pressure relief device 3

[Sn00ze]

What is everyone's views on trips vs alarms from TX gauges.

example i have 2 PRD's one on a selector tank and the other on the diverter tank.

I saw in a double forum that some people have stopped putting these PRD's as trips because they get a lot of nuisance trips. But it seems to me these detect "catastrophic" events so i'd say they should trip.

Also, i have a winding temp indicator (CT), one signal is 75 degrees the other 90. Would you put the 90 one as a trip? or both as alarms?

i'd appreciate the input.

thanks

 

[bacon4life]

Due to false trips from bad RTDs measuring transformer temperature, we have removed trips for winding temperature in locations that are monitored by SCADA. I would put them both as alarms unless there is a reasonable chance the transformer would have significant loss of life before someone takes action to reduce the overload. Also keep in mind that transformer temperature measurements are not particularly accurate, so I would be hesitant to trip at precisely nameplate temperature. I have measured two adjacent RTD wells on a transformer that had more than a 10C degree deferential between them.

PRDs also operate for maintenance mistakes like putting in too much oil, or for putting in oil too fast. If you have microprocessor based differential relays along with a sudden pressure/Buchholz relay, I don't think adding the PRD provides much additional value.

 

[prc]

Maloperation from PRD is rare. But it may operate due to installation mistakes like- closing the valve to conservator, air cell in conservator and too high filling oil level at ambient , clogging of breather with free breathing system. Such conditions require a tripping of transformer and hence PRD may be wired for trip. There are also cases where Buchholz did not operate by PRD operated due to low impedance internal faults.
Remember temperature gauges are only indicators and normally only alarm will be required. But trip contacts always wired for trip with amble margins for expected overloads..

 

[stevenal]

We use sudden pressure (also known as rapid pressure rise) relays to trip, pressure relief devices to alarm.

 

[cranky108]

No tripping for anything we can't do a quality test for. So at what pressure should a PRD trip at?

NERC requires a test that can be repeated, so would need a test value.

 

[stevenal]

Contacts should transfer at the relief pressure. Relief pressure should be somewhat below that which will separate the welds on a transformer.

Sudden pressure relays require special test devices, since they work on time and pressure.

 

[ScottyUK]

Anything which causes a PRD to operate is a major fault. I'd expect to see it trip the transformer off the system.

For anyone specifying a PRD, it's worth considering the oil director which fits over the PRD - during a severe fault the valve acts in a manner similar to a fuel injector, creating small droplets which readily ignite.

 

[prc]

cranky, PRD has a test value usually 6-20 psi range. But this is static pressure test value. In reality, with an internal fault, it will see dynamic pressure very much higher than static test value.

 

[stevenal]

Anything which causes a PRD to operate is a major fault

How about a regulator failure on nitrogen blanketed system?

 

[ScottyUK]

That's an interesting one - don't see that design over here, not with an active pressurisation system

Closest I've seen to that failure scenario was when an SF₆-oil bushing failed and a slightly confused oil laboratory reported that we appeared to have SF₆ in the Buchholz and could we explain why? On a non-conservator design I suppose we would have had full gas pressure into the tank until the PRD lifted.

 

[prc]

stevenal&Scotty, a much common PRD mal operation is with sealed conservator by air cell (rubber balloon). When transformer is overloaded with high ambient air, expanded oil volume will be more than the capacity of conservator and due to air cell, it cannot overflow through breather pipe. Result is development of high pressure inside tank and PRD operation. Another scenario is with oil filling more than necessary at the ambient air level. Then when transformer is loaded to full rating, expanded oil volume will be stopped by air cell to create PRD operation. I have seen such failure modes on many occasions.