Ground Fault sensing on secondary of distribution transformer

[rockman7892]

I was having a conversation with someone today who wanted to install a set of CT's and associated ground fault relay on the secondary of a 3MVA 4.16kV - 480V Delta-WYE grounded transformer. The 480V secondary of the transformer is being connected to a 480V SWGR lineup through non-segmented Bus Duct and it has been requested that these CT's and relay be placed at the secondary of the transformer in order to pickup on ground faults occurring within the Bus Duct in order to prevent damage or explosion of the Bus Duct during faults. I have never seen this application on a 480V transformer of this size however it was stated that several Bus Ducts at this location have experienced faults and "blown up" or been significantly damaged.

On the primary of the transformer there is a 4.16kV fused switch directly connected to the transformer primary, and there is a 4.16kV feeder breaker upstream of the switch. The proposed idea would be to trip the upstream breaker when a ground fault was sensed on the secondary CT's through the dedicated ground fault relay.

As I stated I have never seen this sensing used for this size application so I was wondering if this was typical for this type of application? Is there a protection benefit to be gained by sensing this ground fault on the secondary and thus better protecting the Bus Duct? Is this worth the addition cost associated with CT's and relay? Any thoughts or past experiences would be appreciated. Thanks

 

[davidbeach]

You're not going to, in any practical manner, detect low-side ground faults from the high side. So, if it is desired to clear low side ground faults between the transformer and the first breaker, you're going to need CT(s) and relay(s). If all you want to detect and interrupt are ground faults, you can use a single CT in the transformer neutral. If you want phase protection too, then you'll need a three-phase set of CTs.

 

[Kiribanda]

rockman7892,
See what Davidbeach had mentioned. When there is a SLF on the secondary, the corresponding line current on the primary side (DELTA)
will be 58% of the secondary fault current multiplied by the transfomer turns ratio. That means if the SLF on the secondary is say 1A,
then the corresponding A-phase line current on the primary is (1/1.732) x 0.115A= 0.066A which is not enough not blow the primary fuse.
Also B-ph=0A and C-ph= 0.066A. Therefore, you need to have a coordinated ground fault protection scheme using line & neutral CTs on
the low side. The best possible method is to have a REF protection (87N) scheme on the secondary side.

 

[ScottyUK]

Kiribanda,

REF is an ANSI 64 code, or is over here.

A CT on the neutral point is often used in a standby earth fault protection implemented using a simple relay, independent of the multifunction wonder doing the main protection functions, as a last line of defence in schemes where a full duplicate scheme can't be justified.

 

[rockman7892]

Thank you for the replies.

I see your points above that the fuse will not provide adequate protection for secondary ground faults.

By picking up on ground faults at the transformer secondary will this prevent or minimize damage to the Bus Duct as opposed to relying on the primary fuse. I'm guessing that if the ground fault is set low enough it can minimize damage to the bus duct where as without it, the primary fuse would not pick-up for a ground fault and the ground fault would then escalate into a 3-phase fault which could then cause catastrophic damage to the Bus Duct.

Obviously for pickup settings you will need to coordinate with the ground fault protection downstream so I'm guessing that as far as a pickup and time delay we'd be looking at something around 10000-1200A pickup and a 0.3 to 0.5s time delay? Is this fast enough to prevent damage, or would an 87N be a better scheme?

Also in this proposed application the neutral CT is being field fit so it wont be installed on the transformer neutral (for different reasons) and will be installed in the transformer throat connection. The transformer neutral will therefore not be bonded to ground at the transformer but will be carried to the Switchgear and be bonded to ground at the switchgear with the neutral CT on the bonding bus between the transformer and Switchgear. Are there any issues with this arrangements?

 

[ppedUK]

rockman,

Not sure what you mean by "transformer throat connection", but earthing of the Neutral at the switchboard is common practice (at least in the UK) and a standby E/F CT can be fitted on the N-E link bar in the switchgear to detect earth faults on the entire LV system, but if you go with REF (best option in my opinion), then you need to fit 4 CTs (and a suitable relay) at the switchboard.

If you fit a CT in the earth bond between the switchboard and the transformer as you seem to be describing, then to detect an E/F in the busduct, with the N-E connection at the switchboard, the busduct would have to be isolated from the switchboard earthing system, otherwise an E/F in the busduct will partially bypass the CT and return to the TX star point via the Neutral conductor.

You can fit just one CT in the Neutral-Earth connection at the switchboard and fit an Standby Earth Fault Relay as the others have said, but this will have to grade with all downstream E/F protection and is normally time delayed as a result. As Scotty says, a Standby E/F scheme is the last line of defence.

Based on your previous fault history, I would go with REF every time (Scotty I use 64 for REF too). An REF scheme can be instantanteous without having to grade with downstream E/F devices and can be set at a very low current threshold.

 

[jghrist]

You could put in transformer differential protection and include the bus duct in the zone. That way you can get fast phase and ground protection for bus faults without having to coordinate with downstream overcurrent relays. You would also get faster protection for transformer faults.

Another alternative would be to use zone interlocking protection (fast bus protection) where high side phase and neutral ground instantaneous overcurrent protection sees through faults and trips the high side breaker, but is blocked by low side feeder overcurrent relays picking up for feeder faults. If you already have high side and neutral ground overcurrent protection that can accommodate the necessary blocking signals and has the logic, you could save some money, but I would prefer the transformer differential solution.

 

[rockman7892]

This conversation has come up again and I wanted to make sure I had a good understanding of what others mentioned above.

On the attached documents the 2nd picture is how we are proposing to configure the grounding of this separately derived system per the NEC. It was suggested that if we put the CT somewhere along the system bonding jumper shown in the diagram at the Switchgear than we will be able to detect low level ground faults anywhere between the transformer and the Switchgear. Am I correct in suggesting that we can located this CT here and detect all ground faults?

Others are still suggesting that we locate the CT somewhere in the busway itself although I wont have these final details until I perform a site visit later in the week. If the CT were to be located in the busway then I'm guessing that it would have to be located somewhere along the supply side bonding jumper shown in the diagram between the transformer and switchgear (perhaps on busway ground bus). ppedUK point out above however that if the CT is located on the supply side bonding jumper then the busway would have to be isolated from the switchgear ground connection or else current would bypass and return on neutral. Can someone explain further how some of this current would bypass CT in this instance I just don't see if from the diagram? It appears to me that all ground fault current would have to pass through system bonding jumper regardless of fault location.

In reading through some of the responses above it appears that the REF protection is a better option for this instance. Can you use a similar in-expensive ground fault relay for the REF as you would with just the neutral CT sensing so that the added cost will really only be the addition of the (3) phase CT's?

Thanks for the help.

 

[waross]

Confusion reigns! Neutral busses are for neutrals and ground busses are for grounds.
Mixing grounds and neutrals on the same bus makes a well planned ground fault sensing system difficult.
Try this:
Locate or install an insulated/isolated neutral bus. Connect all neutrals and only neutrals to the insulated/isolated neutral bus. Insulated/isolated neutral busses may feed other insulated/isolated neutral busses. The main insulated/isolated neutral bus may be located at the switch gear or at the transformer. The best location is often close to the main ground bus.
Locate or install a ground bus.
Connect all grounds and only grounds to the ground bus. One ground bus may feed downstream ground busses.
The one and only exception to the above rules:
Run a system grounding jumper from the main neutral bus to the main ground bus. Install the ground fault sensing CT on this system grounding jumper.
With this arrangement any neutral current may return to the transformer without passing through the GF CT.
Any and all ground fault current must return to the transformer through the GF CT on the system grounding jumper.
Your proposed setting seems a little high. 2 or 3 times the capacitive charging current may be all that you need. 5 Amps, 10 Amps 20 amps, somewhere along that line.
Neither drawing has any single wire that does not pass neutral current but does pass ALL the ground current from any event.
The next words I expect to hear are;
"But we have no line to neutral loads."
My answer;
"YET"

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