Does anyone have any recommendations for the frequency of sampling transformer oil? We just had a transformer red tagged out for having too high of an explosive gas count. When I asked the electricians how often they sample the oil, they said annually. This seems to me to be not nearly enough for something so critical, considering that I sample the majority of my gearboxes and reservoirs quarterly. What is the industry standard for transformers?

While I was employed at a large power generating utility in SA we use to sample all transformers, e. g. generator and unit transformers (420kV/700MVA, 275kV, 88kV/10 or 20 MVA), on a 6 monthly basis. Resamples / sampling frequency increases were only conducted where any out of specification condition was detected.

Based on the duty cycle as well operational and safety considerations these transformers could / should not be sampled at a lower frequency. Distribution transformers on the national grid were sampled annually but frequencies were decreased to 6 monthly on suspect units.

There is a British Standard providing basic guidelines with regard to sampling procedures and frequencies - BS5263 : 1975. I do not have a copy and cannot garuantee that it will give you exactly what you want but it will point you in the right direction.

BE SAFE

quote:

---

Originally posted by Mark K.:
Does anyone have any recommendations for the frequency of sampling transformer oil? We just had a transformer red tagged out for having too high of an explosive gas count. When I asked the electricians how often they sample the oil, they said annually. This seems to me to be not nearly enough for something so critical, considering that I sample the majority of my gearboxes and reservoirs quarterly. What is the industry standard for transformers?

---

Tranformer oil, being a closed system six months oil sampling frequency is generally followed. The important tests that to be monitored is Water/moisture, Dielectric strength, TAN value and insolubles. Mostly oil life comes to years but leakage is to be watched.
Kumar
kumartr2@yahoo.com

Good Day group,
We have been sampling transformers and OCB's for the last 10 years or so. The interval has always been 1 year betweeen samples, as previously noted the interval gets reduced if a problem is encountered.
Generally the issue of sampling frequency is brought to conclusion when it is understood what you are looking for and at what point can you identify the problem and how long is it from the point of detection until a failure occurs. This is the infamous RCM P/F interval where P=the point of detection and F=the catastropic failure point. So if your F is 6 months after the P then you should consider sampling at 3 month intervals minus preperation time. Fortunatly we have found that 1 year intervals has kept us operating without a major outage.

If you have any questions please contact me.

regards.......

[QUOTE]Originally posted by Mark K.:
...We just had a transformer red tagged out for having too high of an explosive gas count...QUOTE]
How long it take to concentrate enough gas to explode?
The answer to my question less 50% is a good aproach to answer your question.

As this is what I do for a living, my comment would be that it depends. For DGA on transformers, resampling frequency depends on how the transformer is gassing. If normal, than once a year, if at nuclear facility then the insurance company dictates and it is every 3 months. Some of our sampling frequencies are as short as 1 day depending on the rate of gas generation. For regular oil quality, especially in sealed transformers, yearly is probably adequate. For OCBs and LTCs there are special diagnostic testing programs using fluid anlaysis to help determine the frequency on which you should test or actaully go into the unit.