Significance of CO2 when CO/CO2 ratio is low 2

[electricpete]

Let’s say we have a conservator type GSU with steady state levels on the order of
500ppm CO
18,000ppm CO2

The CO exceeds the IEEE threshhold by a small amount.
The CO2 exceeds the IEEE threshhold by a large amount.

My understanding is the CO and CO2 are both byproducts of thermal aging of cellusoe
insulation, but that ratio of CO/CO2 increases as the aging occurs at higher temperature.

CO/CO2<(1/7) is a good thing per IEEE standard.

If I looked at CO only, I would say we are only slighly above the limit.
If I look at CO2, it says way above the limit.
However the low CO/CO2 suggests lower temperature

QUESTION: Does adding CO2 to the picture cause any higher of concern compared to looking at CO only?

[yes, I realize there are many aspects of evaluating aging such as complete oil history and.. furanic compounds... the question is about the significance of CO2 in the above case... if we add it into the picture does it make the picture any worse than if CO2 was not monitored]


=====================================
(2B)+(2B)' ?

 

[Jebb]

You have to look at the big picture (i.e. all the gases) as a whole and the complete historical trend from the sampling data. Singling out one gas doesn’t add much value. It looks to me that you might have an overall cooling issue or general overheating. Has this been increasing over a long time? Is there some reason why you don't or can't monitor the CO2?

 

[ScottyUK]

My understanding is that GSU transformers in baseload stations are expected to have higher CO and CO2 levels than most system transformers because GSUs typically run fairly warm 24/7/365. What is the DGA trend doing? A high value in itself may just be an indication of an aging transformer which is still working hard. A trend showing the full DGA results would be a far better tool than spot values.

Are you familiar with the Rogers Ratios for hydrocarbons? How do they look? What is the furan trend doing?

 

[electricpete]

Jebb - Thanks for your comments. You'll see my original question was very narrowly focused on the difference between evaluation based on CO and by including both CO and CO2 (not asking you to analyse a transformer for me).

Let me enlarge my question to say I am trying to understand the significance we should attach to the two parameters CO and CO2, individually and in combination.

I think the interpretation of CO2 may be a source of some disagreement within the industry.

IEEE includes a limit on CO2... our insurance does not (they only have a limit on CO).

My quote above was based on IEEE which seems to suggests CO2/CO > 7 is good. (I added the comments about temperature but may have been offbase...I'll see if I can track down a refernce.)
But I get a different slant from Bureau of Reclamation CO2/CO>3 is good. CO2/CO<7 is good. Outside these bounds a problem

Let me list some quotes:

IEEE Std C57.104-2008]
The ratio of CO2/CO is sometimes used as an indicator of the thermal decomposition of cellulose. This ratio is normally more than seven. For the CO2/CO ratio, the respective values of CO2 and CO should exceed 5000 μL/L (ppm) and 500 μL/L (ppm) in order to improve the certainty factor, i.e., ratios are sensitive to minimum values. As the magnitude of CO increases, the ratio of CO2/CO decreases. This may indicate an abnormality that is degrading cellulosic insulation.

Formation of CO2 and CO from the degradation of oil impregnated paper increases rapidly with temperature. CO2 /CO ratios less than three are generally considered an indication of probable paper involvement in an electrical fault (arcing or sparking), along with some carbonization of paper. Normal CO2 /CO ratios are typically around seven. Ratios above 10 generally indicate a thermal fault with the involvement of cellulose. This is only true if the CO2 came from within the transformer (no leaks), and these ratios are only meaningful if there is a significant amount of both gases. Caution must be employed because oil degradation also produces these gases, and CO2 can also be dissolved in the oil from atmospheric leaks. The oil sample can also pick up CO 2 and O2 if it is exposed to air during sampling or handling at the lab. If a fault is suspected, look
carefully to see if CO is increasing. If CO is increasing around 70 ppm or more per month (generation limit from IEC 60599), there is probably a fault. It is a good idea to subtract the amount of CO and CO2
shown before the increase in CO and CO2 began, so that only gases caused by the present fault are used in the ratio.
This will eliminate CO and CO2 generated by normal aging and other sources. When excessive cellulose degradation is suspected (CO2 /CO ratios less than 3, or greater than 10), it may be advisable to ask for a furan analysis with the next DGA. This will give an indication of useful life left in the paper insulation [12].

=====================================
(2B)+(2B)' ?

 

[electricpete]

Hi Scotty – I guess we were typing at same time. See clarifications above. Sometimes the motivation for the question is necessarily vague.

I did review Duval’s “Review of Faults Detectable by Gas in Oil Analysis in Transformers”. It gives a number of laboratory studies and case studies which seem to support the conclusion that the ratio CO2/CO produced goes down as the temperature of the thermal aging reaction increases.


=====================================
(2B)+(2B)' ?

 

[iwright]

I've got a poster from Siemens that shows various standards and ratios and it gives the following two references for CO2/CO ratios
IEEE C57.104:
"CO2/CO Ratio of <3 Excessive
CO2/CO Ratio of >7 - <10 Normal
CO2/CO Ratio of >10 Excessive
*These ratios are only valid when levels exceed minimums: CO>500ppm and CO2>5,000pmm"

CIGRE SC15 "New Guidelines for Interpretation of Dissolved Gas Analysis in Oil-Filled Transformers, (Electra NO.186 October 1999):
"Key Gas Ratio #4, CO2/CO, >10 Indicates overheating of Cellulose"

In response to your question "Does adding CO2 to the picture cause any higher of concern compared to looking at CO only?" No the CO2 doesn't add much to the picture. The elevated CO already tells you that there is overheated cellulose, what would be of more use is the H2 level. If the H2 is elevated you have a thermal fault, if the H2 is not elevated you just have cellulose aging. With just the CO and CO2 you don't know if it is excessive aging or a thermal fault.

 

[electricpete]

Thanks iwright, good info.

I've got a poster from Siemens that shows various standards and ratios and it gives the following two references for CO2/CO ratios
IEEE C57.104:
"CO2/CO Ratio of <3 Excessive
CO2/CO Ratio of >7 - <10 Normal
CO2/CO Ratio of >10 Excessive

Interesting. That seems to mirror the Bureau of Reclamation document (available on-line free). I don't se that in IEEE Std C57.104-2008.... all I see is the part I quoted above which has a different spin. I wonder if it was in an old version or I'm just missing it.

By the way, I noticed in front of the guide Steven Alexanderson. Could that be our own stevenal?



=====================================
(2B)+(2B)' ?

 

[stevenal]

I don't write standards, but I vote on a few. (good way to see advance copies) Guess they include ballot members.

 

[iwright]

Electricpete, I looked a little closer at the fine print and the part reference to >10 is actually from IEEE PC57.104 Draft 11D, a different part of the poster references IEEE C57.104.1991. I guess the >10 ratio got voted down in the 2008 version. I'll have to read the fine print a little closer in the future.

 

[prc]

Pete, C57.104 -2008 is the latest IEEE standard on the subject.IEC 60599-2007 gives 90%typical gas concentration values as 3000/14000 ppm for CO/CO2 and typical annual increase as 260-1060/1700-10,000 ppm for Power Transformees. No typical values are given except that less than ratio 3 is an indication of cellulose ageing.Standard also mentions-
"CO and CO2 formation increases not only with temperature but also with the oxygen content of oil and the
moisture content of paper." "Air-breathing equipment, for example, saturated with approximately 10 % of dissolved air, may
contain up to 300 μl/l of CO2 coming from the air" "If necessary subtract last values from present ones before calculating ratios, particularly in the case of CO, CO2"

There are also cases when abnormal values can be due to other reasons. In India in some 420 kV reactors co/co2 values of 1247/17,705 and 784/18,459 ppm were reported with in a year.Other thermal gases were insignificant.Finally it was found out that these values were from synthetic rubber(with inadequate vulcanisation) and resin impregnated wood used in reactor. Lab tests conducted on these materials(ASTM 3455-02 80C at 164 hrs)with oil showed high content of CO/Co2 under DGA.When these materials were replaced with good quality material CO/co2 became normal.

In a traction transformer where oil content is low, showed Co/CO2 of 292/1237 ppm with thermal gases Ch4 = 509,C2H2 =8,C2H4=799,C2H6=150 ppm after a few months of service.Inside, charring of wood was noticed at one place where a hot metal was touching wood.Almost similar results got in another unit where there was no involvment of any cellulose or wood but only a heated joint in a lead connection.

Hence it seems CO2/Co may be misleading during analysis.It can be affected by stray gases,passivators added in oil,certain paints ,gaskets,wood etc.