Transformer Oil Analysis

[wirenut1980]

Hello, I am dealing with an issue of a utility transformer pad-mount bay-o-net fuse blowing once ever six months or so (on average) at a couple car wash locations. Back in 2002, an oil sample was analyzed and showed 14 parts per million (PPM) of acetalyene. Relatively low as I understand it. I just had another sample analyzed this month and the report came back less than 1 PPM. Is it possible for impurity levels to decrease over time? I wonder if the transformer was changed out sometime between 2002 and now...

Thanks!

 

[zog]

Hello, I am dealing with an issue of a utility transformer pad-mount bay-o-net fuse blowing once ever six months or so (on average) at a couple car wash locations. Back in 2002, an oil sample was analyzed and showed 14 parts per million (PPM) of acetalyene. Relatively low as I understand it. I just had another sample analyzed this month and the report came back less than 1 PPM. Is it possible for impurity levels to decrease over time? I wonder if the transformer was changed out sometime between 2002 and now...

Thanks!

Well there are a few conflicting standards on what is acceptable or not but 14PPM of C2H2 is high and likely due to internal arcing. However, the trending of the gasses, the other gasses present, the sampling method, and analysis method used are all big factors here.

No one gas value for a single sample will tell the whole story.

What were the other levels?
How confident are you of the compentency of the person who drew the DGA sample?
What analysis method are you using?

 

[Microwatt]

Transformer oil sample

Transformer oil sample

I am dealing with a problem with a similar transformer. It's not blowing fuses, but its secondary voltage drops significantly as it gets loaded up. I was told I should get an oil analysis. Not being farmiliar with this I am looking for a little insight on it.

What should I test for? I've looked into it and it looks like there are many different things to test for, all with a different price. ASTM# D-1533,D-971,D-974,D-1500,D-1524,D-1298,D-877,D-1816,D-924,D-924,D-3612,D-3635,D-2668,D-1698,D-5837,D-1275

Who are some of the better companies to have perform the analysis?

Can the sample be taken from an energized transformer?

 

[11Haze29]

S.D. Meyers is a good sample company and one of the few that will take samples from an energized transformer. I don't think any oil sample test will tell you anything about excessive voltage drop. To check voltage drop, first look at the transformer impedance. From no load to full load, the % voltage drop will equal the % impedance on the nameplate. So if you have 5.75% impedance, and at no load you have 480VAC, at full load you would see a 5.75% reduction in voltage. You usually adjust transformer tap positions to give you the operating voltage you need at full load, and suffer a higher unloaded voltage condition.

 

[zog]

What should I test for? I've looked into it and it looks like there are many different things to test for, all with a different price. ASTM# D-1533,D-971,D-974,D-1500,D-1524,D-1298,D-877,D-1816,D-924,D-924,D-3612,D-3635,D-2668,D-1698,D-5837,D-1275

Who are some of the better companies to have perform the analysis?

Most power system testing companies are experienced in drawing these samples, just tell them you want a DGA analysis. Most of them will send it to the same labs anyways, most use Wiedman.

Sorry I don't know any in ND but there is Dynex in Minneapolis.

Can the sample be taken from an energized transformer?

Yes and maybe no. If done right it will not damage your transformer if drawn energized, if done wrong by someone not knowing what thye are doing you can suck air in and have a total failure, so get the right people here.

Depending on the configuration it may violate NFPA 70E to draw a sample live, some basic info about your transfomrer will determine if a S/D is required or not. Post a photo.

 

[dbuckley]

I am dealing with a problem with a similar transformer. It's not blowing fuses, but its secondary voltage drops significantly as it gets loaded up. I was told I should get an oil analysis.

I'm by no means an expert on this, but I don't see how oil "quality" is going to affect the regulation abilities of a transformer. There's no harm in getting the analysis done, however.

Is this transformer being operated within its nameplate ratings, and does the primary voltage hold up as the secondary voltage drops?

 

[mxslick]

I am dealing with a problem with a similar transformer. It's not blowing fuses, but its secondary voltage drops significantly as it gets loaded up. I was told I should get an oil analysis. Not being farmiliar with this I am looking for a little insight on it.

I'm by no means an expert on this, but I don't see how oil "quality" is going to affect the regulation abilities of a transformer. There's no harm in getting the analysis done, however.

Is this transformer being operated within its nameplate ratings, and does the primary voltage hold up as the secondary voltage drops?

If there are loose connections internally not only will the voltage sag under heavier loading, but the arcing WILL generate gases and byproducts that can be detected with a DGA.

It is one of the most valuable tools to determine the condition of a transformer.

Another trick one old-timer showed me to detect internal arcing in a transformer: use a portable AM radio tuned near the high end of the band, but not on a station, will have a very raspy buzz or even high-pitched static when brought near a transformer with internal arcing. Otherwise it will have a pronounced "60Hz" hum.

 

[Besoeker]

If there are loose connections internally not only will the voltage sag under heavier loading, but the arcing WILL generate gases and byproducts that can be detected with a DGA.
It is one of the most valuable tools to determine the condition of a transformer.

I agree. One of our customers may have a problem with a 2MVA unit. It's part of a fairly large VSD (c8,000 HP) we manufactured some years ago. There is a problem with insulation failures on one of 16 small signal isolation transformers that isolate the power circuit gate drives from the control electronics. It is the same one every time and has been replaced three times now. We have made quite a few drives of this type in operation and it is the only failure of this particular component we have experienced.
Hardly what you'd consider a random failure.
So suspicion has fallen on the 2MVA transformer. It's 11,000V to two 950V windings (Ddyn0) with a grounded interwinding screen between the HV and the LV. If the signal transformer insulation failures this most likely excessive voltage. One possibility is a breach of the integrity in the interwinding screen and possibly HV to LV PD.
Currently, we ware waiting for the DGA results from the second sample.

Another trick one old-timer showed me to detect internal arcing in a transformer: use a portable AM radio tuned near the high end of the band, but not on a station, will have a very raspy buzz or even high-pitched static when brought near a transformer with internal arcing. Otherwise it will have a pronounced "60Hz" hum.

A 60Hz hum would definitely indicate an unusual problem in this case...

 

[mxslick]

A 60Hz hum would definitely indicate an unusual problem in this case...

LOL I put it in quotes because it depends on who you talk to..I've always called it a 60Hz but it's really 120hz....I think. I posted last night after a long day on only three hours sleep. :grin:

 

[zog]

LOL I put it in quotes because it depends on who you talk to..I've always called it a 60Hz but it's really 120hz....I think. I posted last night after a long day on only three hours sleep. :grin:

Still 60Hz, Hertz is a cycle per second, + and - pulse. But 120 "beats" per second if you talking about a tone, that is nearly a perfect B flat.

 

[wirenut1980]

Well there are a few conflicting standards on what is acceptable or not but 14PPM of C2H2 is high and likely due to internal arcing. However, the trending of the gasses, the other gasses present, the sampling method, and analysis method used are all big factors here.

No one gas value for a single sample will tell the whole story.

What were the other levels?
How confident are you of the compentency of the person who drew the DGA sample?
What analysis method are you using?

I feel like the guys taking the sample are pretty competent. How does one screw up taking an oil sample anyways? Maybe they are not competent for that task...

We use Weidman for the DGA analysis. I don't have an actual copy of the report from 2002, just an old email describing the acetalyne level.

The report for this recent sample just says "less than 1 PPM" and does not list any specific gases.

The transformer was not changed out according to the utility local district. The transformer nameplate says 1997 on it.

I am not sure on what you mean by analysis method. I am not really competent to decipher oil sample analysis. What standards are out there that could help me develop a method?

Thanks Zog!

 

[Besoeker]

LOL I put it in quotes because it depends on who you talk to..I've always called it a 60Hz but it's really 120hz....I think. I posted last night after a long day on only three hours sleep. :grin:

In this case, either would be surprising...

 

[zog]

I feel like the guys taking the sample are pretty competent. How does one screw up taking an oil sample anyways? Maybe they are not competent for that task...

You realize it is not a matter of filling a jar with oil right? You need to draw a DGA sample with a syringe, using all the correct fitting and flushing procedures prior to drawing the sample. Any air in the sample will throw off your DGA results. [/quote]

We use Weidman for the DGA analysis. I don't have an actual copy of the report from 2002, just an old email describing the acetalyne level.

You need to look at the entire report, no one gas at any level can tell you what is going on. You need to look at the concentrations of all the gasses present.

The report for this recent sample just says "less than 1 PPM" and does not list any specific gases.

It does not sound like a DGA, just the common oil analysis. No DGA report from Wiedman would just say that.

I am not sure on what you mean by analysis method. I am not really competent to decipher oil sample analysis. What standards are out there that could help me develop a method?

Unless you are some master chemist you won't be developing a method, there are several used already. I can't tell you the best one because I only know a fraction of the story.

Most common is the interpretative or key gas method. This looks at concentrations of 9 key gasses (H2, O2, N2, CO2, CO4, CH4, C2H2, C2H4, C2H6) to determine the condition and history of the events the transformer has seen.

Other methods include the Rogers Ratio, Dornenburg ratio, etc

The standards to look at are IEEE/ANSI C57.104

 

[Microwatt]

Most power system testing companies are experienced in drawing these samples, just tell them you want a DGA analysis. Most of them will send it to the same labs anyways, most use Wiedman.

Sorry I don't know any in ND but there is Dynex in Minneapolis.

Yes and maybe no. If done right it will not damage your transformer if drawn energized, if done wrong by someone not knowing what thye are doing you can suck air in and have a total failure, so get the right people here.

Depending on the configuration it may violate NFPA 70E to draw a sample live, some basic info about your transfomrer will determine if a S/D is required or not. Post a photo.

No photo but here is some literature. http://www.cooperpower.com/Library/pdf/21012.pdf

 

[zog]

No photo but here is some literature. http://www.cooperpower.com/Library/pdf/21012.pdf

We stopped doing these padmount oil samples energized a few years ago. Can't justify doing it energized per 70E and it really is dangerous. Usually you have to move those cables around to get to the sample port and I have seen too many close calls on these.

There is a retrofit kit that moves the sample valve outside the door, most companies do that and put a lock on the valve. Then you an safely sample them energized.

 

[wirenut1980]

You realize it is not a matter of filling a jar with oil right? You need to draw a DGA sample with a syringe, using all the correct fitting and flushing procedures prior to drawing the sample. Any air in the sample will throw off your DGA results.

You need to look at the entire report, no one gas at any level can tell you what is going on. You need to look at the concentrations of all the gasses present.

It does not sound like a DGA, just the common oil analysis. No DGA report from Wiedman would just say that.

Unless you are some master chemist you won't be developing a method, there are several used already. I can't tell you the best one because I only know a fraction of the story.

Most common is the interpretative or key gas method. This looks at concentrations of 9 key gasses (H2, O2, N2, CO2, CO4, CH4, C2H2, C2H4, C2H6) to determine the condition and history of the events the transformer has seen.

Other methods include the Rogers Ratio, Dornenburg ratio, etc

The standards to look at are IEEE/ANSI C57.104

Thanks Zog. At least now I know what I don't know.

I think you are right that they did not do the DGA, just an oil analysis. I probably did not specify. I'll have to check and see how the oil was taken from the transformer.

If I were a master chemist, I would probably set my sights on turning lead into gold.:grin:

 

[Besoeker]

Thanks Zog. At least now I know what I don't know.

The beginning of understanding is the acknowledgment of ignorance.
Not always easy. In my twenties I considered it a weakness that I was unwilling to to admit.
From my thirties on I saw it as a means to learn more.

If I were a master chemist, I would probably set my sights on turning lead into gold.:grin:

That would make you an alchemist.

 

[jcole]

What does DGA stand for? What is the difference between an oil analysis and a DGA analysis?

We just had our transformers done at our plant by SD Myers. We had the same problem that the original poster had (high values one time, low the next). Nobody seemed to care. So we didnt care either.

 

[Besoeker]

What does DGA stand for?

Dissolved gas analysis.

 

[zog]

What does DGA stand for? What is the difference between an oil analysis and a DGA analysis?

We just had our transformers done at our plant by SD Myers. We had the same problem that the original poster had (high values one time, low the next). Nobody seemed to care. So we didnt care either.

A DGA samples the gasses in the transformer, the 9 key gasses will tell you if you have arcing, celluose breaker down, over heating, etc.... It is like a blood sample at the doc. Normal oil sample just checks the dielectric strength of the oil and if there are any contaminates present. Alway do a DGA. There is no way you had high C2H4 and then it was gone, someone messed up a sample.