Transformer Protection 3

[m4verick]

I'm aware of the difference between a Buchholz relay and a differential relay. I've read that in case of internal faults such as short circuits in a transformer Buchholz relay act as the primary protection and a differential relay as the secondary source of protection but I don't understand why. Isn't the electrical quantities/signals supposed to be more accurate and faster than the mechanical/pressure signals?

 

[crshears]

If I understand it correctly, not all pressure waves in transformers are caused by short circuits; a high resistance internal connection that fails open just enough to allow the propagation of an arc may generate a significant pressure wave that very correctly trips the transformer off potential, despite the fact that the power in = power out therefore no differential condition exists. Under the condition cited, which electrical quantity change would you recommend be used to initiate a protection operation?

Others may have better answers...

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[davidbeach]

What do you want, to try to start some sort of religious war? ;-) Or, maybe some sort of Mac vs. PC debate?

There are those who claim that the Buchholz or other oil relay is absolutely essential in all cases and there are others who are of the opinion that relays are now capable of detecting all faults that would also cause the oil relays to operate. Oil relays are subject to misoperation due to through faults and seismic events. Various means are employed to desensitize the oil relays to prevent the false operations; do these also desensitize the relays against faults for which they should operate?

One of the more promising developments in relaying is the ability to do a differential in the negative sequence domain. In the phase and positive sequence domains the differential has to ignore inrush, excitation, and transformer losses. That means those differential algorithms are going to miss many types of low magnitude faults. In the zero sequence domain there can be currents on one side that aren't on the other. In those cases the zero sequence has to be filtered out of the phase quantities. REF type protection can sensitively detect ground faults on windings that are connected to a neutral that are brought out and grounded.

But in the negative sequence domain a healthy transformer is only an impedance; no load, no source. The phase shift of the negative sequence current is opposite that of the positive sequence, but otherwise the negative sequence is unaffected. One of the most difficult transformer faults for a phase differential to detect is a turn-to-turn fault. The change in terminal currents due to a turn-to-turn fault is small enough that it will be ignored in the phase or positive sequence domains. But those small changes in terminal current show up strongly in the negative sequence domain. With no need to desensitize the negative sequence differential against normal conditions these small changes become actionable.

There's fair amount of debate as to whether or not the negative sequence differential will allow the retirement of the oil relays, but change comes slowly.

 

[krisys]

"davidbeach (Electrical)"

This concept appears to be new. I am not sure whether this is a time tested one. But the Buchholz relay should not be considered as the back up protection. It is a complementary protection. Experience has shown that this can operate as fast as a cycle or two after detecting the fault.

Fault due to the core burning cannot be detected by the differential protection at the early stage. The Buchholz relay can at least alarm such symptoms. Hence in my view one should not remove the Buchholz protection.

 

[prc]

Both Buchholz and differential relays are required and they turn alternately primary and secondary protection depending on the type of fault. Transformer engineers want both these relays and I don't think any one of them is going to outdate the other as they always act as standby in case one of them do not act, sometimes say, due to mal settings etc.

In case of a flashover from primary bushing tail to tank- Differential primary, gas relay secondary
A loose joint in lead or tap changer contact getting overheated- gas relay primary, differential may come in case the contact opened out in due course.
An inter turn fault- If a sectional fault, differential primary ,gas relay secondary. But if only one or two turns are involved, the extra current drawn from line will be small and below setting of the differential relay. Then gas relay may or may not pick up depending on the severity of fault. Recently most of the differential relays have negative sequence current detection feature that claims inter turn fault sensitivity up to 1% of total turns.
A fault in core - only gas relay will pick up.

 

[cranky108]

If the Buchholz is an actual relay with sensitivity settings, please let us know how to test these with an actual test set.

It's not that I don't trust these, it's that we don't know how to control how they are set.

I've seen the differential operate faster than the pressure relay, so I like to say both are primary, with different operating inputs.

In many cases the differential can also cover an area between the breaker and the transformer, where the pressure relay can't.

 

[waross]

The Buchholz responds to both gas volume and oil velocity. I understand that the gas volume to trip is mostly set by the volume of the chamber. This setting may not be adjustable.
The other function of the relay is to respond to higher than normal oil velocity from the tank to the conservator. There are claims that an external fault may cause the oil in the tank to expand fast enough to cause the oil velocity feature to activate. This condition may not be picked up by differential protection. This feature may back-up inverse time relays particularly if the inverse time setting is less than optimum.
I haven't tested one for many years but as I remember we injected a small amount of dry nitrogen into the tank. We opened the bleed screw to vent off the accumulated gas and put the Buchholz back in service.
The relay did not have the oil velocity feature. It was mounted at a high point rather than in the line to the conservator.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[prc]

Buchholz relay settings cannot be adjusted at site. But you can get relays with different sensitivity - say tripping with oil velocity of 1 or 1.5 or 2 meter per second( see EN 50216 -2002 Buchholz Relays) So transformer maker selects the relay. If you select 2 m/s relay, it is less sensitive than 1 m/s relay. But some times such less sensitive relay will be required to avoid mal operation with switching of cooler pumps.

Relay manufacturer type test these relays on his test bench where specified oil velocity is maintained in pipe and relay tripping is checked.

 

[vijaypandiyan]

In case there is a little arc or insulation faults between turns, break down of core of transformer, core heating, the transformer insulating oil will be decomposed in different hydrocarbon gases, CO2 and CO. The gases produced due to decomposition of transformer insulating oil will accumulate in the upper part the Buchholz container which causes fall of oil level in it and maybe causes air bubbles inside the transformer that can be easily picked up by the Bucholz relay. Differential relay cannot find this kind of internal arc faults unless they draw current.

Thanks,
Vijay