Impact of proposed reductions in 'maximum Zs' figures?

Many will be aware that proposals for Amendment 3 of BS7671:2008, due to appear next year, include a reduction in maximum permissible Zs figures (to achieve required disconnection times), so that the required disconnection times are still achieved when supply voltage is appreciably below the nominal voltage (230V).

The current proposal involves an adjustment factor (‘minimum voltage factor’, Cmin) of 0.95, which would mean that disconnection time requirements would be guaranteed to be satisfied at supply voltages as low as 218.5V. I have suggested to the authors of BS7671 that it would seem more logical to utilise a factor of 0.94, such that disconnection times would be achived down to the lowest permitted supply voltage of 216.2V (230V - 6%).

The following tabulation indicates what the new figures in BS7671 would (give or take rounding errors) look like (for Type B MCBs) with the current proposal (Cmin=0.95) and with my suggestion (Cmin=0.94). I also include a tabulation of what I assume would be the corresponding ‘OSG figures’, lower by a factor of 1.24, to take into account the fact that BS7671 gives figures for conductors at ‘operating temperature’ (often 70 °C), whereas the OSG gives figures for 10 °C.

I was wondering to what extent (if any) electricians feel that achieving these reduced ‘maximum Zs’ figures will represent a significant problem in practice.

Kind Regards, John

If you are doing a rewire, I don't see too bigger problem with dealing with the proposed Zs that will be required to keep within, at the design stage.
On older installation I suspect that a report would remark on the age of installation and the requirement of that time, rather than current. Plus if the addition of RCD protection were introduced, I would assume in most cases the Zs would be acceptable.

PrenticeBoyofDerry said:

If you are doing a rewire, I don't see too bigger problem with dealing with the proposed Zs that will be required to keep within, at the design stage. ... On older installation I suspect that a report would remark on the age of installation and the requirement of that time, rather than current.

Click to expand...

That all makes sense, but I wasn't really thinking that either of those situations (doing a rewire or a report) would be the biggest potential problem. I was thinking more about adding to/modifying existing circuits - is there then not theoretically a requirement for the Zs to be satisfactory (in terms of current regs) when measured from the new socket, light or whatever, regardless of when (and under what regs) the rest of the circuit was installed?

PrenticeBoyofDerry said:

Plus if the addition of RCD protection were introduced, I would assume in most cases the Zs would be acceptable.

Click to expand...

We've discussed this one many times before, and I find the regs a little unclear. It is certainly stated that RCDs should only be used for 'additional protection' and I'm not sure that it's within the spirit (even if within the word) of the regs to rely (in a TN installation) on an RCD to achieve required disconnection times which would otherwise not be achieved because of too high a Zs, is it? What do you think?

Kind Regards, John

Back in my college days (C&G236 don't you know) there were three methods of achieving an acceptable EFLI/disconnect time
1. use thicker conductors
2. use a shorter cable route
3. install an RCD

I'd have to unearth the 16th to see if the 'additional protection only' concept had been dreamt of back in the day…...

Taylortwocities said:

Back in my college days (C&G236 don't you know) there were three methods of achieving an acceptable EFLI/disconnect time
1. use thicker conductors
2. use a shorter cable route
3. install an RCD

Click to expand...

Indeed. I suppose the one possible issue is that, whilst (1) and (2) are hard and never-changing solutions, there seems to be this belief (correct or not) that the in-service reliability of RCDs is fairly poor.

Taylortwocities said:

I'd have to unearth the 16th to see if the 'additional protection only' concept had been dreamt of back in the day…...

Click to expand...

As I said, I don't find the current regs very clear. Although there is this 'additional protection only' business, there are also definitely places where it talks about achieving disconnection times with an RCD when/because they cannot be achieved with an OPD.

Of course, there is a subset of us, including me, for whom there is no issue, since we have TT installations! I'm sure that some would argue that if an RCD is considered to be a satisfactory primary (not 'additional') means of achieving disconnection times with TT, then so it should be with TN. Having said that, I suppose the counter-argument could be that an RCD is not truly 'satisfactory' but that with TT one has no better alternative!

Kind Regards, John

Firstly, I don't agree with allowing reliance on RCD (except unavoidable TT) rather than achieving satisfactory EFLI.

To the main subject; I don't see a problem with the new values.
After all, it only results in maximum circuit lengths being reduced by 7% or 8% or so (because of Ze being part of the circuit).

Perhaps cable should, at the same time be introduced with larger CPCs.

EFLImpudence said:

Perhaps cable should, at the same time be introduced with larger CPCs.

Click to expand...

Easy, just install 3-core flex instead of T&E.

(I'll get me coat).

EFLImpudence said:

Firstly, I don't agree with allowing reliance on RCD (except unavoidable TT) rather than achieving satisfactory EFLI.

Click to expand...

As I implied in my reply to TTC, I'm inclined to agree - fatter cable or shorter cable routing are totally foolproof and everlasting ways of ensuring that required disconnection times will be achieved 'for ever' (or as long as the OPD is functioning 'per spec'). There are far more uncertainties with RCDs.

EFLImpudence said:

To the main subject; I don't see a problem with the new values. After all, it only results in maximum circuit lengths being reduced by 7% or 8% or so (because of Ze being part of the circuit).

Click to expand...

I think that your mathematical reasoning has gone a little awry there Contrary to what you imply, the fact that Ze may be an appreciable contributor to the total Zs means that a small reduction in permitted maximum Zs can, in some situations, have quite a dramatic effect on maximum possible cable lengths....

... for illustration, take a fairly extreme example - of a TN installation with a Ze of, say 0.65Ω (not impossible) and a B50 MCB. Currently, the maximum ('OSG') Zs is 0.74Ω, leaving 0.09Ω for the R1+R2 of the cable. With the proposed (Cmin=0.5) change, maximum Zs will fall to 0.70Ω, leaving only 0.5Ω for the R1+R2 of the cable. The maximum possible cable length would therby be reduced by about 44%. That's a fairly extreme example, but if Ze is fairly high, the changes in the regs could easily reduce maximum possible cable length (for a given csa) by a lot more than the "7% or 8%" that you suggest - particularly if they accept my argument that Cmin should be 0.94 .

EFLImpudence said:

Perhaps cable should, at the same time be introduced with larger CPCs.

Click to expand...

That would certainly help to mitigate the effects of the change! If that every did happen, I imagine that there would be decades of confusion (and mistakes) whilst cables with two different CPC sizes were knocking around (paricular if mixed onthe same circuit!

Kind Regards, John

Taylortwocities said:

Easy, just install 3-core flex instead of T&E. (I'll get me coat).

Click to expand...

That would certainly help a lot (other than with 1mm²)

I've often pondered the question of why 3-core flex has, AFAIAA, always had a CPC csa equal to that of the live conductors, in contrast with the situation that was deemed to be 'adequate' for T+E. Do you know the answer?

Kind REgards, John

Another though/question .... I wonder why the BS7671 tabulations of 'maximum Ze at cable maximum operating temperature' actually exist, since it seems only sensible to use the 'OSG' (10 °C) figures if one wants an 'always safe' situation'

As I've said, the proposed 'maximum Zs' changes make total sense to me (or, they would if they gave Cmin as 0.94 ). It always seemed daft to me that one could design a BS7671-compliant circuit, only to find that the disconnection times were inadequate because the supply voltage was (as it is allowed to be) appreciably below 230V.

Similarly, it would seem daft to design a circuit that only achieved the required disconnection times if the cable was 'hot' - there is no law saying that a fault can't arise on a minimally-loaded (or totally unloaded), hence 'cold', cable.

It therefore seems only sensible to always use the OSG figures - so am I missing some reason why the BS7671 ones are as they are?

Kind Regards, John

JohnW2 said:

... for illustration, take a fairly extreme example - of a TN installation with a Ze of, say 0.65Ω (not impossible) and a B50 MCB. Currently, the maximum ('OSG') Zs is 0.74Ω, leaving 0.09Ω for the R1+R2 of the cable. With the proposed (Cmin=0.5) change, maximum Zs will fall to 0.70Ω, leaving only 0.5Ω for the R1+R2 of the cable. The maximum possible cable length would therby be reduced by about 44%. That's a fairly extreme example, but if Ze is fairly high, the changes in the regs could easily reduce maximum possible cable length (for a given csa) by a lot more than the "7% or 8%" that you suggest - particularly if they accept my argument that Cmin should be 0.94 .

Click to expand...

True, in an extreme example.

I was working on a more usual 0.2Ω Ze.

EFLImpudence said:

True, in an extreme example. I was working on a more usual 0.2Ω Ze.

Click to expand...

Fair enough, but the situation doesn't have to be all that extreme for the effect to be 'significant'. With a 'Cmin' of 0.95, if Ze were zero, then the reduction in permissible R1+R2 (hence cable length) would obviously be 5% (or 6% with Cmin=0.94), and that reduction would increase progressively as Ze rose above zero. I would imagine that it could become a potential issue with high current circuits in some TN-S installations.

Kind Regards, John

JohnW2 said:

Taylortwocities said:

Easy, just install 3-core flex instead of T&E. (I'll get me coat).

Click to expand...

That would certainly help a lot (other than with 1mm²)

I've often pondered the question of why 3-core flex has, AFAIAA, always had a CPC csa equal to that of the live conductors, in contrast with the situation that was deemed to be 'adequate' for T+E. Do you know the answer?

Kind REgards, John

Click to expand...

Maybe it's due to robustness I.e not to be damaged when flexed rather than electrical reasons.

AndyPRK said:

JohnW2 said:

I've often pondered the question of why 3-core flex has, AFAIAA, always had a CPC csa equal to that of the live conductors, in contrast with the situation that was deemed to be 'adequate' for T+E. Do you know the answer?

Click to expand...

Maybe it's due to robustness I.e not to be damaged when flexed rather than electrical reasons.

Click to expand...

Maybe - I certainly haven't yet thought of any electrical reason.

Kind Regards, John

JohnW2 said:

EFLImpudence said:

True, in an extreme example. I was working on a more usual 0.2Ω Ze.

Click to expand...

Fair enough, but the situation doesn't have to be all that extreme for the effect to be 'significant'. ... I would imagine that it could become a potential issue with high current circuits in some TN-S installations.

Click to expand...

OK - some chapter and verse, illustrating the extent of reduction in maximum permissible cable length with the proposed Cmin=0.95 (the reductions would obviously be a bit greater with 'my' proposed Cmin=0.94), using the 'OSG' (10 °C) 'maximum Zs' figures. It's probably only going to have a potentially large impact with shower, and some cooker, circuits (B40 and B50 MCBs) in installations with Ze greater than about 0.35Ω-0.4Ω. With higher Ze, as may sometimes be seen with TN-S supplies, there may be a little impact in relation to 32A circuits. The second graph is just a 'magnified' version of the first:

Kind Regards, John