Does anyone have some advice / information on monitoring transformer oil? Also what concerns should one have for PCB's while performing such inspections?

We have a consultant who samples our transformer oils and conducts various tests.
Dissolved Gas Analysis ( DGA ) by ppm
Standard KV Dielectric Breakdown
Moisture ( Karl Fischer )
Neutralisation Number

other tests as options
Interfacial Tension
Dissolved Metals ( ICP ) by ppb
Furan Analysis
Liquid Power Factor
Oxidation Inhibitor
PCB Content ( for contamination purposes )

Contact your local Transformer Consultant for these types of testing.

ON PCB's.
Avoid contact - enough said.

What would you expect to see most often in a sample that was failing?

I have looked over several on-line services that provide oil testing for transformers, but question the methods. There are many tests that can be performed. However, are test such as wear debris analysis really necessary in such systems? Here is some of the more popular test I keep coming across.

Dissolved Gas in Oil Analysis (DGA) (Oil Gas Analysis)
Buchholz Gas Analysis
PCB Testing (Polychlorinated Biphenyls)
Electric Strength
FFA
Resistively
Water Content
Acidity
Fiber Estimation
Color
Metal in Oil Analysis

DGA is indication of transformer fault. C2H2 key gas for arcing. H2 key gas for partial discharge. C2H4 key gas for hot metal in contact with oil. CO and CO2 key gases for solid insulation overheating. Analysis can be complex. Usually lab will provide some expert analysis on these results if abnormal.

Moisture is indication of abnormal water intrusion which can compromise the dielectric strenght (can lead to arcing), long-term can lead to corrosion and oil degradation.

KV is direct measure of dielectric strength required to prevent failure. S

Neutralisation Number – normally < 0.1 mm/kg KOH. Higher starts to indicate oil degradation.
Interfacial Tension – an earlier indicator of oil degradation
Metal particles – an indicator of oil pump bearing problems.
Furan Analysis – an advanced indicator of solid insulation conditionn.
Liquid Power Factor – Typically not done during lab tests. A quick easy field test which tells you roughly the same thing as dielectric strenght (primarily looking for moisture).

Waylon,
I am from near your area and have successfully used a company for my transformer testing. (77 transformers) PM me and we can talk

For transformers, is there a capacity/size limit for lubricant analysis to be economical below which the strategy is just change the oil periodically?

Is there a standard or paper on Lubrication for electrical equipment other than motors, generators & alternators?

This message has been edited. Last edited by: Josh,

I'm not trying to be picky on terminology, but I want to mention that the oil is not used as a lubricant in transformers, it is used as a coolant and dielectric (electrical insulator)

I haven't ever heard of a periodic oil changeout for transformers. But if the transformer is non-critical you may just sample very infrequently or not at all. The primary purpose of sampling is to check the condition of the equipment (signs of electrical faults developing). Checking the condition of the oil is a secondary function. The faults can develop over short time periods. Degradation of oil is usually a very slow process (except perhaps in certain LTC's, OCB's or other equipment where frequent arcing is expected).

IEEEC57-104 describes DGA.

IEEE 62-1995 describes all the other electrical tests and gives typical limits for new and in-service oils.

As far as selecting which tests to do, I would give a rough order from most important to least important:

water - very important - most frequent contaminant
dga - very important - best diagnostics
dielectric - moderately important

color/visual - cheap easy - might as well check it

IFT and NN - somewhat important - keeps an eye on long-term oil degradation

metal in oil analysis - not important except on very critical transformers and where oil pump bearing problems are suspected (certain types of sliding brass bearings)

particle count - not important - checks for contamination which is not likely

furanic compounds - not important unless you see something on CO and CO2 that makes you suspect long-term aging of the solid insulation and you want to evaluate further

PCB's - Does not have to be an ongoing test. Once the transformer is cleared of PCB's, it should be labeled as such and need not be tested further.

power factor - not recommended as a lab test. This is sometimes a convenient test to be done in the field using test equipment which is available for transformer testing... checks similar characteristic as the dielectric withstand.

Your lab will probably have some recommendations. A lot of times they price groups of tests in a package.

Some links:

A primer on DGA
http://www.nttworldwide.com/tech2102.htm

example report with selected limits:
http://www.nttworldwide.com/docs/nttdiag.pdf

Great info in section 6 of the following document:
http://www.usbr.gov/power/data/fist/fist3_31/fist3-31.pdf

Also for those who might be just getting started, remember that drawing samples requires some training both for equipment/personnel safety and to get good results.

This message has been edited. Last edited by: electricpete,