Rate of Dissolved Gas Increase

[czping]

Hi all,

Just need to verify some doubts..

IEC 60599 Guide to interpretation of Dissolved Gas Analysis
Page 49.

I have 2 readings of Ethane taken in Nov 2003 & Dec 2003.
PPM is as such:
Nov 2003 - 52ppm
Dec 2003 - 106ppm
Oil Vol - 7950 cubic meters
number of days interval - 15

Using the formula in page 49:
rate = 28.62 ml per day..
which exceeds the acceptable level.
m i right?

 

[jbartos]

Suggestion: Please, provide more information regarding the gas and formula. Visit

http://www.usbr.gov/power/data/fist/fist3_30/3_30_4_4.htm

etc. to see what is actually involved.
Also, IEC standard is:
Guide to the interpretation of dissolved and free gases analysis, IEC 60599, Second edition, 1999-03.

 

[xabproject]

Dear czping,
We too had an alarming situation in 3 nos 140MVA transformers sometime back, very similar to what you have explained, but with other gases - Hydrogen and methane. We had carried out samplings on the 15 day intervals. On detailed investigation, we found that
1) One of the samples had been taken from Tank bottom (not a recommended place for DGA sample!)
2) the loads on these Transformers were so small that a good thermo-siphon action/proper distribution & mixing of gases to uniformly occupy the available volume of the oil could not have taken place.
The samples, however carefully drawn, would yield incongruent results!
3) the period shall be not less than 15 days if load levels are good ( not on the "15 days period" by clock)

Does any of this is similar to what is prevailing at your end? We have now decided not to sample earlier than three (3) months if the load levels continue to be the same.
Though a lot of literature is available on the web, you may have to arrive at certain benchmarks all by yourselves, in consultation with the Manufacturer, as each installation offers certain unique conditions! It is often seen that the Vendor of the DG Analyser equipment will normally play insignificant role in offering useful guidelines/solutions!

best regards,

 

[electricpete]

Volume Rate = Concentration Rate * Volume
= (50*PPM/15/DAY)* 7950*M^3
* (1E-6/PPM)*(1E6*mL/M^3) (unit conversions)
= 26500 mL/DAY

Your answer is factor of 1000 too low.

But 7650M^3 is an astrnomical oil volume. Roughly a cube 20 meters on each side. Much larger than even a Generator Stepup Transformer.

If you meant 7650 liters, then your answer is in the ballpark.

 

[rconnett]

Based upon my experince, sudden rates of change on dissolved gases in oil are a very serious concern. My recollection is that ethane is a product of internal overheating as opposed to corona (hydrogen) or arcing (acetylene). I have found the gas tests to be a very sensitive test of developing problems.

The comment about oil circulation is a good one. However, sample points are usually located at the bottom of the tank.

Is this a new transformer or one whiich has been in service for a period of time?

 

[chiefwvfc]

Also, time of day can make a difference in the sample results, we have seen that samples late in the afternoon after the sun has heated the transformer will result in higher gas levels then those in the morning.

This is also due to the fact that the samples are taken from the bottom of the transformer where circulation is poor.

Regardless of the reason, you definitely have an indication that there may be PD in the transformer, and additional testing of the transformer is required during this stage to ensure that a catastrophic failure will not occur from the rising gas levels.

You did not mention the size of this transformer or if it has an LTC in the tank. LTC switching can cause sudden changes in oil condition as well.

 

[czping]

Hi, thanks for all the replies!

jbartos: Sorry for being lazy. I should have the full title instead of the shortcut.

xabproject: We carry out DGA yearly but due to the increase of Ethane from 8ppm in yr 2002 to 52ppm in yr 2003, I carried out another one 15 days later to confirm the results. Just like to find out why is tank bottom not recommended? Bad circulation?

electripete: Yup, you are right. It is in litres. Thanks!

rconnett, chiefwvfc: This is a 23/6.6kV, 15MVA, ONAN Unit transformer in our power station, no OLTC. Oil sample is taken from Main Tank (Bottom). It has been in service for about 3 years now, load are constant.

I have asked 2 separate contractors to do the 2nd sampling to verify the results.
1st sample done by contractor A. Result is 52ppm in Nov 2003.
2nd sample done by contractor A again 15 days later. Result is 44ppm but another contractor, contractor B also collect on the same day as contractor A for the 2nd sample and result is 106.4ppm.

I am puzzled by the difference in the result. Contractor B has mentioned that different DGA anaylser might produce diff results.

My questions:-
1. Which results should I take in?
2. Is there a cause of concern that the ppm increase from 8ppm to 52ppm over a year?
3. Will there be a time whereby this ppm will be constant over the years, but still above the acceptable IEC level. The existence of this gas might be due to the material used for the transformers, say, insulating paper etc.

Sorry for the long posting.

 

[czping]

Also, we are using : Guide to the interpretation of dissolved and free gases analysis, IEC 60599, Second edition, 1999-03. as our reference.

 

[rconnett]

FYI, if the sampling procedures were not consistant between the contractors(for example if one contractor delayed longer than the other contractor in running the analysis in the lab) sometimes certain gases are more volatile than others - they can go back out of solution & thus give differing results.

 

[xabproject]

Dear czping,
Thanks for your additional inputs.

a) reply to your question:
Some maintenance practices call for comparison of DGA results to that of gas accumulated in the Buchholz. So a tank top sample is considered to be more representative. You see how 'poor circulation' is correlated!

b) I have reproduced an extract of available literature with me. Hope this will be of some use. Please note the "longer than..."

...................
Dissolved Gas Analysis on
transformer oils from the main
tank of a transformer
...................

Sampling has to be done in accordance with
IEC 60567 or ASTM 3613
Analysis has to be performed in accordance
with IEC 60567/599(??) or ASTM 3612

DGA should be performed as frequently as
possible, but at least once a year as long as
the values remain below the limits of Table 1.
When values exceed the limits of Table 1, a
more frequent analysis is necessary (contact
Manufacturer).

Table 1 Limits for normal values
THRESHOLD VALUES
Rates of Concentrations
production(*,**) ppm
ml/day
GAS
H2 5 150
CH 5 110
CO2 200 13000
CO 50 900
C2H6 2 90
C2H 2 280
C2H2 0.1 50 (***)


#) If the measurement (one of the gasses)
exceeds one of the threshold values
(concentrations OR rates of production) in
Table 1 (and measurement of the second
sample confirms these results), a new
sample should be taken and analyzed.
##) If the values of the new sample still
exceed the threshold values in table 20,
Manufacturer should be contacted as soon as
possible.
Remarks:
1)If C2H2/H2 ratio is higher than 2, a
contamination from the OL TC
compartment is probable. If this is the
case, DGA of the main tank can be
unreliable.
2)If CO2/CO ratio is lower than 3, paper
deterioration is possible. This has to be
checked with a FAL analysis (specific
analysis for paper deterioration.


*) Rates of production ml/day = (dc x V)/(M x 1000)

where,
dc = difference in concentration over the
observed period in ppm
V = volume of the transformer litres
M = expired period between measurements in days

**) To have a reliable rate, intervals should be
longer than 2 weeks

***)If a communicating OL TC conservator is in
use, this value will be 270 ppm



MOST IMPORTANT:
1) Were you associated with all the earlier measurements on this particular transformer, to the same capacity/extent as in the recent ones?

2) Have attempted to analyse the Buchholz sample and compare with your results?

best regards,

 

[chiefwvfc]

It is not too unusual to see differences in data between test labs and from different methods of taking samples. One of the DGA manufacturers gave a paper a few years ago showing the differences between labs with known samples. TJH2B also has some good papers on their web site about taking samples and analyzing them.

Either way you do show that the transformer is creating gas and it must be further analyzed to determine the source.

 

[jbartos]

Suggestion: The Contractor with very high values should provide references from the past where the readings were found acceptable. If no references are provided a different Contractor should be hired.