Do you delay tripping on low oil? How long? 1

[stevenal]

We had a substation transformer manufacturer question our specification on this point. I believe the delay is to prevent tripping due to oil slosh from a small seismic event. Do other folks in seismic areas have similar requirements, and what's a reasonable delay? Thanks.

 

[davidbeach]

I don't know about low oil, but we do delay sudden pressure to avoid false operation.

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[wattyeng]

You would only delay any functional trip application if:
> it is safe and reasonable to do so
> it is marshalled and presented via SCADA etc to someone who can evaluate the situation
> it is part of the design spec

Many conditions can momentarily slip out of tolerances or limits during normal ops and it would be reasonable to add hysteresis or timed delay to prevent such activity from shutting down complete processes or systems

Fail safe systems cater for alarm situations before trips are initiated to allow corrective action or notification

As with all things consult with equipment OEM

I.Eng MIMMM MIET MIPowerE AIOSH

 

[cranky108]

We only alarm on low oil, but for some reason we do trip on sudden pressure. Sudden pressure, as far as I have seen only has misoperated.

 

[dpc]

I have not seen a time delay on the low oil level trip that I can recall. There's normally also a low level alarm set above the trip level. Perhaps the manufacturer wants to avoid buying a timer that they forgot to include in their bid. That said, I would not have any serious objection to a short time delay on the oil level trip.

 

[bacon4life]

I have seen some transformers trips on low-low oil and/or pressure relief valves. Hopefully the alarm on low level will allow a crew to respond before the transformer reaches the low-low setpoint. I have heard certain oil gauge designs are mostly immune to oil sloshing and other easily trip. Unfortunately I don't know specifics of which designs are which designs are more likely to trip.

As to a time delay, earthquakes can last 10s of seconds. On other sorts of devices, lengthy pickup times have resulted in confusion where there is a significant delay between an action and a system response. Perhaps some multiple of the sloshing period of oil inside the transformer rather than riding out the earthquake?

I keep wondering if the Metcalf transformers had low oil trips or if the lowering oil level eventually caused electrical faults.

 

[stevenal]

wattyeng,
Did you read the question? What is safe and reasonable are exactly my questions. The question refers directly to both specifications and consultation with the manufacturer.

Others,
Probably not much point in trying to ride out a severe event since there will likely be damage to the system. Yes there are two levels for alarm and trip. The seismic oil wave that triggers both won't give an operator much time in between. It is hoped that a leak will allow sufficient response time before the windings are exposed, but the trip function is there in case this fails to occur. Perhaps we should be specifying slosh resistant gauges rather than time delays. I have heard that a large truck rolling near transformer on a pad on bad ground caused a problem.

 

[wattyeng]

stevenal - yes, I read your question and answered it - it reads that you are specifying a low level trip delay on a transformer and the OEM who is quoting it, has questioned that function

the only persons who can give you that accurate and detailed response is the transformer OEM - they have all the relevant design detail and it's their kit that would be affected by a trip delay - if YOU are asking for it to be included, then I'm sure the OEM will for a price, provided it will not adversely affect equipment safety



I.Eng MIMMM MIET MIPowerE AIOSH

 

[OffTap100]

How low is your oil to begin with? I assume it has a nitrogen blanket?

 

[stevenal]

How low prior to a seismic event? I would assume not low at all, but in the normal range. Yes, we use N2 oil preservation systems.

 

[Palletjack]

I’m trying to wrap my head around a seismic event that’s violent enough to cause the transformer oil to slosh around enough to get the float on a oil level alarm bobbing up and down enough to alarm and trip the breaker or S&C. XF full level is usually around 11”-13” below the top on an N2 blanketed XF. That’s a lot of sloshing..

if a seismic event is that violent, wouldn’t you want it to trip because there are probably broke poles, broke bus in the station, maybe a breaker or whatever type of circuit protection you have broke off the rack and laying in the sub. Regulators may be laying on the ground also. Depends on how good your guys bolt them down..

 

[davidbeach]

In a significant seismic event I really don't want undamaged transformers tripping on oil abnormalities. And I really, really, really, don't want a denergized but still connected transformer tripping on oil. There's already going to be a huge list of transformers to test, don't need random undamaged transformers tripping on slosh.

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[Palletjack]

East coast...
We panic over here if we feel a large dozer vibrating the ground close by..



So back to the original question..
For those in a seismic area what’s a normal delay? Several minutes?

I’m assuming monthly station inspections should pick up on any oil leaks that are problematic so a few minutes delay to allow for sloshing to settle wouldn’t be a problem?

I can see where sudden pressure would be a problem there also.

 

[thermionic1]

On some of our larger and more critical units, we install an "Active Leak Detection" scheme. It's the Xfmr Low (Alarm) contacts along with a pressure transducer installed in the lower manifold, via an overbuilt PLC that converts the pressure to a contact. When both signals are sent to the Xfmr relay in the house. If both signals are high for a defined time(relay logic)-TRIP 86T, target Active Leak. If we get Low-Low and Pressure, the relay logic trips immediately.

I believe the scheme was developed by one of the Transformer Monitoring vendors. There seems to be a lot of unnecessary vendor related PLC hardware in the cabinet for what it does, but this is what is used. It's factory installed on new units, and retrofitted on older ones.

 

[unclebob]

Did a maintenance job in Eastern Canada, in a petroleum company. We were to test an oil-filled transformer and most of the meters. Main breaker was rack-off and locked. This was a partial shut-down, so only our transformer and the 4.16 kV mcc load were de-energized. Everything went ok, until I decided to unscrew the oil level meter on the oil conservator and rotate it to activate the contact. It worked, the BF on the locked-out breaker activated too tripping an upstream breaker, and a larger part of the plant.

There are days when you wished you should have stayed home.