GIS Bus - Trip On Low-Low Gas Pressure? 1

[rcw retired EE]

Our client has a 350 kV class design Gas Insulated Substation (GIS) operating at 230 kV. It is a breaker-and-half arrangement with three diameters, two outgoing lines to the utility and three GSU transformer feeders to the generating plant.

The local utility added a requirement for immediately tripping the transmission lines via transfer trip if any GIS bus section on the line side of the breakers has a Low-Low SF6 Gas Density alarm (Stage 2 Alarm).

Is this typical to trip on low SF6 pressure at a manned facility?

All the sections have a Low Pressure Stage 1 alarm in the DCS that sounds in the continuously manned control room. The Stage 2 alarm tells the operator it is urgent.

Inside the plant the philosophy is to alarm only and give the operator a chance to correct the problem. (All breakers have dual interlocks to block operation on Low-Low SF6). If the bus or VT/CT enclosure is leaking gas the operator will get a low gas alarm and can take action to isolate the affected section for maintenance. Some of our thoughts are:

1) Most leaks are not large, unless the enclosure is damaged and then it has probably already flashed over.
2). Since the system is designed for 350 kV and operating at 230 kV, the insulating quality of SF6 gas at one atmosphere pressure is sufficient per the GIS vendor, assuming the SF6 is clean.
3) The enclosure cannot leak out to below one atmosphere.
4) The operator should have time after the first alarm to get the problem corrected or at least isolated.
5) The station building has large storage of SF6, spare parts, a transfer/filter cart and trained staff on duty to make a timely repair.
6) Tripping the transmission line sheds other customers and trips a lot of generation.
7). Low Gas Pressure systems are a source of nuisance trips.

What is your philosophy for GIS? Trip or alarm?

Of course, I would not be asking this question if we didn't have a momentary false actuation that tripped one line and should have dumped both of them. A concurrent timing aberration in a teleprotection system appears to have saved the other line and a lot of utility customers. We have confirmed that a real trip signal would have operated correctly.

 

[slavag]

Hi rswilson.
We never used SF6 low-low alrms for the trip.
Only for the BFP acceleration.
I think, sure, it's a very bad solution.
I told about substaion up to 500kV, double bus, one-and-half CB. For the manner plant and one-and-half CB arrangement...operation personal have enough time for solve the problem and usually first come low alarm, up to low low personal have a "lot" of time.
Best Regards.
Slava

 

[marks1080]

You can consider tripping on low gas alarm and not tripping on low low gas alarm. The manufacturer can help you with this, along with operational procedures in your area.

"1) Most leaks are not large, unless the enclosure is damaged and then it has probably already flashed over."
- Do you have evidence of this? This statement does not seem to be based on fact, rather someones opinion.

"2). Since the system is designed for 350 kV and operating at 230 kV, the insulating quality of SF6 gas at one atmosphere pressure is sufficient per the GIS vendor, assuming the SF6 is clean.
3) The enclosure cannot leak out to below one atmosphere."
- Just because pressure in the enclosure won't drop below atmospheric pressure does not mean there will always be a sufficient amount of SF6 gas to break a fault. If there is a leak, chances are the gas will be dirty with moisture.

"4) The operator should have time after the first alarm to get the problem corrected or at least isolated."
- The only thing an operator can be expected to do in this situation is manipulate other elements to lesson the impact of a break fail due to low SF6. An operator cannot fix a breaker from his control room. There are two types of alarms the system will send to an operator in most cases - Sustained alarm or Trip alarm. I believe Low gas should be sustained, and low low gas should be trip.

"5) The station building has large storage of SF6, spare parts, a transfer/filter cart and trained staff on duty to make a timely repair."
- To use this as an operational standard is very poor practice in my opinion.

"6) Tripping the transmission line sheds other customers and trips a lot of generation."
- Unless other outages are concurrent, a 230kV system should not trip a line due to one single breaker tripping. If that is the case than your SF6 issue is only a very small part of a much larger operational issue.

"7). Low Gas Pressure systems are a source of nuisance trips."
- This may be true. But this is something that protection engineers always struggle with. My personal opinion is to trip on a low-gas alarm, in general situations. If you get a low gas alarm, chances are you are losing gas. Most of the SF6 switchgear I've seen are designed to be able to break fault current during a low-gas alarm scenario. They aren't always able to break a fault during a low-low gas alarm. If your low gas alarm is an indication of a more serious problem that could happen than it is better to trip that single element while you can. Otherwise, you risk a break fail situation which is generally much worse than a single breaker tripping. Agian, IMO, a 230kV system (no withstanding current outages, etc...) should be able to withstand one breaker opening without effecting load. Most breaker fail scenarios will shed load.


There may be other factors to consider as well depending on your location and station layout. Cold climates are not always suited for SF6 for example. Also consider a loss of AC to the breaker, which would likely result in a loss of the breaker heater. If it's cold enough the SF6 could liquefy. Here you may not want to trip if that triggers a low-gas alarm.

Hope this helps.

 

[marks1080]

Sorry rcwilson, I wrote my above reply thinking it was a single breaker you were dealing with on this issue. After re-reading it you are talking about a bus. Same philosophy applies, but different operational procedures might also be applied.

 

[m3ntosan]

rcwilson,

It makes no sense to trip the lines if you lose a bus-bar, that’s the whole idea of 1 and a half CB running arrangement, you can still operate (with reduce flexibility) even if you lose one or even both bus-bars (through the middle CB, at least with 2 gen units in your case).
IMHO you should only trip the CB on the side of the affected bus-bar.


May you grow up to be righteous, may you grow up to be true...

 

[jghrist]

The concept, I believe, is to trip all sources to the GIS that has low-low pressure. This prevents the possibility of one of the GIS breakers tripping to interrupt a fault when there is insufficient SF6 to have a successful interruption.

Another possibility mentioned was to trip the SF6 breakers with low pressure, but before the pressure gets too low to successfully trip.

Either way would prevent tripping of breakers when there is insufficient pressure, assuming that the pressure loss is not too fast. The first method seems preferable if the station is manned and the problem can be fixed most of the time without any tripping.

 

[rcw retired EE]

Thanks for the comments.

Good ideas Mark & Slavag.

m3ntosan- My use of the word bus may confuse the issue a little. I was talking about the SF6 insulated bus sectiovs that connect the transmission lines into the GIS breaker-and-half switchgear.

Two overhead transmission lines come into the plant through an open air disconnect, metering VT's & CT's and surge arrestors to the outdoor, SF6 bushings. SF6 insulated bus (still electrically part of the transmission line)connects the line into the GIS.

About 20 feet of SF6 bus, two disconnects, VT and a CT enclosures are part of the GIS switchgear, all connected on the transmission line side of the circuit breakers. For protection purposes, they are treated as part of the transmission line, the same as the outdoor disconnect or metering CT. Since the SF6 insulated equipment is part of the transmission line the utility end of the line must be tripped to remove potential from any failed piece.

 

[marks1080]

Perhaps you should defer to the utility standards/procedures if they have protective zones overlapping into your equipment. If your plant can affect the transmission system I would venture a guess you would be obliged to follow their rules. Have you talked with the utility yet?

 

[rcw retired EE]

Mark- (I typed the below comments before seeing your reply)

The two-stage low pressure alarms are monitoring SF6 density in GIS bus sections, disconnect switches, surge arrestors and instrument transformer compartments where the SF6 is used strictly for insulation, not for arc extinguishing.

All disconnects are interlocked to operate under no load with the breakers open. If a flashover occurs in these enclosures, the SF6 will help extinguish the arc, but the fault will be tripped by the dual redundant bus differential protection. Tripping for low, low pressure does not clear the fault faster, but as you state, it may deenergize a component prior to the flashover and avoid the fault.

All SF6 breakers are inhibited from operating when a breaker compartmetn has a low low gas pressure alarm. Any trips are sent to the backup breakers by the breaker failure relays.

This plant's philosophy of not tripping on low-low gas pressure came from the vendor's engineers with many years of GIS field experience. Our lead design engineer (now retired) had designed and commisssioned several GIS stations in Asia. Our philosophy of detecting and repairing leaks instead of tripping is based on opinion as you noted, but it comes from operating experience.

The local utility for this plant has a large GIS station on the same 230 kV system. They have had several faults and flashovers of SF6 bus and breakers, most attributed to contaminated SF6. Their engineers do not recall any low-low pressure trips. They have had false trips from the pressure switches.

Your brought up a valid concern about low temperature conditions which can decreasae SF6 density and its insulating and arc quenching abilities. Except for two short runs of outdoor bus and bushings, all equipment is in a building with redundandt heating and cooling backed up by diesel generator. Low temperatures should not be a concern.

If someone has GIS operating experience where a low pressure trips avoided a problem (probably hard to prove) or where you wished it tripped on low pressure, I like to hear about it.

 

[marks1080]

When using GIS equipment there is a trade off. Your system will be more susceptible to false trips, simply because the system is now more complex. This is a fact of any protection system. The major benefit I see from GIS is saving on space. An equivalent GIS system takes up much less space than a conventional system. I think low low gas in your case would require different philosophies depending on which element of your system is sending the alarm. If it's a breaker, you can't trip. Treat a breaker with low low gas as failed. If it's a section of bus, than that's a different story. If the breakers defining that bus are not alarming on low (or low low) gas, but a section of the bus is alarming low low gas I think you should trip breakers to isolate that section of bus. Sure that bus may potentially be able to supply its load during a low low gas scenario, but you have imposed operational restrictions (or at least you should) due to the low low gas alarm. A single contingency in that state could be disastrous. The clearances expected from a bus with good SF6 gas are no longer valid, and if the system should experience a fault during that time you could experience serious problems. Over voltages could cause flash overs, and I think in general it could be a serious safety issue. I think the likely-hood of getting false low low gas alarms is low. If the system was installed and commissioned properly anyway. You should treat these situations very seriously. For a low gas alarm you don't need to automatically trip, but if you want a maintenance crew to look at and fix the problem the element is going to have to be isolated regardless.

Regardless, the system should be designed to respect the loss of any single bus without losing any load to your plant. If not, than tripping on low low gas is the least of your problems