Shower circuit design.

I think that I have said before that there is no such thing as “nuisance tripping”, just bad design.

I agree that the B32 MCB is unlikely to trip during a shower of a normal duration, but that is not really the point.

I agree with Sparkingchip on this one.

davezawadi (David Stone): 
 

Why worry, it is not dangerous, and the breaker will almost certainly never trip. 

Well … I really don't know what to say about this.

The issue being, that OCPDs heat up. We simply don't have enough information to determine whether it's dangerous … if the mcb's either side of the shower breaker are often well loaded up to a good percentage of their capacity whilst the shower is running the shower breaker ABOVE it's capacity, we really don't know what will happen (except we do, because we've all seen overheating occuring at the CU for various reasons)!

davezawadi (David Stone): 
 

 the thermal section is pretty accurate, because it also needs to be very repeatable.

For the 32 amp MCB the thermal section tripping current is between 41 amps and 24.3 amps depending on the temperature, the stated 32 amps is what it is calibrated to at 30 degrees centigrade.

As well as the ambient temperature within the consumer unit enclosure you also need to consider the grouping factor. Inside enclosure the heating effect of MCBs installed in close proximity (in groups) needs to be taken into consideration. Certain amount of watt loss from each device will contribute to raise the ambient temperature of the breaker above the ambient temperature of the enclosure.

In reality a B40 MCB may have a trip current of around 38 amps just sufficient for a 8.5 kW shower depending on the location of the consumer and the temperature of it as well as what level of loading is on the adjoining MCBs , however on the other hand my consumer unit is mounted on a single brick exterior wall in an unheated garage and the MCBs may be barely adequate for circuit overload protection in the middle of the winter when the temperature of the devices in the consumer unit may be 5 degrees or less.

https://www.hagerelectro.com.au/files/download/0/32587_1/0/TECHINFO_MCBS.PDF

The design current of a circuit is not defined by the CPD Andy, as you know. There is also a great difference between “ambient temperature” and the MCB internally generated heat from the load current as must be fairly obvious. In a domestic CU, the ambient temperature is probably between 20 and 25 C in most houses. The dissipation from all the MCBs is a few Watts maximum, and in the case we are discussing, there is by definition very little other load. If you look at some large industrial boards with many fully loaded circuits the conditions may be somewhat different, but in a domestic the CU never feels hot does it?

The point of the thermal temperature rise applies to all MCBs and all sizes dissipate approximately the same power at the trip point, simply to achieve the same temperature. A bigger value one will get slightly less hot at 30ish amps, but this makes no important difference.

I don't really see why you are arguing so hard to justify your feeling that this is in some way dangerous, it is just a CPD operating in exactly the way designed into it. If the breaker gets warm, the operating current gets less, which is ideal for “safety” isn't it? It is possible that this CPD may trip, given enough time, but that is just a nuisance. My garage RCD trips sometimes due to the inrush to my compressor, I put up with it because the RCD is a good idea, but it is not “dangerous”. It is also not ideal, it is an inherent defect of RCDs that large currents sometimes cause spurious trips.

Following your reasoning the CPD for a ring circuit should not be 32A, it should be the maximum current that could be drawn from all the sockets, with cable size to match!  The point is that 32A protects the CABLE sufficiently, even if I plug in 50A of load. This is exactly the same as these showers.

Did I say it’s dangerous?

It is shoddy design and installation that fails to meet the basic design requirements and does not comply with the Wiring Regulations.

However I see installations installed by what I consider to be very competent electricians and each time I do see it I think to myself “Why did you do that, what was your reasoning?”.

Although I consider it completely wrong to use an under rated MCB I know it is accepted custom and practice, I also expected that I may be told that it is okay, I just didn’t expect such a highly technical argument to be put forward to prove it is not an issue.

Remember we are talking a about a protective device that costs less than three quid including VAT from many suppliers, the correct MCB I have ordered from Screwfix is £2.79 from Screwfix, it’s costing more in postage £5.00 to get it posted to the customers home in case I’m not the guy who goes back to to the remedial work, so that it’s onsite for whoever does go.

I was actually asked to do a 321 mile round trip to inspect the installation, because the customer have lost confidence in the original electrician and actually took the landing carpet and floor up to do a full inspection going beyond what is normally done for an EICR to check the cables under the floor.

Just walking in and seeing a MCG MCB in a Chint consumer unit when the correct device is available from Screwfix immediately red flags the job, when I then saw it was a 32 amp device for a 8.5 kW shower it got its second red flag, then I found the busbar pins had been cut to get the incorrect device to fit and it got its third red flag.

Three red flags and that is only from looking at the MCB.

Now I’m being told I’m over cautious and can lower one of the red flags.

davezawadi (David Stone): 
 

The design current of a circuit is not defined by the CPD Andy, as you know. There is also a great difference between “ambient temperature” and the MCB internally generated heat from the load current as must be fairly obvious. In a domestic CU, the ambient temperature is probably between 20 and 25 C in most houses. The dissipation from all the MCBs is a few Watts maximum, and in the case we are discussing, there is by definition very little other load. If you look at some large industrial boards with many fully loaded circuits the conditions may be somewhat different, but in a domestic the CU never feels hot does it?

The point of the thermal temperature rise applies to all MCBs and all sizes dissipate approximately the same power at the trip point, simply to achieve the same temperature. A bigger value one will get slightly less hot at 30ish amps, but this makes no important difference.

I don't really see why you are arguing so hard to justify your feeling that this is in some way dangerous, it is just a CPD operating in exactly the way designed into it. If the breaker gets warm, the operating current gets less, which is ideal for “safety” isn't it? It is possible that this CPD may trip, given enough time, but that is just a nuisance. My garage RCD trips sometimes due to the inrush to my compressor, I put up with it because the RCD is a good idea, but it is not “dangerous”. It is also not ideal, it is an inherent defect of RCDs that large currents sometimes cause spurious trips.

Following your reasoning the CPD for a ring circuit should not be 32A, it should be the maximum current that could be drawn from all the sockets, with cable size to match!  The point is that 32A protects the CABLE sufficiently, even if I plug in 50A of load. This is exactly the same as these showers.

I think the point that's omitted here, is that a low voltage assembly (CU) has a load diversity factor applied also. For example, if the assembly has a load diversity factor of 0.8, then groups of OCPDs should not exceed that to prevent overheating. If we have one mcb/RCBO that's running at a diversity factor of 1.1, then whilst it might not overheat on its own, two adjacent mcb's running at anything over 0.65 (65%) would tip the balance.

Sparkingchip: 
 

Now I’m being told I’m over cautious and can lower one of the red flags.

I'm not … research into accidents show that they happen because two or more things that could go wrong actually do: the “planets align” if you will.

With three flags, one of which “might be OK”, still leaves two.

Anyway, I don't necessarily agree with DZ(DS) that a minor overload on an OCPD “will probably be OK” - we don't have sufficient information to make that call (although you might Sparkingchip).

Why would an electrician install a 10 mm twin and earth circuit protected by a B32 MCB for a 8.5 kW shower?

Interesting thread, as you have “the shower circuit” with the OPD, the cable and how and what it is installed in, the size you choose, together with a isolator somewhere at 45 amps [or 50amps] rating, maybe, and you have the shower- being the fixed equipment. To match all aspects as a best fit “for fully complying with the regs” can be a struggle as Showers fail and get changed, perhaps as little as every five years. You could tie yourself in knots and, worst case, actually end up with something that does have potential for danger due to the cavorting with the intricacies of trying to be be prescriptive with individual requirements of BS7671. That is why the decision on these things is at install. You know what has been put where and how. There is no guess after the event, perhaps enagaging some kind of “regularisation” process, else you can end up chasing your  tail. Go for what is the least risky approach. I think somewhere in this thread, remedial work is mentioned. Was this upping the ante due to a “recommend improve”? Yes the busbar bodge is bad; that can be dealt with in isolation along with mixed manufacture at the same time, but there has to be 100% confidence in where, through what and how  the cable is installed when increasing the OPD . I am sure it has not been completely excavated from beginning to end to find out. 

So a one size fits all declaration of a 8.5Kw shower being on a 32 amp MCB is bad design cannot hold true. There may have been a very good reason why it is done so.

 It does not help that cable sizes for flat t/e are 6, 10 or 16mm and that most showers installed are 8.5kW +. 16mm cable for a shower circuit is madness and never done in a domestic for many obvious reasons. It was simpler when the only shower show in town was a 7 or 7.5kW. But when we are in the guess what may or may not happen territory in some endeavour to regularise some design after the event with only partial knowledge of what went on originally, you could foresee some unwanted consequence of changing to a 40 amp. 200mm of cable in a shower circuit with a ceiling switch can easily be fully enveloped. We accept this , 100mm in, 100mm out of the switch gives 200mm. Still fine on 40amp on 10mm. But what if its longer, say 400mm? So the original designer considered the cable rating a priority all things considered. There is nothing wrong with that. Next cable size up is not an option, nor is a ceiling isolator the size of a bread bin. It may be habit, to err on less risk approach of protect the cable or, knowing from experience, that when the shower fails the homeowner/ builder/ plumber sticks in a 9.5+ for bigger bang for your bucks. I am not saying that this is what happened in the OP , as we simply do not know all the facts, but simply the least risk, if the 32amp device is showing no signs of distress is too leave the rating as it was. Least risk for you and the cable. 

Zoomup: 

Chint. Oh 'ek.

Posters should be aware that any form of advertising is not allowed and should be removed by a moderator. Makers have a case for compensation from the I.E.T.!

P.S. Whatever happened to OMS and other valuable contributors who kept up the calibre of posts!  

Jaymack