Catastrophic failure SF6 Circuit Breaker

I have a large collection of papers about SF6 circuit-breakers, but I am not clear what you need.

If the circuit-breaker has failed it may or may not have released arced SF6 into the substation. If Arced SF6 has been released there are toxic hazards associated with this, and I hope that you have had the site cleaned up professionally.

The circuit-breaker may have failed for a number of reasons :-

(a) mechanical failure of a component (b) Incorrect construction or maintenance (c) Insulation Failure (d) Loss of the interrupting medium (SF6 gas in this case) (e) The breaker was not suitable for the application. The last cause is probably the most unlikely.

John

Thanks for the reply John. We are in the process of decontaminating the sub station just now. We did not know about the possibility of toxins until made aware of this by the equipment manufacturer, so we have been in and out of the sub, many times before starting clean up. However, very little has been touched. I don't thing this is an installation problem, as the circuit breaker has been in operation for a few years. We are focusing on the following as possible causes: 1.. Synchronising problem 2.. Excessive number of operations, 2439 ops. 3.. Loss of SF6 gas. The extent of destruction of the CB has taken everyone by surprise, including manufacturers and insurance investigators, so one of the questions is why did the switchgear fail so destructively. The CB was used well within its rating. If anyone had been in the sub station at time, it may well have been a fatality. If I get time later today, I will post some photos. onto an image hosting website.

Thanks

| Thanks for the reply John. We are in the process of decontaminating the sub | station just now. We did not know about the possibility of toxins until made | aware of this by the equipment manufacturer, so we have been in and out of | the sub, many times before starting clean up. However, very little has been | touched.

Arced SF6 can produce disulfur decafluoride (S2F10), which is extremely toxic on the scale of phosgene. It had even been considered as a possible chemical warfare agent because, unlike phosgene, it produces no irritation and a lethal level can be absorbed before the victim becomes aware of it. Freezing point is -53C and boiling point is 30.1C. So it could be present in liquid form in the substation, if it was not all gassified by the arc that caused the failure.

| I don't thing this is an installation problem, as the circuit breaker has | been in operation for a few years. We are focusing on the following as | possible causes: | 1.. Synchronising problem | 2.. Excessive number of operations, 2439 ops. | 3.. Loss of SF6 gas. | The extent of destruction of the CB has taken everyone by surprise, | including manufacturers and insurance investigators, so one of the questions | is why did the switchgear fail so destructively. The CB was used well within | its rating.

Maybe the excessive operations had accumulated enough S2F10 products that would produce explosive gasification pressures under another fault.

Another possibility is a structural failure of the gas containment due to other problems like cracking from extremely cold temperatures or from temperature differences. Previous operations in cold temperatures could have created extreme thermal differences, producing cracks in the container or seals that will eventually either fail unless pressures that should otherwise not fail, or allowed the gas to escape or be contaminated.

Did only one phase fail? If so, the other phase breakers should be inspected. I'm sure the manufacturer would be suggesting this.

| If anyone had been in the sub station at time, it may well have been a | fatality. If I get time later today, I will post some photos. onto an image | hosting website.

I would be curious to see just how much explosive pressure was involved. Can you report the normal current and possible fault current at the time of failure?

That's a hell of a lot of ops! In our experience most failures were due to loss of gas pressure ( MG make) although to be fair we have had a miniscule amount go wrong out of several hundred and none were 'catastrophic' in the sense you mean. The early ones were prone to leaking gas and we had a program of using 'sniffers' to detect leaks before the drop in pressure became noticeable. They were worse than useless. We now use Drager tubes to test the gas every six months. SF6 is not poisonous per se but during the arc creates a white powdery substance that is extremely irritating to lung membranes. The latest SF6 units seem reasonably gas tight ...so far..

Jb

A few further questions :-

What technology did the circuit-breaker use ? Puffer ? Rotating Arc ? What was happening when it Failed ? Was it Opening ? Closing ? or just sat there doing nothing ?

How big is the CHP set that it is connecting ? The number of operations may or may not be excessive depending on what duty the circuit-breaker was designed to meet.

Anything which leads to an uncontrolled arc within the switching enclosure is almost inevitably very destructive since it will continue until some back up protection removes the supply and in certain conditions this can easily be 3 seconds. If you assume that the fault current was 10,000 A, and the arc voltage is 100 V., and that all 3 phases are involved, the energy dissipated in 3 seconds is 9 MJ.

John

This may well be the cause as the CB had exceeded recommended number of ops that requires a factory service. The recommended number of ops is 2000, then the switchgear requires a factory service. This is not stated in the O&M manual, but advised by word of mouth. Why it wasn't sent away is another problem.

Not sure yet on this as switchgear has still to be sent back to manufacturers.

This is a 1.2 MW CHP, generating at 415 Volts. It is then stepped up to 11 kV by a transformer. The failed circuit breaker is used for switching onto the bars when synchronisation with the mains has been acheived. Due to the position of the charging springs, it looks like the problem occured when the breaker energised. So at this point load shedding would not have kicked in so the current would have been quite low. However, the CHP service engineer just ahppened to be working on the CHP when all this happened, and his account was that the CHP started, synchronised and energised the circuit breaker. The CHP started generating up to about 50kW, then displayed -200 kW and the engine started back firing. This indicates to me that we were shoving power back into the alternator and driving it as a motor.

The fault current at the transformer on the 11kV side works out to about

8750 Amps.

Puffer I believe. It is a YSF6 circuit breaker.

I think it had just closed judging by the charging springs and position of spring mechanism.

It is about 1.27 MW at 415 Volts. Stepped up to 11kV by transformer.

According to the CHP service personnel and switchgear manufacturers, it is not uncommon to use circuit breakers as synchronising breakers on HV. The number of operations is excessive without a factory service. It is not stated in the O&M manuals, but by word of mouth, that the switchgear is recommended for factory service after 2000 ops.

Image links:

I can post larger pics if required.

TIA

First impression is a switching error, could it have closed whilst an earth was still applied? Although a servicable CB should have handled it without that degree of damage.

The only possible place to apply an earth is that CB, unlesss of course there is an earth fault.

Thanks

| This is a 1.2 MW CHP, generating at 415 Volts. It is then stepped up to 11 | kV by a transformer. The failed circuit breaker is used for switching onto | the bars when synchronisation with the mains has been acheived. Due to the | position of the charging springs, it looks like the problem occured when the | breaker energised. So at this point load shedding would not have kicked in | so the current would have been quite low. | However, the CHP service engineer just ahppened to be working on the CHP | when all this happened, and his account was that the CHP started, | synchronised and energised the circuit breaker. | The CHP started generating up to about 50kW, then displayed -200 kW and the | engine started back firing. This indicates to me that we were shoving power | back into the alternator and driving it as a motor.

Are there any power factor reading logs that might show any rapid changes in reactive power demands?

I see several of these a month across the US.I have many images and case study's I can share. The normal failure is not loss of gas but high contact resistance due to contaminates that are generated in the chamber that will increase the contact resistance thus creating heat that will damage springs and contacts.

These poles can be reconditioned and this event can be avoided. Typically these are only a few types and the same issue is seen over and over again. Let me know if you want to communicate further.

FL

You do not say where you are. The reference to 11 kV suggests an area of UK interest. Several technologies have been used with SF6 in the UK including puffer and rotating arc. Which type of circuit breaker was it that failed, and did it fail during an operation, or when it was just sat on the system ?

John

We are in primarily operating to the US but have an office in Milton Keynes.

The majority of the US SF6 38kV and below switchgear is European design all types of contact interruption design Alstom, ABB and MG..

We see failures in service regularly and with the age of this installed infrastructure the failure rates are accelerating. We have set up a service center to deal with SF6 breaker failures and see them weekly for repair or replacement.

The work we see is repair based on three major groups of root causes.

50% Low SF6 pressure alarms. 30% High Contact resistance generally found during planned maintenance. 20% Catastrophic failure.

Many of the low pressure alarms are on ABB poles that the owners inadvertently released the gas during maintenance.

I have attached an image of a typical failure we see often in the US. FL [IMG]