http://www.machinerylubrication.com/Read/428/varnish-power
Oz lost it's on shore refining capacity, and as ALL the lube base-stocks were being imported, they only imported GrII and up (according to suppliers)...they changed, and we didn't know.
The GrI while not as oxidation resistance, had MUCH better capacity to handle oxidation products. The GrII would show up fine in all the standard testing for oxidation, but couldn't handle the products that
a) were there already; and/or
b) that it made.
So the underground tank got filled up with GrII, without us knowing, then started to be added to the system.
First the centrifuges started showing black mollasses, then the governor filters started to clag up, as the varnich falls out of suspension in the cold spots (where you find it isn't where it's being made in the system).
Bloke who came after me did a full replacement, plush caustic flush of the affected units, and it was back inside a year.
I commissioned an undergrad thesis on Varnish formation, and using a specially calibrated flat bed scanner to carry out patch colorimetry testing (wasn't YET an ASTM standard, and they were hideously expensive to get done, but the colour changes in the insoluble fraction were the only indicator that you would get before you had problems.
We still have problems on start-up of the remaining units, as the (hydrogen) seal oil strainers are very very fine, and clag up until the system gets back to running temperature.