DGA values of transformer oil 5

[edison123]

Is there any standard/reference that lists the dissolved gas values in transformer oil ?

We got a DGA test result as below in a 11 KV/150 V, 3 MVA rectifier transformer.

H2 - 2 ppm

Methane - 155 ppm

Ethane - 683 ppm

Ethylene - 1101 ppm

Acetylene - 1 ppm

CO2 - 480 ppm

Water content - 50 ppm

Are these values worrisome ?

 

[DanDel]

One set of limits which we use are:

Water - max 35ppm for oil; 100ppm for silicone

Hydrogen - 100ppm
Methane - 120ppm
Ethane - 35ppm
Ethylene - 30ppm
Acetyelene - 15ppm (though any acetylene indicates arcing under oil - not a good thing)
CO - 350ppm
CO2 - no limit
O2 - 50000ppm
Nitrogen - no limit

From your results, I would imagine you have some overheating issues, and possibly some water intrusion.

 

[edison123]

Thank you DanDel.

Yes, there has been some overheating issues with the two LV neutral terminals, which carry DC return current of over 6000 amps. I am planning on modifying these terminals from round to flat.

Is there any standard that specifies those DGA values ?

 

[DanDel]

ASTM D1433-77 for the water, and IEEE Insulating Fluids Committee, Project 637, April 1980, for the DGA levels.

These standards may have been updated; I haven't checked recently.

 

[Kiribanda]

edison123,

Ethlyne and water content seems to be on the high side.

Please refer IEEE C57.104-IEEE Guide for the Interpretation of Gases in Oil Immersed Transformers.

You can come to a conclusion.

 

[edison123]

Thanks Kiribinda and DanDel. I will try to get those standards.

 

[ScottyUK]

edison,

You might wish to investigate the Rogers Ratio criteria which, as you probably can guess, look at the ratio between different gasses rather than the absolute values. I think there is a guideline that the total hydrocarbons should be above 500ppm when applying the Rogers tests. I'm pretty sure there is some discussion in the J&P Transformer Book and this document is worth a read:

http://www.usbr.gov/power/data/fist/fist3_30/fist3_30.pdf

----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[mgtrp]

CIGRE published an excellent guide to interpreting DGA test results.

 

[edison123]

Thanks scotty for that excellent usbr link. Quite informative. Interestingly, that FIST manual says DGA is unreliable if the trafo was de-energised and has cooled (at the time of sample extraction) or if it had less than 2 weeks of continuous service after oil processing. I wonder why.

mgtrp - any link to that CIGRE guide ? Thanks.

 

[ScottyUK]

Hi edison,

The likely reasons for the restrictions in the FIST manual are: Oil in a transformer which is out of service and cool doesn't have oil circulation due to convection, resulting in a sample which isn't representative of true conditions. A transformer which is newly returned to service needs time for the dissolved gases to reach equilibrium, and as before taking a sample too early will result in a sample which isn't representative of true conditions.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[edison123]

Scotty. But, dissolved gas is dissolved gas, right ? I mean do these gas elements disappear when the trafo oil is cold ? For that matter, do they get removed when the oil is filtered/treated ? I am trying to understand the chemistry here.

 

[ScottyUK]

I imagine the solubility changes with temperature, but the main reason as far as I understand is that the sample must be taken from 'live' oil which is actively circulating to get a good sample and in a convection cooled transformer the heat from the core and windings is needed to produce that circulation. As far as processing goes it is done with heat and under a high vacuum to remove particulates and dissolved gases. I know some processors just filter the oil but in my opinion that is a waste of time and effort.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Ceast]

Here's one.

 

[prc]

edison, you have found out the cause of heating up.Then why worry with DGA values.DGA is to detect the cause of problem.You have already found it out!Yes the results show over heating of oil with out any arcing or involvement of cellulose.Latest standard on DGA analysis is IEC 60599 with the amendment of 2007,incorporating the CIGRE study results.
The typical allowable values as per IEC is acetylene-2-20ppm, H2- 50-150 ppm, Methane -30-130 Ethylene - 60-280 ppm, Ethane 20-90 ppm, CO-400-600ppm,CO2- 3800-14000 ppm

When oil is filtered,all dissoved gases are removed in the vaccum chamber of the filter.It will take some time for the gaeses to reach the saturation levels.

 

[edison123]

Thanks Ceast for that nomograph.

Thanks prc for those numbers and confirming that the DG's are removed in the vacuum chamber. Yes, we did find the problem of loose connections and fixed it. But the nos. I posted were after fixing the problem (at least as per my client) and hence my query.

I think the oil requires more filtration. I will be visiting the site this week-end and will post here my findings.

 

[alehman]

One oil test is not as meaningful as trends. After you think the problem is corrected the most important thing will be to monitor for changes in DG values over time. Get the oil cleaned to a tolerable state as soon as possible and leave it alone for a while so you can monitor it.