High Zs RCD for fault protection?

[Bradley6969]

Hi, kind of follow on from my last post, but my boss hasn’t explained very well hoping you guys can

High Zs on circuit,
Can you reply on an RCD for fault protection?

He’s told me if the circuit is RCD then you can use that as fault protection, obviously you can’t meet disconnection times for MCBs I’m unsure as how to the MCB would disconnect short circuit and overload, I know the RCD covers L-E and N-E faults, he said something about checking the Line-Neutral loops but I’m unsure.
Thanks

 

[pc1966]

You can use a RCD for fault disconnection, but it is much better not to. Basically the RCD (or equivalent electronic portion of a RCBO) is much more likely to fail than a fuse/MCB!

Normally you should always design so the Zs value is low enough that the OCPD disconnects fast enough to meet the shock protection time limits in the BS regulations of 0.4s (most final circuits) or 5s (sub-main and some fixed circuits). However, you might have a high Zs due to the CPC impedance on a long circuit run in T&E (with the reduced CPC size) or on a high current circuit if on TN-S and the Ze value is on the high side. Or most obvious of all, on a TT system with an earth rod.

If the MCB is selected to provide overload protection of the cable (and not just fault/short protection) then no matter what the PSSC it will eventually trip and save the cable. Its approx inverse-time tripping curve is designed to match cable protection for that reason!

Usually it would also meet the 0.4s disconnection times for shock protection (the issue behind Zs/fault clearing times) on a L-N short as if you feed meets the 5% voltage drop on rated circuit operating current, and the MCB is around that value, then on a short you are getting around 20x the nominal current and that is enough to hit the magnetic "instantaneous" trip of the MCB so disconnection is going to be typically 10ms or less:

• B-curve 3-5 * In
• C-curve 5-10 * In
• D-curve 10-20 * In

However, you don't have to meet the 0.4s /5s disconnection times on a L-N fault but you must be safe. So the OCPD current/time profile should not allow the cable's adiabatic limits to be exceeded and so the insulation damaged (or fire risk)!

 

[Bradley6969]

You can use a RCD for fault disconnection, but it is much better not to. Basically the RCD is much more likely to fail than a fuse/MCB!

Normally you should always design so the Zs value is low enough that the OCPD disconnects fast enough to meet the shock protection time limits in the BS regulations of 0.4s (most final circuits) or 5s (sub-main and some fixed circuits). However, you might have a high Zs due to the CPC impedance on a long circuit run in T&E (with the reduced CPC size) or on a high current circuit if on TN-S and the Ze value is on the high side. Or most obvious of all, on a TT system with an earth rod.

If the MCB is selected to provide overload protection of the cable (and not just fault/short protection) then no matter what the PSSC it will eventually trip and save the cable. Its approx inverse-time tripping curve is designed to match cable protection for that reason!

Usually it would also meet the 0.4s disconnection times for shock protection (the issue behind Zs/fault clearing times) on a L-N short as if you feed meets the 5% voltage drop on rated circuit operating current, and the MCB is around that value, then on a short you are getting around 20x the nominal current and that is enough to hit the magnetic "instantaneous" trip of the MCB so disconnection is going to be typically 10ms or less:

• B-curve 3-5*In
• C-curve 5-10 * In
• D-curve 10-20*In

However, you don't have to meet the 0.4s /5s disconnection times on a L-N fault but you must be safe. So the OCPD current/time profile should not allow the cable's adiabatic limits to be exceeded and so the insulation damaged (or fire risk)!

Ok I kind of understand better now, so the MCB should in theory operate aslong as the correct size cable is used for the rating of the MCB and the Volt drop isn’t exceeded? Appreciate your reply I’ll keep that

 

[pc1966]

Ok I kind of understand better now, so the MCB should in theory operate as long as the correct size cable is used for the rating of the MCB and the Volt drop isn’t exceeded? Appreciate your reply I’ll keep that

Yes, in most cases you should not have any problems to disconnect on the MCB aspect if the other design factors are OK. And if you can then try to stick with that! The OSG Table 7.1(i) has the cable length-limits for most common cases so is the easiest starting point.

But there will be cases when it is impractical/uneconomical to meet Zs by that route and then you may have to depend on the RCD.

The classic case is a TT installation, and my own preference (but more expensive than some might want) is to have the traditional 100mA delay-RCD incomer and an all RCBO board, that way for any fault to Earth above 100mA there are two independent RCD that can trip to disconnect, normally the 30mA RCBO would go, but if has failed then the incomer would cut everything.