accurate for oils using BHT/DBPC (oils RPVOT was developed for) but was not as useful for turbine oils using amine/phenol combinations. Basically when the phenols depleted, the varnish/AN started accumulating even though RPVOT was still in good shape since amines just starting to deplete. Also some of the newer antioxidant systems are acidic causing AN to decrease with initial use followed by increase.
Are you monitoring a steam or gas turbine? When were the RPVOT performed: 2 data points/5 dates? I am not that familiar with particle counts but it looks like you had a very high particle count prior to cleaning with the electrostatic cleaning but the particles started to increase again during the last sample or is increase within reproducibility of test?
I copied the below response from the bottom of page 2 of this forum when sweetening was also discussed. Gives a link to get some of the data that was presented at ASTM.
Rick - Our paper presented at ASTM Norfolk last December showed our results of 3 samples, two used turbine oils and the new oil, sent to 4 different labs, in four different parts of the world. The samples to three of the labs were unmarked to try get get some real "blind" testing results.
Of the blind labs, two of them could not even tell the difference between used oils and the new oil (by RPVOT values, of course the color would have been a dead give-away, but they must of neglected to look at that)
One blind lab maybe looked at the color of the oils, and gave results that sort of made sense, at least the new oil was higher in RPVOT than the used oils. However, they were still way off in the realive RPVOT values.
The only lab to get the results right, was the same lab than blended the oil!!And you guessed it, they were not blind, and knew it was their oil. Lo and behold, they got the new oil value right!
What does that tell you? It tells me that RPVOT does not work with true unknown samples, in true blind testing, and that most RPVOT results are made-up by random number generators or dart-boards!!
(rememeber the accountant job-applicant joke..how much do you want it to be, boss?)
email me at andy.sitton@focuslab.co.th if you want to see the data