Cpc size on unfused spur

[John D v2.0]

So the cpc in t&e is sized to be sufficient not to melt before the MCB trips.
We can get away with unfused spurs in rfcs because of the single accessory special dispensation (i don't believe there's a normal reg that allows it for a socket outlet is there?)
If the load was hard wired and unlikely to overload eg water heater we could also have a 2.5 cable, we could also take account of the special dispensation i think?
However the cpc in the cable is only 1.5mm therefore wouldnt comply regarding minimum cpc size. The minimum may up to around 1.8 for a 32a mcb.
Is that just classed as a fudge to allow old rfcs, or is that a more general exception?

 

[EFLImpudence]

What makes you think it IS undersized - for any of the things you have mentioned?

 

[John D v2.0]

Ok i think i got it, the up front fuse <=125A in the cutout would limit the energy further than the mcb, so it would be ok! But I'll leave this thread for posterity!

 

[EFLImpudence]

The cut-out fuse is nothing to do with it.

It has to do with the fault current (Uo/Zs) and the time in which the device will trip at that current (or its energy let-through) as to whether the CPC will be damaged.



Also remember 'r' (rfc) stands for radial as well.

 

[John D v2.0]

What makes you think it IS undersized - for any of the things you have mentioned?

Sorry missed your post, i was looking at the adiabatic equation and didn't realise the let through energy depends on the up front fuse.

 

[JohnW2]

... We can get away with unfused spurs in rfcs because of the single accessory special dispensation (i don't believe there's a normal reg that allows it for a socket outlet is there?)

Eh? One is allowed one double socket (or one single socket) on an unfused spur from a ring final. That reg is presumably about L/N operating current, not CPC fault current.

... However the cpc in the cable is only 1.5mm therefore wouldnt comply regarding minimum cpc size. The minimum may up to around 1.8 for a 32a mcb.

I'm not sure what regulation (regarding 'minimum CPC size) you are talking about. If one undertakes the adiabatic calculation (rather than relying on the tabulated, 'very conservative', figures for those too lazy or unable to do the calculations), one usually finds that the CPCs in T+E have an appreciably (or much) greater CSA than is actually 'necessary'.

Don't forget that we are talking only about fault currents (due to an L-CPC fault of 'negligible impedance' - hence cleared by an OPD very rapidly). Essentially by definition, overload currents (which may be much more protracted) cannot flow through a CPC.

... or have I misunderstood your concerns?

Kind Regards, John

 

[John D v2.0]

My posts keep crossing others'!
Basically my concern was if the pefc is high enough, eg 10kA, due to a low Ze, the let through of the mcb due to tripping time would be enough to melt the cpc.

 

[JohnW2]

Sorry missed your post, i was looking at the adiabatic equation and didn't realise the let through energy depends on the up front fuse.

It doesn't if the 'up front' (I presume you mean cutout) fuse doesn't blow (which it very very rarely will, if a local OPD operates)!

Kind Regards, John

 

[John D v2.0]

It doesn't if the 'up front' (I presume you mean cutout) fuse doesn't blow (which it very very rarely will, if a local OPD operates)!

Kind Regards, John

In that case let through energy wouldn't be relevant because it would just come from the current squared time the tripping time.
If that energy is higher than the pre arc energy for the cutout fuse then that would go as well.

 

[JohnW2]

My posts keep crossing others'! Basically my concern was if the pefc is high enough, eg 10kA, due to a low Ze, the let through of the mcb due to tripping time would be enough to melt the cpc.

Fair enough.

As I say whenever such things are mentioned, although my experience is totally limited to domestic installations, I have never seen a PEFC appreciably above 1kA (even in houses right on top of a substation), let alone anything approaching (or exceeding) 10kA.

Maybe I've just been lucky. However, if you think about it, a PEFC of 10kA implies a Zs (from the point of the fault) of 0.023Ω - and, if the fault is on a final circuit, even if Ze were zero, the R1+R2 of any final circuit of appreciable length would probably be more than that.

Kind Regards, John

 

[JohnW2]

In that case let through energy wouldn't be relevant ....

The let-through 'energy' of the cutout fuse (which hadn't blown) would obviously be irrelevant, but the let-through 'energy' of the OPD which cleared the fault would be highly relevant.

... because it would just come from the current squared time the tripping time.

Exactly - the tripping time of the device which actually operated. One of the problems with trying to undertake adiabatic calculations is in deciding what 'tripping time' to use - the availble curves for MCBs are always essentially 'vertical' from 10 seconds downwards.

[by the way, I put the 'energy' of "let-through energy" in quotes, since I²t does not have the dimensions of energy]

I would also add that, for what it's worth, I have never experienced, or heard of anyone experiencing, a cutout fuse blowing as a result of a fault on a final circuit in a domestic installation. In fact, the only time I've been aware of a cutout fuse blowing as the result of a fault was when my daughter's RCD developed an L-N fault (when she pressed the test button, with a frightening result!) - upstream of any OPD in the installation, so the cutout fuse had to 'deal with it'!

Kind Regards, John

 

[John D v2.0]

domestic installations, I have never seen a PEFC appreciably above 1kA

True I checked at work on a random desk mounted socket and I think it was about 0.15 Zs, presumably the transformer is in the basement. The Zdb must be lower.

even if Ze were zero, the R1+R2 of any final circuit of appreciable length would probably be more than that

True but remember that faults can occur anywhere on a final circuit, and the Worst case for adiabatic is actally at the MCB rather than at the most distant point. So you have to use the Zdb/Ze and assume a very short length of CPC on the final circuit.

The let-through 'energy' of the cutout fuse (which hadn't blown) would obviously be irrelevant, but the let-through 'energy' of the OPD which cleared the fault would be highly relevant.

Yes of course, my point is the curves of I2t of a 32A MCB and a 100A cutout fuse cross over at some relatively high but not impossible current. So the cpc can be protected from melting by the cutout fuse if that is the lower I2t at the given current.

the availble curves for MCBs are always essentially 'vertical' from 10 seconds downwards

Yes you wouldn't use those curves above 500A or so, there are alternative ones which you can read the I2t directly off.

since I²t does not have the dimensions of energy

Interesting, didn't spot that, but clearly it makes sense upon reflection. I think the term was actually energy limiting, and let through ad I mixed them.

 

[JohnW2]

True but remember that faults can occur anywhere on a final circuit, and the Worst case for adiabatic is actally at the MCB rather than at the most distant point. So you have to use the Zdb/Ze and assume a very short length of CPC on the final circuit.

Agreed. However, with the possible exception of London etc., I suspect that Zes anything like as low as 0.023Ω (hence PEFCs ≥10kA, even if R1+R2 is zero) are virtually never seen in domestic installations (maybe, outside of London).

Yes of course, my point is the curves of I2t of a 32A MCB and a 100A cutout fuse cross over at some relatively high but not impossible current. So the cpc can be protected from melting by the cutout fuse if that is the lower I2t at the given current.

True in such a scenario but, as I've said, even if 'not impossible' it is probably a very long long way off anything I have personally ever encountered.

Yes you wouldn't use those curves above 500A or so, there are alternative ones which you can read the I2t directly off.

Not just wouldn't, but couldn't - those curves effectively become vertical at 5In, so are simply not usable above such a current. As for the curves from which one "can read the I²t directly off", I don't think they are routinely available to most electricians, so I confess to feeling a bit guilty when I/we accuse people of 'being lazy' by not doing the calculations!

Interesting, didn't spot that, but clearly it makes sense upon reflection. I think the term was actually energy limiting, and let through ad I mixed them.

Almost everyone seems to talk of I²t as being "let-through energy", and it clearly is a quantity which is of relevance in the context we are discussing - but, as I said, it is not 'energy'.

Kind Regards, John

 

[EFLImpudence]

True I checked at work on a random desk mounted socket and I think it was about 0.15 Zs, presumably the transformer is in the basement. The Zdb must be lower.

Even that is only 1.6kA

True but remember that faults can occur anywhere on a final circuit, and the Worst case for adiabatic is actally at the MCB rather than at the most distant point. So you have to use the Zdb/Ze and assume a very short length of CPC on the final circuit.

Uo/Zdb, yes. As above.
You can, of course, calculate values for near-end faults but in domestic situations it is unlikely the CPC will not be large enough (otherwise they would have not made it as small as it is) so more attention is given to far-end points to make sure the Zs is low enough.

Yes of course, my point is the curves of I2t of a 32A MCB and a 100A cutout fuse cross over at some relatively high but not impossible current. So the cpc can be protected from melting by the cutout fuse if that is the lower I2t at the given current.
Yes you wouldn't use those curves above 500A or so, there are alternative ones which you can read the I2t directly off.

I'm not sure where you're going with this.
For 32A MCBs anything over 160A will result in an instantaneous trip - usually taken as 0.1s but actually may be considerably quicker, 0.01s.
It is only generally below 0.1s (if that results in inadequate CPC size) that you need to go to the manufacturers let-through values and calculate that way.

 

[JohnW2]

I'm not sure where you're going with this. .... For 32A MCBs anything over 160A will result in an instantaneous trip - usually taken as 0.1s but actually may be considerably quicker, 0.01s. It is only generally below 0.1s (if that results in inadequate CPC size) that you need to go to the manufacturers let-through values and calculate that way.

Whilst the curves in BS 7671 do, indeed, stop at 0.1s, I've seen others which go down to 0.01s, and the 'curves' remain as essentially 'vertical straight lines' down that far (i.e. still at ~5In for a Type B) - so it may well be possible to use 0.01s.

Kind Regards, John