Earth fault on a TN-S system

[Marcopolo]

So here is the situation
22kV/415V delta-star transformer with neutral point earthed on secondary
Downstream breaker of transformer 630A with earth fault setting 0.2xIn @ 100ms
This supplues a number of outgoing loads with respective 50A breakers which only have thermal and magnetic settings.
An earth fault occurred downstream the 50A breaker tripping the 630A and damaging the feeder to the load.
Is it recommended to remove the earth fault or adjust it higher?

 

[Marcopolo]

Can you show a picture of the 630A device. Sounds like it has an rcd module fitted.

It is an mccb abb tmax t6ss630 with trip unit pr222ds/pd-lsig

 

[Andy-1960]

How do you know it’s tripped on earth fault?

 

[snowhead]

And if it is a known Earth fault why not remove/repair it?
What exactly was the damage?

 

[Julie.]

Your earth fault protection is not co-ordinated. Lifting the upstream protection may not solve the issue as it may be a time thing.

Not 100% familiar with ABB units, but I think they have a separate section on the OCD for earth fault, usually identified as G let's say this setting is set at. 0.2, and 0.1, this means it will trip above. .2x 630 = 126A with a 0.1 second delay, there is usually a higher setting for instantaneous operation - this would usually be disabled on outgoing feeders.

So any protection downstream of this, must operate for earth faults with a lower current, but also much quicker than 0.1 second - because not only does it need to time out, it also has to clear the fault before this operates, so if it's 0.09 second downstream, even though it would trip first, this upstream device would still trip as the time is too close!

In the same way, if you have earth fault protection on the transformer secondary, this must have a setting greater than this 126A, and a delay long enough that these mccbs operate before this transformer protection operates.

The issue you have is determining settings that will operate quick enough to prevent damage, but sufficiently co-ordinated and discriminated to ensure that the downstream circuits operate first, followed by the upstream feeders, and finally the transformer protection - for all levels of current likely to flow.

Upping the current settings on their own won't do this, you need to compare the characteristic trip curves of the devices.

ABB do a pdf describing an approximate /simplified method here https://library.e.abb.com/public/65ddf36f7c3bd0fec1257ac500377a37/1SDC007100G0204.pdf

 

[Marcopolo]

Hello, thanks, I am well aware of what you mentioned. The point is that the 50A breaker only has thermal and magnetic protections, not earth fault.
At first I thought it was a coordination issue ast the settings are as follows:
For 630A breaker
LT 0,66xIn @ 6s
ST 3xIn @ 0.1s
G 0,2xIn @ 0.1s
For 50A breaker
Thermal 0.8xIn
Magnetic 6xIn

 

[Julie.]

Ok, I don't have the characteristic trip curves but if we assume the 50A mccb @ 0.8 has a similar curve to a 40A mcb, then both the upstream overcurrent and earth fault ought to co-ordinate with that, but it doesn't, earth fault would trip in circ 0.1s @ 300A - the start of the instantaneous trip for the 50A mccb whilst the actual earth settings on the 630A mccb are 0.2 = 126A and 0.1sec, so it doesn't co-ordinate at all.

If you set it at 0.55x, this would be 346A, and if you set it at 0.2sec this should also give sufficient time for the downstream circuits to clear whilst still not causing the nuisance trip you had.

In this case the 630A mccb would need to see an earth fault current in excess of 350A, which shouldn't be too much of a problem, if the outgoing circuits are expecting to trip the outgoing 50A mccbs they need to have an earth fault current in excess of 300A - that should be achieved by the design of the installation, - if the downstream circuits have more than 350A fault current, then the circuits would be protected by the outgoing 50A mccbs, with backup via the 630A mccb - which is all you need.

 

[Marcopolo]

Thank you, was thinking of settings rather similar. However I am concerned about the fact the feeder downstream the 50A breaker which got damaged at 126A earth fault.

 

[Julie.]

Thank you, was thinking of settings rather similar. However I am concerned about the fact the feeder downstream the 50A breaker which got damaged at 126A earth fault.

Do you know the fault current level?

126A is merely the setting the device trips at.

How was the cable damaged - if it was damaged by the current level for an earth fault, it is likely to be much worse for a phase - phase or 3 phase fault as more cores are being heated!

Larger feeders should be sized for normal current, voltage drop and fault current withstand in the event of a fault - normally the adiabatic calculation.

 

[Marcopolo]

Do you know the fault current level?

126A is merely the setting the device trips at.

How was the cable damaged - if it was damaged by the current level for an earth fault, it is likely to be much worse for a phase - phase or 3 phase fault as more cores are being heated!

Larger feeders should be sized for normal current, voltage drop and fault current withstand in the event of a fault - normally the adiabatic calculation.

Unfortunately I don't know the earth fault level that caused the damage.

 

[Julie.]

Unfortunately I don't know the earth fault level that caused the damage.

Did/can you measure the Ze?

Usually the same meter provides the pfc.

 

[davesparks]

Can you elaborate on exactly what was damaged and what sort of damage occurred?

 

[Marcopolo]

Ok, I don't have the characteristic trip curves but if we assume the 50A mccb @ 0.8 has a similar curve to a 40A mcb, then both the upstream overcurrent and earth fault ought to co-ordinate with that, but it doesn't, earth fault would trip in circ 0.1s @ 300A - the start of the instantaneous trip for the 50A mccb whilst the actual earth settings on the 630A mccb are 0.2 = 126A and 0.1sec, so it doesn't co-ordinate at all.

If you set it at 0.55x, this would be 346A, and if you set it at 0.2sec this should also give sufficient time for the downstream circuits to clear whilst still not causing the nuisance trip you had.

In this case the 630A mccb would need to see an earth fault current in excess of 350A, which shouldn't be too much of a problem, if the outgoing circuits are expecting to trip the outgoing 50A mccbs they need to have an earth fault current in excess of 300A - that should be achieved by the design of the installation, - if the downstream circuits have more than 350A fault current, then the circuits would be protected by the outgoing 50A mccbs, with backup via the 630A mccb - which is all you need.

To add, I think the 346A setting for the current is too high. Since the cable is likely not designed to withstand this amount of current even for a short period, hence the ~126A damaged it.

@davesparks so an earth fault caused a cable damage, tripping the whole 630A supply tripped due to this. Downstream tge 630A are multiple 50A thermo-magnetic mccbs with settings specified in an early post.

 

[DPG]

What was the damage though?

 

[Marcopolo]

What was the damage though?

The cable downstream the 50A breaker and the trip on 630A breaker