Does SF6 degrade with number of CB operations

[Ephraim]

I have jus had an 11kV SF6 Circuit Breaker flash on our plant in a manner that suggests that either there was loss of SF6 the quality of th gas had deteriorated therefore loosing some of its critical properties. Because the tests showed that there was no loss of SF6 I would like to assume that there was deterioration. Is this a correct assumption and if so what are the facts relating to SF6 degradation with Circuit Breaker operations?

 

[DanDel]

What tests did you perform that showed there was no loss of SF6? A gas sample can be taken for analysis, similar to an oil sample for an oil-insulated switch, but the only method I am aware of to determine SF6 loss is by verifying the pressure in the tank.

 

[Rocketir]

SF6 gas itself will not need replacement because it doesn’t get used up, but every time there is an arc a toxic/corrosive residue is left which after a while will build up.
However water and air molecules in the gas can combine with the contamination and eventually lead to a dielectric breakdown.


Jasjit Khangura P.E.
Electric Distribution Engineer

 

[busbar]

You may try posing the question to an operation like that at

http://www.actiinc.com.

They conduct a number of procedures for SF6 evaluation in electrical environments, including decomposition and dielectric tests.

 

[Ephraim]

Thanks to you guys, DanDel, Rocketir and busbar.

DanDel - We confirmed that the tank had SF6 by checking the tank gas pressure before removing the gas from the tank to enable us gain access to the CB contacts.

busbar - I have sent a querry to ACTI. I'm yet to hear from them. Thanks.

The CB manufacturer's representative on site has indicated to us that the CB may have done its expected number of operations. They are yet to confirm the expected number of operations and what deterioration is expected.

 

[DanDel]

Ephraim, I'm going to guess that you released all the SF6 to the atmosphere and did not collect a sample for analysis. The procedure to take a gas sample is not easy, but the analysis may have given you some answers.

 

[hy]

The fault seams to be like leakage problems, or better say SF6 contamination problem.

If there has been leakage, there could enter moisture inside, many CB manufacturers of HV CBs use dew-point meters for evaluation of gas contamination (with moisture).

 

[Ephraim]

DanDel,

You guessed right. Our minds were in some kind of auto pilot: "Something has mechanically gone wrong with this CB, lets open it and check what we shall find."

We took the gas out and opened the tank only to find a flash and a general layer of carbon. Mechanically, the internals, including the contacts, were okay apart from the carbon layer.

The switchgear manufacturer has since taken the CB away for rehabilitation so I cannot do any more tests. I will just have to wait for their report. Any thoughts and pointers will still help a lot

Thanks

 

[DanDel]

Ephraim, what was the original 'flash' that made you think you had a loss of SF6? Did something flashover externally at the C/B or somewhere else? Or maybe you had a fault downstream that the SF6 C/B didn't clear as you would have expected?

 

[Ephraim]

DanDel, The CB did not flash externally and there was no fault down stream. The CB feeds a 1250HP mill motor. When the CB failed, the incomer tripped on earth fault, as this was seen as a busbar fault. Insulation tests were done on incomers, busbar and feeders, that is how this particular breaker was found to be down to earth even though there were no visiual external indication. Power was restored and after further checks on the mill motor and its cable a spare CB was used to start up the mill and it has been running problem free. Partly that is why we rushed to open the CB, expecting to find mechanically damaged parts touching the tank or something like that.

 

[DanDel]

Ephraim, I'd be interested to hear what the manufacturer says about why the C/B failed and why there was carbon buildup inside.
Who was the manufacturer and what is the model of the C/B?(if you don't mind sharing that info)

 

[Ephraim]

DanDel, I doubt if the manufacturer would want to be discussed so openly, but I promise, I will let you know what the manufacturer will come up with

 

[DanDel]

Thank you.

 

[jbartos]

Suggestion: The flash could happen because of:
1. SF6 insulating deterioration while the voltage was within the circuit breaker design basis
2. The voltage experienced a spike or surge that caused the flash.