Shower cable rating

[daggermark]

I have a 9.5kw shower and have just noticed that it is wired in on 6mm. By my calcs the shower should need 41.3A, so 6mm is pushing it given that a couple of metres of the run is in conduit.

But then I noticed it has a 32a breaker on it to protect the cable, which has never tripped - even when the shower is running on its highest setting. Can I assume that the shower is, for whatever reason, never running at full power?

 

[EFLImpudence]

I have a 9.5kw shower and have just noticed that it is wired in on 6mm. By my calcs the shower should need 41.3A, so 6mm is pushing it given that a couple of metres of the run is in conduit.

The 9.5kW will have been stated at 240V so it's 39.6A (38A at 230V) - but same difference.
There are time factors in the heating process.

But then I noticed it has a 32a breaker on it to protect the cable, which has never tripped - even when the shower is running on its highest setting. Can I assume that the shower is, for whatever reason, never running at full power?

No, the 32A breaker will carry 36.2A (32 x 1.13) for ever and never trip and 46.4A (32 x 1.45) for an hour before tripping so 40A will take a lot longer.
However, it might wear out eventually.

 

[daggermark]

Thanks, I never knew about the time factors.

Do you think this should be rewired with 10mm? The conduit run also contains 1 other circuit (2 cable ring). Also it isn't on the RCD side of the board, should it be?

 

[ericmark]

Most manufactures say their shower should be RCD protected, but when I bought this house only around 1/4 of the circuits were RCD protected, all the main house was still using fuses, some of the flat under the house was RCD protected, but only one RCD.

Now 14 x RCBO's so has 14 RCD's.

But the problem is the older boards were not made to take RCD's, then only designed to take one RCD, and then a pair of RCD's and finally designed to take all RCBO's (RCD and MCB combined) so to upgrade often means a new consumer unit.

So down to a risk assessment and is it worth the cost? Last house around 1992 my 14 year old son passed his RAE radio amateurs exam and I wanted to protect him, so fitted two RCD's before the two fuse boxes, and since then we have had bouts of trips, could go 2 years without a trip, or more, then in 2 months maybe 10 trips, no fault found, then another 2 years.

The problem was when we were not at home, and the freezer defrosted, and the more circuits on the single RCD the more likely it is to trip, in theory we should measure the back ground leakage which when all is running should be less than 9 mA, but I don't have the clamp-on meter to measure that, and we were told to test RCD's with all the MCB's turned off, so we, well at least I, used trust to the lord method to hope they did not trip, I feel it was not my fault, we were taught to test with all switched off, which means we had no idea if within that 9 mA or if DC imposed on the circuit would stop the RCD tripping, we never considered what a solar panel or EV charging point could do to a RCD tripping.

So this house when the roof leaked a RCBO tripped stopping the socket going on fire when the water got into the back, so yes some times good to have RCD protection, and since all RCBO's I could leave supply to sockets off until roof repaired, but had it been a RCD feeding loads of circuits it would have caused a real problem.

In fact it did with my mother, as CU under stairs so in a wheel chair she could not reach it to reset it.

Clearly I have all RCD protection, but I would not say using one or two RCD's is today the way to go, and to change a CU costs money, it took me 6 months to upgrade, and you need to consider if worth spending out for half a job, better to wait a little then get new CU with all RCBO's than have to dump a freezer full of food.

 

[daggermark]

I did wonder if RCBO might be an option, I believe my CU is compatible.

 

[EFLImpudence]

Do you think this should be rewired with 10mm?

No. The figure for 6mm² in conduit is actually 38A, so...

The conduit run also contains 1 other circuit (2 cable ring).

The ring circuit won't make any difference to worry about.

Also it isn't on the RCD side of the board, should it be?

Not if it isn't.

 

[SUNRAY]

No. The figure for 6mm² in conduit is actually 38A,

from 2004 brown book:
Table 4A1 - method 3
Table 4D2A - one 2 core cable 38A
Table 4B2 - 3 cables derating factor 0.7

38x0.7 = 26.6A

 

[EFLImpudence]

from 2004 brown book:

I only have the latest blue one to hand.

Table 4A1 - method 3

Please explain what that is in blue book.

Table 4D2A - one 2 core cable 38A

There you go.

Table 4B2 4C1 - 3 cables derating factor 0.7
38x0.7 = 26.6A

What about time? Is that not for long term 70° cables at maximum rating; not 80%?

The ring only counts as one, so - 0.8.

 

[sparkwright]

I did wonder if RCBO might be an option, I believe my CU is compatible.

I would think a 40amp RCBO would be a big improvement.

 

[SUNRAY]

I only have the latest blue one to hand.

I only have the old brown book.

Please explain what that is in blue book.

I only have the old brown book.

Table 4B2 4C1- 3 cables derating factor 0.7

4B1 apologies for the typo

What about time? Is that not for long term 70° cables at maximum rating; not 80%?

The ring only counts as one, so - 0.8.

The title of table 4B1 in the brown book: 'Correction factors for groups of more than one circuit of single-core cables, or more than one multicore cable...**'.
Then a column heading within the table: 'Number of circuits or multicore cables'.
Footnote to the table: '**When cables having differing conductor operating temperatures are grouped together, the current rating shall be based upon the lowest operating temperature of any cable in the group'.

I believe the ring cables should be classed as 2 cables as they are both current carrying cables and the table refers to: 'Number of circuits or multicore cables', there would be no dispute if it were to be two 16A radials.

The other 2 cables are socket circuit and not assessed for short term load. My interpretation is that applies to all cables in the group.

If those two ring circuit cables are to be classed as one then yes I agree factor of 0.8x38 = 30.4A which is still under the 32A OCPD.

 

[SUNRAY]

Let's not forget that also applies to the ring circuit cables, presumably 2.5mm² T&E.

Table 4A1 - method 3
Table 4D2A - one 2 core cable 23A
Table 4B1 - 3 cables derating factor 0.7

23x0.7 = 16.1A or 23x0.8 = 18.4A...

 

[JohnW2]

... What about time? Is that not for long term 70° cables at maximum rating; not 80%? ... The ring only counts as one, so - 0.8.

I believe the ring cables should be classed as 2 cables as they are both current carrying cables and the table refers to: 'Number of circuits or multicore cables' ....

I have to say that I have always regarded Table 4C1 (in current book) as being about as clear as a lot of the other mud in BS7671. By making a couple of words almost invisible, you have removed a fair bit of the confusion, but when one puts them back ('Number of circuits or multicore cables') (restoring what the book actually says) the intended meaning becomes far less clear.
...... After all, you could just as easily have modified it to be "Number of circuits or multicore cables".

Everything (including the applicability of any of the Table) becomes uncertain when one reads the first note, which reads:

NOTE 1: These factors are applicable to uniform groups of cables, equally loaded

... since, in the real world, one will probably rarely have groups of "equally loaded" cables, one has to wonder when, if ever, the 'factors' in the table are "applicable"!! ... and nor it is clear what one is meant to do if the tabulated factors are not "applicable"!

I know it rarely works, particularly when some of the ultra-cautious are around, but cannot one appeal for a but of electrical common sense? In fact, although I may be wrong, I strongly suspect that few (electricians) even consider cable grouping in the vast majority of domestic installations, not the least because such grouping often only exits for a few inches of a long cable (e.g. close to where it leaves a CU).

There is also, as EFLI has suggested, the issue of time. Shower circuits usually only carry (any!) current for very short periods of time (a few minutes), so I personally would not lose any sleep for not having 'de-rated' it's cable, or the other cables involved, if it were 'grouped' for at least some of its length with cables of other circuit(s) (none of which are likely to be 'fully loaded' for long periods of time, if ever, anyway).

Kind Regards, John

 

[SUNRAY]

I have to say that I have always regarded Table 4C1 (in current book) as being about as clear as a lot of the other mud in BS7671. By making a couple of words almost invisible, you have removed a fair bit of the confusion, but when one puts them back ('Number of circuits or multicore cables') (restoring what the book actually says) the intended meaning becomes far less clear.
...... After all, you could just as easily have modified it to be "Number of circuits or multicore cables".

Everything (including the applicability of any of the Table) becomes uncertain when one reads the first note, which reads:
... since, in the real world, one will probably rarely have groups of "equally loaded" cables, one has to wonder when, if ever, the 'factors' in the table are "applicable"!! ... and nor it is clear what one is meant to do if the tabulated factors are not "applicable"!

I know it rarely works, particularly when some of the ultra-cautious are around, but cannot one appeal for a but of electrical common sense? In fact, although I may be wrong, I strongly suspect that few (electricians) even consider cable grouping in the vast majority of domestic installations, not the least because such grouping often only exits for a few inches of a long cable (e.g. close to where it leaves a CU).

There is also, as EFLI has suggested, the issue of time. Shower circuits usually only carry (any!) current for very short periods of time (a few minutes), so I personally would not lose any sleep for not having 'de-rated' it's cable, or the other cables involved, if it were 'grouped' for at least some of its length with cables of other circuit(s) (none of which are likely to be 'fully loaded' for long periods of time, if ever, anyway).

Kind Regards, John

I started with the title which seems to make the distinction of multicores, or, circuits of singles. Deliberately turned round for emphasis.
Then the heading makes more sense.

If the consensus is to ignore the table, where do we stop ignoring the regs? Why not save money and use 4mm² (and a 40A OCPD) as that is rated at 40A (ref1) for the 42A (potential) load?
In fact why not abandon the regs altogether, use 2.5mm² and not bother with those expensive things in the box that cause so many questions on this forum? After all it is a resistive load that's impossible to draw over current and the DNO does provide a fuse.

 

[JohnW2]

I started with the title which seems to make the distinction of multicores, or, circuits of singles. Deliberately turned round for emphasis. .... Then the heading makes more sense.

[ it's slightly different in the current version of the regs, since the mention of "singles" has disappeared from both Table title and column headings ]

Anyway, you seem to be saying exactly the same as I did. You took a title which, as written, is difficult to make sense of (and/or is ambiguous) and modified it into something which "made more sense". However, that was just a decision on your part, which may or may not correspond with the intent of the regs. As I illustrated, you could have modified it in a different way, again resulting in something which made much more sense, but with a very different meaning. How can we tell which is correct?

If the consensus is to ignore the table, where do we stop ignoring the regs?

I'll talk a bit about the general situation below but, in the specific current context, as the disagreement between EFLI and yourself illustrated, one cannot even know whether or not one IS "ignoring the reg" (actually "ignoring guidance in a non-Normative Appendix of the regs"? We can presumably agree that, in the case of, say, a ring final, one can have "two multicore cables" which relate to "one circuit" - so, if it is just those two cables that are concerned, which column of 4C1 do we look at - the one headed "Number of circuits or multicore cables = 1" OR the one headed "Number of circuits or multicore cables = 2" ??

Also, as I asked, how can we work out if/when any of the grouping factors in the table are "applicable" (and hence whether one is "ignoring the guidance"), given ..

NOTE 1: These factors are applicable to uniform groups of cables, equally loaded.

??

So, it's not really a question of whether or not one "ignores" the guidance but, rather a problem in working out what the guidance is trying to say! In such an (unsatisfactory) situation, the best one can do is to try to make a judgment as to what the guidance is trying to say - and, as EFLI and yourself have illustrated, different people may come to different conclusions.

If the consensus is to ignore the table, where do we stop ignoring the regs? Why not save money and use 4mm² (and a 40A OCPD) as that is rated at 40A (ref1) for the 42A (potential) load? In fact why not abandon the regs altogether, use 2.5mm² and not bother with those expensive things in the box that cause so many questions on this forum? After all it is a resistive load that's impossible to draw over current and the DNO does provide a fuse.

In many senses, this is an inappropriate discussion to be having in this forum, since the 'official line' has to be that one should follow the regs (and maybe even accompanying 'Informative guidance') to the letter, and a good few electricians are also constrained (by conditions of employment and/or scheme membership) to do just that.

For a start, to be able to do that requires that the regs should be clear, unambiguous and comprehensive and, above all, that they should correspond with 'electrical common sense' - which, unfortunately, is sometimes very far from the case.

Has it ever occurred to you that the reasons for the Appendices of BS7671 are only "Informative Guidance" (rather than "Normative") might be to allow 'flexibility' (individual judgement) on the basis of 'electrical common sense'?

For example, in context, do you really think that it would be reasonable/necessary (or consistent with 'electrical common sense') to, say, substantially de-rate the CCC of a 30 metre length of cable because a couple of inches of it were 'grouped' with some other cables, particularly if the cable was never going to be carrying any current for more than a few minutes?

Kind Regards, John


-

 

[daggermark]

Wow this has generated quite the debate, thanks everyone for your comments.

Are the regs applicable relevant to the circumstances of the usage of the circuits? For example, as noted above, the shower is only ever on for 5-10 mins at a time. The ring it runs alongside in the ~2m conduit run is supplying some sockets in the conservatory, which are almost never used.