Oil analysis results

[Gauss2k]

I just received the results of the oil analysis we made on a 2.5 MVA transformer. I'm a little puzzled by the results and I would like to have your opinion.

We have the results from the last 4 years.
The DGA show stable gasses (low concentrations) and no problem have been observed on that transformer.

But look at the oil properties:
(values for 2003 2004 2005 2006)
Dielectric breakdown (kV) :
51.5 46.9 35.2 18.6
Neutralization number (mg KOH/g):
0.028 0.036 0.044 0.052
Color:
1.5 1.5 1.5 1.5
Interfacial Tension (Dynes/cm):
26.2 27.9 25.3 26.9
Water (ppm):
7 3 8 14

Oxidation Inhibitor (% DBPC):
(2005 2006)
0.18 0.16

There is more than enough oxidation inhibitor and we observe what looks like oxidation results (Neut. N. increases, DB lowering, low IFT), but the DB is falling really fast! I think a treatment with fullers earth is necessary, but can't explain why this happened with enough oxidation inhibitor!

Anything else could explain those results (sample contamination, analysis errors?)

 

[Zogzog]

Let me start off that I am not an expert in this area but you asked for opinions, so here is mine.

The increase in water contant would seem to be related to dielectric breakdown decreasing. The breakdown of the oils insulating properties would cause sludge deposits and a increase in your neutralization number.

Instead of putting a bandaid on this you need to find the source of the problem, where is the water coming from? Any leaks? What was the P/V gauge for each year? How often (if at all) have you had to re-pressurize the system (or refill the N2 bottle if this has that type of system)?

I doubt sample contamition would cause these results, water content yes, but the rest of the analysis seems to agree with the water in the transformer, not just the sample. If the samples were tested by the same lab I doubt an error occured, these results look real to me.

Again, just my opionon, but I do review alot of samples.

Scott Peterson
Training Manager
Power Plus Engineering

http://www.epowerplus.com

 

[Gauss2k]

Hi Scott,

Thanks for the reply.

You think that the water content can be significant in this case? It has raised by 11 ppm to reach 15 ppm in 2 years, which I believe is not that much. I don't think a 15 ppm water content could create such a drastic change in the oil properties.. I have seen several transformers with about 25-30 ppm of water without big variations in DB

I would need to check our files to get the pressure readings during each of those samplings.

 

[electricpete]

How about some conductive contaminant (maybe carbon particles from tracking somewhere)?

The change in dielectric breakdown as you point out is severe.

The change in the other parameters don't seem that severe from my memory. I thought 0.1 max was an upper limit for nuetralization number - you're still below that.

You might want to to a retest to confirm the values (especially dielectric). While you're at it maybe there are some more special tests you can do to narrow down contaminants. Talk to your oil lab about what they can do and what they recommend. Possibly mass spectrometry by ICP method or RDE method. Possibly FTIR. A sample of new oil is often helpful for comparision on these types of analyses.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Zogzog]

Recent changes in sources of world supplies of crude oils refined to produce insulating oils have allowed some oils to be used in transformers that contain concentrated amounts of corrosive sulfurs. These oils have been in the supply system since the year 2000.

The problem with corrosive sulfurs in transformers (and line reactors) is that the sulfurs react with the copper (especially in oxygen deficient systems like a sealed tank transformer) causing the occurrence of copper corrosion and copper sulfide (Cu2S) deposition in paper insulation which may lead to insulation failures (Usually across the turns of the HV winding). This process is accelerated by excessive heat found in warm regions of the world and on overloaded units.

Unfortunately, most failures usually occur without any warning signs such as excessive gassing or Partial Discharge, even up to the day before the failure occurred.

Currently, the best means for detecting corrosive-sulfur problems has been by internal investigation. Testing the oil can determine if the apparatus could develop a problem, but it is the internal inspection that has revealed the evidence of copper-sulfide formation.

The oil analysis used to detect the presence corrosive sulfur compounds is ASTM D1275. On March 1st 2006 an alternative, more in depth method of analysis was approved; it is referred to as method “B”.

 

[unclebob]

Make sure the sampling is done correctly, i.e., rinse your
vessel and drain some oil (half liter) before taking the sample. Do the dielectric test yourself, just to be sure.
( I know you have a dielectric tester...)

 

[Gauss2k]

The person who took the sample tells me it was done correctly, but we'll double check with a new sample. Good idea to confirm with our dielectric tester

I talked to Morgan Schaffer about those results and they told me it was most likely a water contamination in the plastic jar (even if the water content was low in the analysis, the water ppm is taken in the syringe and not in the jar). But the weather was sunny and dry on that day...

Anyway, we don't have much choice but to confirm with a new sample!

Thanks for your answers everyone!

 

[mar888]

I suggest rto perform oil particle analysis folowed by ferography or/and ICP it seemes a contamination or non compatible materials inside the trnasformer to oil.
If you have a similar transformer it will help to examine same test.

 

[prc]

A BDV of 18 kV at 2.5 mm gap is pretty bad.IFT has not deteriorated, indicating no need for reclamation using fullers earth.Take oil sample when transformer is fully loaded/hot and check moisture content.When transformer is hot,oil will come from paper to oil and will give actual water content. Any way immediate step is to take a shut down,do a hot oil filteration using a good oil filter machine having vacuum chamber. OIl BDV comes down with high water content and /or particle contents.

Zogzog,latest studies show even testing as per Method B of ASTM is also not adequate to rule out corrosive sulphur.A new test method has been finalised by CIGRE study committee and the same is now under consideration of IEC working Group.

 

[Zogzog]

prc, can you send me some details or tell me where to find them?

 

[electricpete]

I've heard just a little bit about sulfur lately but I'm interested to hear more about the concern that sulfur poses as well.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[prc]

Zogzog,please give your e-mail id so that I can send you.If you are a CIGRE member you can try at

http://www.cigre.org

electricpeten copper corrosion please visit doble site for some preliminery information.

 

[jnims]

Here is a presentation made an the Fall 2005 transformer committee meeting:

http://www.transformerscommittee.org/info/f05-presentations.htm

 

[dewone]

Maybe try a Fault metal test ASTM D-3635icp, and a Corrosive Sulfur test ASTM D-1275. Agree with the above retest DB on your own just to compare. Is this a conservitor type trans or sealed with a top air space.