Frame Leakage protection: 3 phase fault 1

[Prot123]

Hi All,
Just wondering if the switchgear Frame leakage protection is good for 3phase bus fault? Cheers

 

[davidbeach]

Very little of the fault current will return via ground. The closer to perfectly balanced the fault, the closer to zero the current detected by the protection.

 

[Prot123]

Thanks David. So what do you think, is Frame Leakage protection good to protect against accidently earthing of 3phase bus via earthing truck??

 

[davidbeach]

My system doesn't have any frame leakage protection and I'm pretty sure that the rest of the protection system won't miss anything that a frame leakage system might pick up. Away from generators everything is solidly grounded; a better approach as far as I can tell. To each their own.

 

[RRaghunath]

Frame earth leakage protection is for unearthed systems I suppose and will help detect first phase-to-earth fault in the system.
I only read about this in text books. Is this kind of protection used these days! For that matter, even the unearthed systems are no more preferred I understand.

Rompicherla Raghunath

 

[marks1080]

Frame leakage is meant to protect against ground faults. I wouldn't use it for phase. Frame leakage, specifically, is for when you have live tank equipment with CT on only one side.

 

[veritas]

Agreed - forget about frame leakage for phasefaults. It is intended for groundfault detection. Having said that I do not see any of them around nowadays. Replaced a few a couple of years ago with a standard EF protection scheme. Not seen or heard of any since. FL are notoriously difficult to manage. A ladder left leaning against the swbd could defeat the FL scheme for example.

 

[Prot123]

The pickup current for frame leakage protection is very small and doesn't have any intentional time delay. A 3 phase fault within the bus bar wont be balanced completely, because of arc, and a small current will flow through the ground which will be seen by the F/L relay. Only case when there will be no neutral current flowing is when we have a perfectly balanced 3 phase fault/ or LL fault , happening of which I think is very rare/impossible. So with this what do you think is FEL still reliable?
We have FEL protection on our older switchboards, more than 35 years old. The newer ones have High Z protection. Also bit curious on how you replaced FEL protection with EF protection. My understanding is FEL protection is to protection the switchgear and needs to be quick; If you replace it with EF I assume it will be slow as you cant trip the switchboard instantly on EF for grading reason.

 

[davidbeach]

But it got replaced with some form of differential protection that can be as fast as you can make it. Frame leakage protection seems to be one of those buggy whips that we haven't completely dispensed with yet.

 

[marks1080]

As mentioned before, Frame Leakage is specifically for live tank equipment with a CT on only one side. The Frame Leakage zone is only half of a protection zone. IE: You only trip the half of the zone behind the CT. It protects the very small blind spot you end up with when using only a CT on one side of a live tank breaker.

 

[veritas]

Prot123 - We replaced the FL with standard EF as well as a blocking scheme I should add. You seem to be fixated with retaining your FL scheme even though the wider engineering community views it as dated technology that should be discouraged as their are better ways of doing swbd groundfault protection nowadays.

 

[rcw retired EE]

1940 vintage large AC-DC conversion equipment like mercury arc rectifiers and the associated 6-phase, rectifier transformers had earth leakage relays (64G) for fast tripping on insulation failures. Equipment was isolated from ground except for a bar type CT feeding the 64G overcurrent relay. Transformers were mounted on railcar wheels with insulated bearings.

When the mercury arcs were replaced with diodes & 208V cooling fans were added, we discovered the CT & 64G would trip the 50MW potline when a 1/4 HP fan motor failed. The CT could not distinguish between 120V and 13.8kV ground current.

The CT's also didn't do very well on DC faults, they tended to melt, leaving the neutral side of the CT circuit the only ground on the transformer. That ground path melted the relay wiring and a large hole in the relay panel. Wiring was changed on all the other units after that discovery.

I've tried adding differential relays to medium size (8-10 MVA) rectifier & VFD transformers and not had much luck. The high harmonic waveforms on the LV side confuse the digital relays causing constant 87T trips. We had to set the harmonic inhibit low to block during normal load and allow tripping if the fundamental fault current was large enough to swamp the % harmonic current. That setting results in not much better protection than a straight overcurrent.

 

[Prot123]

Thanks all. Wanted to clarify Veritas on my attachment with FEL.
I'm not in favour of FEL, however, for the switchboard I have on hand there is no way I can integrate any new bus protection scheme, because of CT limitations .Also with reverse blocking on this sectionalized busbar with multiple incomers and no bus tie CT protection security will be 0. With this situation, till the time this board is replaced, I have no choice without relying on FEL.

 

[submonkey]

Hi Prot123,

A major Australian utility where I used to work applied earth leakage
protection quite extensively on double bus 11 kV metalclad switchgear.

Frame leakage was used because of the difficulty of applying high impedance
protection on a board where breakers could be racked to either of two buses.

The scheme works, but there are a number of traps:-

- Need to ensure good frame isolation from ground. A frame - earth insulation
of > 10 ohms was the set standard, and the schemes were installed with a built
in system for periodically measuring the frame isolation impedance.

- The need to work out what should be tripped. A selection scheme was used which
tripped any feeding transformers with starpoint earth current flowing.

- Frame currents associated with feeder faults. The scheme had a feature which
delayed FL tripping when significant earth current flowed in an outgoing feeder.

The scheme cannot be relied upon to detect phase faults not involving earth. Most
faults in metalclad boards involve earth, and the intention is to speed up clearance
for these. Three phase faults and pure phase-phase faults need to be cleared with
overcurrent.

If I were designing a new station, I would look closely at schemes like Low Z
differential and OAFD as alternatives. I wouldn't completely dismiss FL though -
especially for brownfield - it has its uses.

Thanks,
Alan

 

[Prot123]

Thanks Alan for the detailed reply. Just a question, do you know of any standards which confirms the FEL insulation to be >10 ohms?

 

[submonkey]

Hi Prot123,

The 10 ohm figure was an internal utility requirement. Unfortunately I'm
not aware of any public documents which list a frame insulation resistance.

New switchgear installations designed for FL were mounted on an insulating
mat and achieved far better than 10 ohms. Some FL installations were retrofits
to existing switchboards, and if the frame insulation was > 10 ohms, use of FL
was deemed viable.

Thanks,
Alan

 

[DiscoP]

10 ohms is the recommendation in the GEC Protective Relays Application Guide.

If you have access to this book, it gives an explanation as to why this is normally OK.

It is OK protection for switch gear faults, but I wouldn't be relying on it to operate if all three phases are shorted together and earthed, in theory it shouldn't see this fault.

 

[Prot123]

Thanks DiscoP, got hold of the book and will go through it now.

 

[stevenal]

NPAG

See section 15.6.

Note: This document was freely downloadable from the Alstom web site, so I believe it's okay to post this link.