Diverting xfmr lighting arrestor discharge through a ground CT & relay 2

[bdn2004]

Most of the installations at this plant have the discharge of the transformer mounted lightning arrestor going through a case ground relay. See the attached picture.

The purpose for this I'm told is that should the arrestor fire, it can help pinpoint the issue via the alarms - that the problem is at the xfmr not somewhere else along the distribution path.

All of these are old installations. What is the parameter on the ground CT that is important to know for this installation, should we duplicate it on a new install? And when the arrestor discharges what is the current that would be flowing? I realize the arrestor works as a switch but does it also act as any kind of impedance?

 

[burnt2x]

Seeing the attached picture, the amount of current flowing into the CT primary will be depending on the grounding resistance of your ground rod and the transformer casing -transformer mounting rails-to ground resistance. Whatever LA discharge current occurring on any lightning event will be split between the CT - ground rod path and the transformer casing to ground. (If you have the chance (perhaps during any outage), try measuring your "ground rod resistance" and compare with your measured ground resistance taken from the "transformer casing-rail" section. Then you can calculate the current split, hence the possible current that would be flowing into your CT. Of course you need to assume the lightning impulse voltages or currents.)

My experience with lightning protection installations is that some OEMs provide lightning counters to give a little idea of whether the protection functioned/ was hit. If you find it of use and your company have some money to burn, you can duplicate that setup (if you can find vendors of those ground discharge sensing relays).

 

[odlanor]

In my opinion this is a gap or spinter instead of a surge arrester.
Travelling wave voltage due to lightning discharge on line when exceed eletric space voltage of the gap, current impulse will diverted to ground. the current will evolve for the industrial frequency, if become a short-circuit to earth detected by CT which will operate a ground relay.

 

[bdn2004]

Odlanor,

Specifically they are ABB XPS Station Class surge arrestors.
The new installation is calling for the transformer to have an insulating material under it, so it should be isolated from the ground except this one connection. The picture is typical of another install, not this one.

The vendor is calling for the arrestors to go to their own ground rod, tied to the system but not through this case ground cicuit. If the arrestors conduct they will cause a phase to ground fault and will trip our breaker, but the breaker is 1/4 mile away, and lots of stuff along this path that could cause this same phase to ground signal. Thus my questions why can't we run this through this CT?

 

[stevenal]

Is this an air gap as depicted? Or an MOV?

 

[DTR2011]

The discharge time of an MOV is significantly faster (micro seconds) than any relaying you may have on that line.

The document listed gives a good overview of surge arrestor performance.

http://www.geindustrial.com/publibrary/checkout/Arresters?TNR=White Papers|Arresters|generic

 

[bdn2004]

Stevenal,

They are MOV, I have the wrong symbol?

 

[odlanor]

They are MOV so there is no short circuit to ground. It is a good idea transformer to have an insulating material from transformer , so cable should be isolated from the transformer tank and connected direct to ground cable before CT.
This will reduce refleted voltage wave back to origin.

 

[stevenal]

Even if the relay could sense it, your power frequency rated CT will not reproduce the high frequency of an MOV discharge very well. The CT as shown will only sense a faulted arrester or a faulted transformer.

I agree with the vendor. Arresters work best if lead lengths are as short and direct as posible. The extra (high frequency) impedance of your wound primary (I assume) CT will ensure a higher voltage at the transformer terminals in an overvoltage event. Of course you would need to set your arresters on insulators to keep them isolated from the tank if you follow the vendor's advice.

This type of transformer ground protection is called "case ground." IEEE warns: "The system should be tested periodically to
determine that no accidental grounds have been added. Incorrect operations can result from accidental grounds from power tools and transformer auxiliary equipment."

 

[bdn2004]

Stevenal,

This is the old way they have connected everything likely 1950's vintage when MOVs weren't even invented. With the old type silicon gap arrestor would that have been a concern?

So what you are saying is keep the arrestors isolated from the tank and go directly to ground as the vendor specifies.

 

[burnt2x]

Please understand that you are trying to protect those transformers from the ill effects of lightning strokes! The best location for those arresters are right beside the protected transformers. You also need to provide as short a discharge lead to ground as possible, no tight turns!

 

[bdn2004]

Smallgreek,

What is the discharge voltage and current? And how do you calculate this? For example, What if the impulse were 50kA and the resistance to ground is 10 ohms. I tried to read through the paper you reference. I'm not getting it from that.

There is a question of what size this wire should be to the ground rod. On the arrestors themselves it allows for #2 all the way to 1000MCM.

 

[bdn2004]

burnt2X,

The surge arrestors are mounted on top of the transformer within about a foot of the high voltage bushings. And they are mounted on the far side, so it will be a straight shot down to the ground.

The vendor is also recommending these be insulated and (3) separate wires to a ground rod. I'm thinking about putting bare leads in a PVC conduit. They don't tell us the size wire to use though. I would like to have some backup materials saying why a certain size was picked as opposed to some "well that's how the rest of them are done"...

 

[burnt2x]

The vendor is also recommending these be insulated and (3) separate wires to a ground rod. I'm thinking about putting bare leads in a PVC conduit. They don't tell us the size wire to use though. I would like to have some backup materials saying why a certain size was picked as opposed to some "well that's how the rest of them are done"...

bdn2004,
Regarding wire sizing for lightning ground leads, the criteria being used is "fusing current" capacity of wires and not "current carrying" capacity. E.g. most substation system ground wires utilize 4/0 copper clad steel wires which has a fusing current of about 49kA in 30 cycles (0.5 seconds). If you are sure that the lightning discharge current is lesser in your case, you can use smaller wires, 7-strand #6 (31kA in 0.5 seconds) perhaps.

Please find attached link for fusing currents data for the different sizes of copper-clad steel wires used as ground and OHG wires.

On the decision to use "insulated" versus "bare" wires, I guess you will have to inquire from your vendor why they recommended those "insulated wires" as ground wires when these are not current carrying wires!

 

[bdn2004]

I think the issue is they want to keep the arrestor leads away from the case ground which is connected to the tank, thus the insulated wires. And could there not be some arcing.

And burnt2X thanks for the great info.

 

[stevenal]

I don't think you'd want to ground SiO2 arresters through the CTs either, but I'm not aware of their characteristics.

The reason for insulating the ground leads is to prevent the core protection from failing to operate properly for a faulted transformer. PVC conduit would do the trick. For the same reason you will need to insulate the arrester bases from the transformer tank they are mounted on.

 

[magoo2]

I think you mean silicon carbide not SiO2. They have been replaced by metal oxide varistors or MOVs.

 

[stevenal]

Yep.

 

[odlanor]

bdn2004
ABB XPS surge arrester are applied from 3kV to 500kV system.
1- what is high voltage of transformer ?
2 - what kind of grounded for high voltage system?
3- what are duty cycle or MCOV of XPS surge arrester?
These are the data necessary to answer your qeustions.

 

[rcw retired EE]

I thought the surge arrestors are to protect the transformer windings. Isn't the voltage stress on the windings between the energized winding and the tank? Wouldn't it be better to connect the arrestor directly to the tank to minimize that voltage? Also, the steel tank probably has better high frequency impedance than the ground wires.

Surge currents through the case ground CT will create a tank to ground voltage rise, but that will not be seen by the windings, unless the winding is externally grounded through a neutral terminal.

A minor problem with case ground CT protection is it also responds to ground faults on auxiliary equipment like a faulted transformer cooling fan. The CT and relay can't tell the difference between 120V and 115 kV fault current.